Remembered Parenting Styles and Adjustment in Middle and Late Adulthood

Abstract

Background

Authoritative parenting is the parenting style often associated with positive outcomes for children and adolescents. This study considers whether remembered parenting styles in childhood predict multiple dimensions of functioning in adulthood.

Methods

We used the 1995 National Survey of Midlife Development in the United States data set (N = 2,232) to assess the association between parenting behaviors remembered from childhood—classified as authoritative, authoritarian, indulgent, and uninvolved—and psychological well-being, depressive symptoms, and substance abuse, in a subsample of mid- and later-life adults. Differences in outcomes by sex, race, and childhood socioeconomic status were also examined across parenting styles.

Results

Adults who remembered authoritative compared with authoritarian and uninvolved parents reported greater psychological well-being and fewer depressive symptoms, and those with uninvolved parents noted greater substance abuse. No outcome differences were found between remembered authoritative and indulgent parenting styles. A few sex and race interactions were identified: Authoritative parenting (compared with uninvolved) was more strongly associated with men's psychological well-being than women's, and authoritative parenting (compared with authoritarian) predicted reduced depressive symptoms for Whites more than non-Whites.

Conclusions

There is some support that remembered parenting styles continue to be related to functioning across the lifespan. There is also evidence of resiliency, flexibility, and malleability in human development.

DEVELOPMENTAL research demonstrates that authoritative parenting predicts many positive outcomes for children and adolescents when compared with authoritarian, indulgent, and uninvolved parenting (Coplan, Hastings, Lagace-Seguin, & Moulton, 2002; Steinberg, 2001). Authoritative parenting, characterized by moderate to high levels of control along with high levels of support, is related to greater psychological well-being and lower levels of depressive symptoms and substance use (Maccoby & Martin, 1983). Given current efforts to enhance adult well-being in later life and lifespan development theories that stress the influence of early childhood experiences well beyond adolescence (Elder, 1998), the link between early remembered parenting behavior and adjustment in middle and late adulthood is worthy of examination. Yet, little is known about the connection between parenting styles experienced in childhood and psychological and behavioral outcomes in adulthood.

The present study advances previous work by (a) assessing the association between remembered parenting styles and multiple indicators of mid- to later-life adjustment; (b) utilizing the four established parenting styles from the child development literature, instead of only focusing on one dimension of parenting (i.e., support or control) and its connection to a single outcome; (c) examining a large community sample of adults; and (d) identifying sex, race, and childhood socioeconomic status (SES) differences in the link between earlier remembered parenting behavior and later psychological and behavioral functioning.

LINKS BETWEEN DEVELOPMENT AND PARENTAL SUPPORT AND CONTROL

Most theories of family socialization view parenting behavior as multidimensional, involving primarily the domains of parental support and control (Baumrind, 1971; Maccoby & Martin, 1983), which are both considered critical to children's development. Parental support is important because it predicts a strong sense of self-worth and security, greater psychological well-being, and other positive outcomes (Coplan et al., 2002; Steinberg, 2001). Without adequate parental support, children are likely to experience anxiety, insecurity, aggression, hostility, low self-esteem, and inadequacy (Baumrind).

Parental control helps to shape responsible conformity and self-control in children. The rules and guidelines parents set and enforce teach children about group and societal standards for behavior (Baumrind, 1996). Over time, children's experience with rules and the consequences for breaking them help them to develop independent decision-making skills and to internalize control of their own behavior (Baumrind). Inadequate parental control, on the other hand, may make it difficult for children to manage effectively in the outside world where behavioral rules and standards of conduct exist in nearly all social settings (Barber & Olsen, 1997). In combination, therefore, parental support and control are believed to foster children's emotional, psychological, and behavioral well-being and development.

Based on a cross-classification of high and low levels of parental support and control, child development scholars have developed typologies of parenting styles. Consistent with the work of Maccoby and Martin (1983), we distinguish among four “styles of parenting” based on the balance between high and low levels of parental responsiveness (i.e., support) and demandingness (i.e., control). Authoritative parents display high levels of both responsiveness and demandingness. They are warm, nurturing, and sensitive to their child's needs and consistently consider the child's age and maturity when forming behavioral expectations. Authoritarian parents display low responsiveness and high demandingness. They are often cold, unsupportive, insensitive to the child's needs, and demanding in their control. Indulgent parents use high responsiveness but low demandingness; they believe that fewer rules and expectations for appropriate behavior benefit children's development. Finally, uninvolved parents display low levels of both responsiveness and demandingness. They generally fail to connect with their child, are emotionally detached and withdrawn, and have few rules and expectations. Rather than the child-centered motivations that characterize indulgent parenting, the reasons behind uninvolved parenting tend to be related to parental psychopathology or other problems.

Developmentalists have empirically validated the importance of parenting style for numerous child and adolescent outcomes. (This literature is briefly summarized subsequently.) Yet, despite life course theorizing that early life experiences can have a protracted influence on later development (Elder, 1998), no attempts have been made to link early parenting styles to developmental outcomes in mid- and later adulthood. Parental support during childhood is known to predict psychological well-being, self-worth, and personal control in young adults (Wickrama, Lorenz, & Conger, 1997). Further, adults reporting positive interactions with parents during childhood are more likely to enter into social relationships that are supportive, foster well-being, and contribute to better health and less depression (Shaw, Krause, Chatters, Connell, & Ingersoll-Dayton, 2004). Hence, an empirical foundation exists for testing a wider array of long-term outcomes that may emerge from different styles of early parenting.

EVIDENCE OF THE IMPORTANCE OF PARENTING STYLES FOR CHILDREN AND ADOLESCENTS

Consistent with theoretical expectations, empirical work indicates that authoritative parenting is generally associated with the most beneficial child and adolescent outcomes when compared with nonauthoritative parenting styles (Coplan et al., 2002; Steinberg, 2001). Children and adolescents reared by authoritative parents compared with nonauthoritative parents report better developmental outcomes, including greater self-esteem, self-control, moral development, social maturity, and school performance (Radziszewska, Richardson, Dent, & Flay, 1996), as well as greater psychological well-being and less depression and substance use (Shah & Waller, 2000).

Demographic Differences in Parenting Styles and Outcomes

Researchers have raised two important questions regarding variations in parenting styles as they relate to subgroup differences: First, do preferred parenting styles vary across racial/ethnic and social class groups and are different parenting styles used with males versus females? Second, do the effects of various parenting styles differ across these groups? To address these questions as they relate to adult development, our study examines sex, race, and childhood SES differences in the relationship between remembered parenting style and psychological well-being, depressive symptoms, and substance abuse in mid- and later-life adults.

Studies of children and adolescents indicate that preferred parenting styles may differ somewhat based on parents' race/ethnicity and SES. Authoritative parenting tends to be more common among Whites and authoritarian parenting more typical of Blacks and Hispanics (Radziszewska et al., 1996). Interestingly, Asian Americans are said to be more permissive when children are very young but to shift toward more authoritarian parenting as they reach “the age of understanding” (Garcia Coll & Pachter, 2002, p. 198).

SES differences may be confounded with some of the findings attributed to race/ethnicity (Kelley, Power, & Wimbush, 1992). Several studies of Black parents use low-income samples, which are generally more authoritarian than high-income ones (Radziszewska et al., 1996). Further, many studies noting a dominant pattern of authoritative parenting among Whites use middle-class samples, calling into question their applicability to more diverse samples (Bornstein & Bornstein, 2007). Specifically, high- versus low-SES parents are more likely to use warmth, stimulation, explanations, and praise with their children and to focus on children's internal characteristics (e.g., autonomy). Conversely, low-SES parents generally focus more on children's external behaviors (e.g., obedience) (Hoff, Laursen, & Tardif, 2002).

In terms of race/ethnic variations in parenting style outcomes, much research has addressed the conditions under which authoritarian parenting may be more desirable than, and even preferred over, authoritative parenting. A point often made is that authoritarian parenting styles used by low-income minority parents living in high-risk environments may actually benefit children's safety and development and promote school achievement (Baldwin, Baldwin, & Cole, 1990). Finkelstein, Donenberg, and Martinovich (2001) also found that Black adolescent girls whose mothers used high control reported fewer depressive symptoms than Latino and White girls. Thus, for some populations, authoritarian rather than authoritative parenting may be more advantageous for child and adolescent outcomes.

Findings on sex differences in child and adolescent outcomes as they relate to parenting style are inconsistent; some studies report interactions (Hart, Newell, & Olsen, 2002), and others note that parenting styles have similar effects for both sexes (Finkelstein et al., 2001). When differences are noted, patterns are consistent with the behavior problems typical of each sex. Girls tend to exhibit more internal problems (e.g., shyness, somatic complaints), whereas boys’ problems tend to be displayed through external behaviors enacted against others or their environment (Hart et al., 2002). Thus, less than ideal parenting appears to exacerbate problems in the behavioral domains in which each sex is already at risk.

EARLY PARENTING BEHAVIOR AND OUTCOMES IN ADULTHOOD

Investigations of parenting behavior experienced in childhood and subsequent adult development are often limited to parental discipline and considered in association with a single negative outcome, such as personality disorders, depressive symptoms, or substance abuse (Carter, Joyce, Mulder, & Luty, 2001; Sakado, Kuwabara, Sato, Uehara, Sakado, & Someya, 2000). For instance, adults who recalled overcontrolling parents that used negative discipline reported greater anxiety in adulthood (Gerlsma, Emmelkamp, & Arrindell, 1990). Parents' use of either excessive or inadequate control also predicted increased depressive symptoms in adulthood (Sakado et al., 2000).

Limited work has examined the relationship between adult outcomes and recalled parenting styles that encompass both parental support and control. Moreover, studies generally include a single negative outcome and frequently are limited to clinical samples. Using a sample of clinic outpatients along with a healthy community sample, Shah and Waller (2000) considered the risk of depressive symptoms as they related to both remembered parental care (e.g., warmth) and overprotection (e.g., control). Depressed adults were more likely than healthy ones to recall their parents as uncaring and overprotective. Further, greater parental care rather than less overprotection was associated with lower vulnerability for depressive symptoms.

Carter et al. (2001) investigated the link between remembered parental care and overprotection and later personality dysfunction in depressed outpatients. They found that recollections of low parental care predicted uncooperativeness, use of immature defenses, and more symptoms of personality disorder compared with parental overprotection. Parker, Boak, Griffin, Ripple, and Peay (1999) also observed that patients in their study (most of whom were depressed) who recalled their parents as highly indifferent (low care) and overcontrolling were more likely to be diagnosed with personality disorders, anxiety, depressive symptoms, and self-defeating behavior problems. In contrast, patients recalling more optimal parenting—defined as high affection and balanced levels of control—were least likely to be diagnosed with any of these dysfunctions.

McCrae and Costa (1988) further found that adults who recalled high parental support were less likely to report neuroticism and scored high on extraversion, agreeableness, and openness to new experiences. Additionally, adults who recalled demanding parents were more conventional, conscientious, and assertive. Finally, adults who remembered more indulgent parents developed into confident but self-centered adults. These authors caution, however, that effect sizes for these findings were small, and thus parenting experienced in childhood may have a limited effect on adult personality. Instead, they posit that environmental contexts, genetic factors, and social processes might interact and influence personality in adulthood.

Given this literature, the current study employs the four parenting styles widely used in the child literature based on a combination of high and low levels of support and control to consider the association between parenting styles and multiple dimensions of functioning in mid- and later-life adults. The following hypothesis is tested: Adults who remember parenting behaviors classified as authoritative report greater psychological well-being, fewer depressive symptoms, and less substance abuse than adults who remember parenting behaviors categorized as authoritarian, uninvolved, or indulgent. We also ask whether sex, race, and childhood SES moderate the connection between remembered parenting styles and psychological well-being, depressive symptoms, and substance abuse, controlling for variables that are known to be associated with adult functioning.

METHODS

Data

The analyses used secondary data from the 1995 National Survey of Midlife Development in the United States (MIDUS), which was designed to assess psychological well-being, physical health, and social responsibility in midlife. Using random digit dialing, MIDUS data were collected via 40-minute telephone interviews (response rate 70%) followed by self-administered mail-back questionnaires (response rate 87%). The sample was limited to English-speaking, noninstitutionalized adults between the ages of 25 and 74 years (N = 4,244).

Sample

Given our focus on the link between psychological and behavioral functioning and childhood parenting styles for middle-aged and older adults, we limited the analytic sample to respondents aged 40 and older. This included 1,138 men and 1,094 women who completed both the phone interview and relevant sections of the self-administered questionnaire. The mean age of the respondents was 54 years (SD = 9.51). The majority (89%) of respondents were non-Hispanic White, and most (66%) were married (19% separated/divorced, 9% widowed, and 6% never married). Regarding their level of education, 10% had less than a high school degree, 50% had a high school degree or GED, 27% had either a 2- or 4-year college degree, and 13% had a master's degree or higher. Reported annual incomes were highly variable—43% earned <$15,000, 7% earned >$75,000, and the modal (21%) income category was $30,000–$49,999.

Independent Variables

Remembered parenting styles from childhood.—

Four parenting style categories were developed based on the parenting dimensions of responsiveness and demandingness, as laid out by Maccoby and Martin (1983). (These two distinct dimensions of parenting were confirmed by a factor analysis conducted on the nine parenting items described subsequently.) First, we constructed the parental support and control scales separately for mothers and fathers. The support scale included six items, referring to the years the respondent was growing up, “How much did your mother/father understand your problems and worries? How much could you confide in her/him about things that were bothering you? How much love and affection did she/he give you? How much time and attention did she/he give you when you needed it? How much effort did she/he put into watching over you and making sure you had a good upbringing? How much did she/he teach you about life?” Responses ranged from 1 = a lot to 4 = not at all. Items were reverse coded so that higher scores indicate greater parental support. The remembered parental support scale was created by calculating the mean across the six items, similar to the approach used by other researchers (Ryff, Singer, & Palmersheim, 2004). For households with two parents, the mean was calculated across the maternal and paternal support scale (alpha was .91), consistent with the extensive work of Steinberg, Lamborn, Darling, Mounts, and Dornbusch (1994) on parenting styles.

The parental control scale consisted of three items that asked about mother and father separately: “How strict was she/he with her/his rules for you? How consistent was she/he about the rules? How much did she/he stop you from doing things that other kids your age were allowed to do?” Responses ranged from 1 = a lot to 4 = not at all. Again, items were reverse coded so that higher scores indicate greater parental control, and the scale was created by calculating the mean across the three items (Ryff et al., 2004). For households with two parents, the mean was calculated across the maternal and paternal control scales (alpha was .79).

Next, we used the median split on the two parenting dimension scales to establish low and high levels of support and low and high levels of control. Finally, we developed four parenting style categories based on distinct combinations of the levels of parental support and control: high support and high control parenting were classified as authoritative (29%), low support and high control were classified as authoritarian (19%), high support and low control were categorized as indulgent (21%), and low support and low control were classified as uninvolved (31%).

Based on other categorization approaches used in the child development literature, we also ran the analyses using a tertile split (Steinberg et al., 1994). That is, we divided the support and control scales into low, medium, and high levels of support and control. Then, only the upper and lower tertiles were used to create the four parenting styles, and respondents falling into the medium tertiles (nearly 50%) were deleted from the analyses. This resulted in a sample of 1,118 adults. The tertile split is often used in child studies because it assures four very distinct parenting styles (Steinberg et al). Authoritative parenting included respondents in the highest tertiles on both support and control, authoritarian parenting included those scoring in the lowest tertile on support and highest tertile on control, uninvolved parenting included those scoring in the lowest tertiles on both support and control, and indulgent parenting included those scoring in the highest tertile on support and lowest tertile on control. Results based on this classification scheme were similar to those obtained using the median split. One exception was that adults who remembered uninvolved parents reported greater anxiety compared with adults who remembered authoritative parents using the median split but not the tertile split. To maintain the more complete sample and increase generalizability of the findings and power, we chose to report the results obtained using the median split approach.

Dependent Variables

Psychological well-being.—

Psychological well-being was measured using a mean composite score of the six dimensions of Ryff's (1989) shortened psychological well-being scale, consistent with the approach used by other researchers (e.g., Marmot & Fuhrer, 2004). The use of a composite score also reflects our interest in “overall” psychological well-being rather than particular subscales of well-being. Each scale consisted of three items: (a) positive relations with others (e.g., “I have not experienced many warm and trusting relationships with others”), (b) self-acceptance (e.g., “I like most parts of my personality”), (c) autonomy (e.g., “I tend to be influenced by people with strong opinions”), (d) personal growth (e.g., “For me, life has been a continuous process of learning, changing, and growth”), (e) environmental mastery (e.g., “In general, I feel I am in charge of the situation in which I live”), and (f) purpose in life (e.g., “Some people wander aimlessly through life, but I am not one of them”). Responses ranged from 1 = strongly agree to 7 = strongly disagree. Mean imputation was used to calculate the sum score when an item was missing. Items were reverse coded so that higher scores indicate greater psychological well-being (alpha was .76).

Depressive symptoms.—

Depressive symptoms were assessed by summing responses to seven yes/no questions regarding depressed mood, energy, appetite, sleep, concentration, feeling down, and suicidal thoughts (e.g., “During two weeks in the past 12 months, when you felt sad, blue, or depressed, did you lose interest in most things?”). Higher values indicate more depressive symptoms (alpha was .53). Depressive symptoms were assessed with the World Health Organization's Composite International Diagnostic Interview Short Form, which has shown good test–retest reliability and clinical validity (Kessler, Andrews, Mroczek, Ustun, & Wittchen, 1998).

Substance abuse.—

First, alcohol abuse was assessed via yes/no responses to five alcohol abuse screening items (e.g., “Did you ever, during the past 12 months, have such strong desire or urge to use alcohol that you could not resist it or could not think of anything else?”). Second, illegal drug abuse was assessed via yes/no responses to five drug use screening items (e.g., “Did you ever, during the past 12 months, have any emotional or psychological problems from using these [illegal] substances?”). The substance abuse scale was created by adding the number of yes responses to the alcohol and drug abuse questions (alpha was .76).

Control Variables

Age, marital status, education, and income were used as control variables in the analyses because past work reports differences in psychological and behavioral functioning based on these variables. Older adults, married persons, and those with higher education and income tend to exhibit greater psychological and behavioral functioning (Brown, Bulanda, & Lee, 2005; Rieker & Bird, 2005; Savla, Almeida, Davey, & Zarit, 2008). A dummy variable, “unmarried,” was created for marital status with 0 = currently married (66%) and 1 = not currently married (34%); age, education, and income were used as continuous variables. We used a mean imputation for missing income values (6%); results did not differ when using the imputed variable compared with analyses where cases with missing values were omitted.

Moderator Variables

Analyses also tested the effects of sex, race, and childhood SES, along with their interaction with parenting style as predictors of adult functioning. Sex was coded 0–1, with 1 representing women. Race was coded 0–1 with 1 representing non-Whites. Childhood SES would ideally be based on parents' education, income, and/or occupation. However, in the MIDUS data, we were limited to a single item asking, “When you were growing up, were you better off or worse off financially than the average family was at that time?” (1 = a lot better off, 7 = a lot worse off). This item was reverse coded to indicate higher childhood SES level.

Data Analyses

To test our hypothesis and answer our research question, we conducted separate hierarchical multiple regressions for psychological well-being (Table 1), depressive symptoms (Table 2), and substance abuse (Table 3). For all dependent variables, we included as control variables in Model 1: age, marital status (unmarried), education, and income. These factors are not central to our findings so we do not discuss findings from Model 1 in the results. In Model 2, we added the parenting styles to test our hypothesis. To evaluate whether sex (women), race (non-White), and childhood SES moderate the relationship between parenting styles and the dependent variables, we first added the main effects in Model 3 and then the nine interaction terms in Model 4 (i.e., sex by the three dummy-coded parenting styles, race by the three dummy-coded parenting styles, and childhood SES by the three dummy-coded parenting styles). Given that the large number of interactions in Model 4 may limit statistical power, we ran each moderating variable in separate models as well as together. Because there were no differences in the results, we present the findings for all interactions together in Model 4.

Table 1.

Hierarchical Regression Coefficients for Psychological Well-Being (N = 2,231)

	Model 1		Model 2		Model 3		Model 4
	B	SE B	B	SE B	B	SE B	B	SE B
Control variables
Age	0.03***	0.01	0.02***	0.01	0.03***	0.01	0.03***	0.01
Not married	−0.67***	0.11	−0.64***	0.11	−0.68***	0.11	−0.67***	0.11
Education	0.16***	0.02	0.16***	0.02	0.16***	0.02	0.16***	0.02
Income	0.03***	0.01	0.03***	0.01	0.03***	0.01	0.03***	0.01
Parenting style
Authoritariana			−0.96***	0.14	−0.97***	0.14	−2.03*	0.80
Indulgenta			0.07	0.13	0.07	0.13	−0.56	0.72
Uninvolveda			−1.01***	0.13	−1.01***	0.13	−3.08***	0.71
Women					0.18	0.10	−0.09	0.17
Non-White					0.02	0.16	−0.20	0.24
Childhood SES					0.02	0.04	−0.04	0.07
Interactions
Authoritariana × Women							0.32	0.28
Indulgenta × Women							0.19	0.26
Uninvolveda × Women							0.54*	0.25
Authoritariana × Non-White							0.28	0.44
Indulgenta × Non-White							0.11	0.43
Uninvolveda × Non-White							0.69	0.40
Authoritariana × Childhood SES							0.06	0.12
Indulgenta × Childhood SES							0.05	0.11
Uninvolveda × Childhood SES							0.13	0.10
R2	.07		.12		.12		.12
F	43.53***		43.46***		30.43***		16.76***

Notes:

^aComparison group is authoritative parenting.

^*p < .05; ***p < .001.

Table 2.

Hierarchical Regression Coefficients for Depressive Symptoms (N = 2,231)

	Model 1		Model 2		Model 3		Model 4
	B	SE B	B	SE B	B	SE B	B	SE B
Control variables
Age	−0.03***	0.00	−0.02***	0.00	−0.02***	0.00	−0.02***	0.00
Not married	0.53***	0.09	0.52***	0.09	0.50***	0.09	0.50***	0.09
Education	−0.00	0.01	−0.00	0.01	−0.00	0.01	−0.00	0.01
Income	−0.02***	0.00	−0.02***	0.00	−0.02***	0.00	−0.02***	0.00
Parenting style
Authoritariana			0.31***	0.11	0.30**	0.11	−0.96	0.61
Indulgenta			0.12	0.09	0.10	0.09	0.09	0.47
Uninvolveda			0.29**	0.09	0.29**	0.09	0.04	0.45
Women					0.13	0.08	0.07	0.12
Non-White					−0.07	.11	−0.39**	0.13
Childhood SES					0.00	0.03	0.05	0.04
Interactions
Authoritariana × Women							0.35	0.21
Indulgenta × Women							−0.23	0.18
Uninvolveda × Women							0.15	0.18
Authoritariana × Non-White							0.84*	0.35
Indulgenta × Non-White							0.52	0.34
Uninvolveda × Non-White							0.22	0.26
Authoritariana × Childhood SES							−0.05	0.09
Indulgenta × Childhood SES							−0.05	0.08
Uninvolveda × Childhood SES							−0.06	0.07
R2	.04		.05		.05		.06
F	17.28***		11.56***		8.88***		5.44***

Notes:

^aComparison group is authoritative parenting.

*p < .05; **p < .01; ***p < .001.

Table 3.

Hierarchical Regression Coefficients for Substance Abuse (N = 2,231)

	Model 1		Model 2		Model 3		Model 4
	B	SE B	B	SE B	B	SE B	B	SE B
Control variables
Age	−0.01***	0.00	−0.01***	0.00	−0.01***	0.00	−0.01***	0.00
Not married	0.16***	0.04	0.16***	0.04	0.21***	0.04	0.20***	0.04
Education	0.01	0.01	0.01	0.01	0.01	0.01	0.01	0.01
Income	0.00	0.00	0.00	0.00	−0.00	0.00	−0.00	0.00
Parenting style
Authoritariana			0.03	0.04	0.04	0.04	0.22	0.26
Indulgenta			0.02	0.04	0.02	0.04	0.17	0.28
Uninvolveda			0.09*	0.04	0.09*	0.04	0.62*	0.31
Women					−0.22***	0.04	−0.15**	0.05
Non-White					−0.03	0.05	0.04	0.09
Childhood SES					0.00	0.02	0.00	0.02
Interactions
Authoritariana × Women							−0.15	0.08
Indulgenta × Women							−0.03	0.08
Uninvolveda × Women							−0.13	0.09
Authoritariana × Non-White							−0.08	0.14
Indulgenta × Non-White							−0.12	0.17
Uninvolveda × Non-White							−0.18	0.15
Authoritariana × Childhood SES							0.04	0.03
Indulgenta × Childhood SES							0.01	0.04
Uninvolveda × Childhood SES							−0.03	0.04
R2	0.3		0.3		0.5		0.5
F	13.64**		7.98***		7.12***		4.10***

Notes:

^aComparison group is authoritative parenting.

*p < .05; **p < .01; ***p < .001.

RESULTS

Psychological Well-Being

Model 2 in Table 1 shows that adults who remembered authoritarian or uninvolved parenting as children reported lower psychological well-being compared with adults who recalled their parents being authoritative. There were no differences in well-being between those who had indulgent compared with authoritative parents. Model 4 in Table 1 indicates that sex was the only variable that moderated the relationship between well-being and remembered parenting style with regard to uninvolved versus authoritative parenting. Post hoc analysis indicated that although men and women recalling uninvolved parents did not differ in well-being, the benefits of authoritative parenting for well-being were significantly greater for men than women.

Depressive Symptoms

Model 2 in Table 2 indicates that adults who remembered authoritarian or uninvolved parents reported more depressive symptoms than did adults who remembered authoritative parents. There was no difference in depressive symptoms between respondents remembering indulgent compared with authoritative parents. Model 4 in Table 2 indicates that only race moderated the relationship between depressive symptoms and parenting style with regard to the authoritarian versus authoritative parenting groups. Post hoc analysis indicated no race differences among respondents remembering authoritative parenting, yet among those who remembered authoritarian parenting Whites reported more depressive symptoms than did non-Whites. Therefore, compared with authoritative parenting, authoritarian parenting appears to contribute to more depressive symptoms for Whites but not non-Whites.

Substance Abuse

Model 2 in Table 3 shows that adults who remembered uninvolved parents reported more substance abuse compared with those who remembered authoritative parents. There were no differences in substance abuse for adults in either the authoritarian or the indulgent parenting group compared with those in the authoritative parenting group. Model 4 in Table 3 shows that neither sex, nor race, nor childhood SES moderates the relationship between substance abuse and any of the parenting styles.

DISCUSSION

Ample evidence shows that authoritative parenting predicts more positive outcomes for children and adolescents than other types of parenting styles (Coplan et al., 2002; Steinberg, 2001). This study advances that body of work by assessing whether established parenting styles from childhood—authoritative, authoritarian, indulgent, and uninvolved—as remembered by adults, predict psychological well-being, depressive symptoms, and substance abuse in midlife and beyond. We also tested whether sex, race/ethnicity, and childhood SES moderate the association between remembered parenting styles and the multiple dimensions of adult functioning. Our findings add to the literature on parenting styles in childhood and functioning in mid- to later life by providing evidence that remembered authoritative parenting is linked with more positive adult outcomes compared with recalled authoritarian and uninvolved parenting behaviors. However, in only two situations did sex and race act as moderators.

Remembered Parenting Styles and Adult Functioning

We found some support for our hypothesis that adults who remember parenting behaviors classified as authoritative report greater psychological well-being, fewer depressive symptoms, and less substance abuse than adults who remember parenting behaviors categorized as authoritarian, uninvolved, or indulgent. There were significant effects of parenting style for psychological well-being and depressive symptoms, but only regarding authoritarian and uninvolved parenting compared with authoritative parenting. As expected, adults who remembered more authoritarian or uninvolved parenting reported lower psychological well-being and more depressive symptoms than those remembering authoritative parenting in childhood. Additionally, adults recalling uninvolved parenting in childhood rather than authoritative parenting reported greater substance abuse. These findings, which are consistent with the child development literature and studies of adults (Parker et al., 1999; Shah & Waller, 2000; Steinberg et al., 1994), support principles of the life course perspective (Elder, 1998) by suggesting that positive outcomes associated with authoritative parenting in childhood can extend into adulthood—at least as they relate to remembered parenting styles—and merit continued investigation.

Our finding that parenting style was less influential in adults' reports of substance abuse than psychological well-being and depression was unexpected but may reflect the lifespan development notion that human development is remarkably plastic and shaped by a diverse set of influences (Baltes, 1987). The life course perspective also suggests that life experiences may provide opportunities for change that can redirect trajectories such that positive functioning in adulthood can be achieved even if early experiences are less than ideal (Elder, 1998). For instance, as persons move into adulthood, their range of experiences and exposure to nonparental socializing agents expands dramatically. These new situations provide alternative influences that may contribute to development, thereby reducing the link between early parenting and diverse outcomes. For persons experiencing nonoptimal parenting early in life and problems in childhood and adolescence, formation of positive relationships and introduction to stable socializing experiences as adults can be major turning points (Uggen & Massoglia, 2003).

The fact that we found fewer significant parenting style differences in relation to substance abuse than to our other outcomes could also be due to the MIDUS sampling frame. MIDUS used a community sample that excluded institutionalized adults. Given the link between criminal behavior and alcohol and drug abuse (Goode, 2007), the heaviest drinkers and drug abusers may be more involved in crime and other negative behaviors that remove them from the general population and thus our sample. Adults not included in the MIDUS survey may have selected themselves out through illegal behavior and imprisonment or institutionalization for mental health problems. Selection bias may have reduced our ability to reveal significant associations.

In addition, and contrary to our expectations, there were no outcome differences between adults who remembered authoritative parents and those who remembered indulgent parents. This is also inconsistent with findings that children and adolescents whose parents are considered indulgent versus authoritative have less positive outcomes (Steinberg, 2001). One possibility is that high levels of support provide a buffer for future negative outcomes, even in cases where children lack firm parental discipline. For example, children whose parents are supportive share values similar to their parents and are more open to parental influence (Steinberg, 2008). Thus, even if parents are supportive but lack control, their children's long-term behavior may not be a problem because over time they may model their supportive parents’ behavior and adopt prosocial ways. Additionally, as they mature, children may benefit from nonfamilial environments that provide or require discipline and control, such as school and jobs. Thus, even children who lack early parental control may eventually develop responsibility in accord with the expectations of nonfamilial settings.

Moderators Between Parenting Styles and Adult Functioning

We also tested whether sex, race/ethnicity, and childhood SES moderate the relationship between the four parenting styles and psychological well-being, depressive symptoms, and substance abuse. No clear and overarching patterns were found. Both sex and race revealed one significant interaction, but none were found for childhood SES. Comparisons between adults recalling authoritative versus nonauthoritative parents indicated that authoritative behaviors are predictive of enhanced psychological well-being and fewer depressive symptoms. These benefits, however, were limited to particular subgroups of the adult population. Authoritative parenting appears to benefit men more than women, but only in terms of greater psychological well-being, and authoritative parenting benefits Whites more than non-Whites regarding reduced depressive symptoms.

Similar to our findings, sex and race differences in the child development literature are ambiguous in terms of parenting styles and outcomes (Hart et al., 2002). Some researchers note sex and race differences in outcomes based on parenting styles (e.g., Hart et al.), whereas others report limited or no sex and race differences (Finkelstein et al., 2001). Regarding the lack of race differences in our study, it should be noted that the vast majority of participants were non-Hispanic White, which may have masked finer group differences. Studies with larger numbers of racial and ethnic groups may increase the statistical power to detect significant differences in adult functioning. Moreover, the lack of a moderating effect for childhood SES may be related to our measure. MIDUS included few questions regarding respondents’ parents, especially pertaining to information from childhood. A more comprehensive measure that includes parents' education, income, and financial status from childhood may reveal differences in the link between adult functioning and remembered parenting styles.

Another limitation of our study is the retrospective accounts of parental behavior in childhood. The validity of childhood memories is often called into question because respondents' present status (e.g., psychological well-being) may shape childhood memories (e.g., Halverson, 1988). However, we do not attempt to treat our retrospective data as a proxy for longitudinal data or as a means for establishing causality. Henry, Moffitt, Caspi, Langley, & Silva (1995) also noted that retrospective measures “may constitute valid indicators of the individual's current perception of those features [of interest to social scientists], and as such, may be useful in understanding psychological development or adjustment” (p. 93). In addition, some evidence suggests that concerns about retrospective childhood data are exaggerated (Brewin, Andrews, & Gotlib, 1993). McCrae and Costa (1988) found strong concordance in reports of recalled parenting styles in childhood between siblings, which provides some evidence of the accuracy of retrospective accounts of parenting styles.

Our sole reliance on adults' reports of their parents' behavior during childhood also poses a limitation. However, Block (1971) showed a strong correlation between retrospective accounts of childhood experiences and objective assessments by mothers that were taken in childhood. Schwartz, Barton-Henry, and Pruzinsky (1985) also found adolescents’ reports of parental behavior to be more accurate than those of parents. Another limitation is that our decision to use the framework of Maccoby and Martin (1983) for identifying parenting typologies begs the question of whether our results would be similar using the approach of Baumrind (1971) to defining parenting styles. Specifically, Baumrind used both psychological control and behavioral control to distinguish authoritarian parents from other parents. Our data, unfortunately, did not address psychological control, which prevented us from using Baumrind's typology. Future researchers should consider various approaches to developing parenting typologies to investigate the relationship between remembered parenting styles and diverse psychological and behavioral outcomes in adulthood.

Despite these limitations, this study is the first to investigate the link between four typologies of remembered parenting styles from childhood and psychological and behavioral functioning in mid- to late adulthood. Despite generally small effect sizes, our results provide some support that remembered parenting styles are related to adult functioning, which supports findings in the child development literature and adult literature based on clinical samples and single outcomes. In addition, our findings seem to suggest that human development is characterized by resiliency, flexibility, and malleability, at least for community dwelling adults. It may be that some adults, at least those living in community settings, are able to rise above negative influences in childhood by seeking out or attracting environments, opportunities, and relationships that promote adult functioning. More research is clearly warranted in this area to gain a better understanding of factors (both current and retrospective) that enhance positive adult development.