Conflict Strategies in the Parent–Adult Child Tie: Generation Differences and Implications for Well-Being

Abstract

Objectives

Irritations often arise in intergenerational ties. Middle-aged individuals report that their relationships with adult children are more important and more negative than ties with aging parents. However, it is unclear whether midlife individuals use different interpersonal conflict strategies with adult children compared to aging parents, and whether the strategies used have implications for psychological well-being.

Method

This study examined middle-aged individuals’ reports of conflict strategies with their adult children and their aging parents and their own depressive symptoms. Participants were from the Family Exchanges Study; middle-aged adults (N = 365, ages 45 to 66 years) reported on the conflict strategies used with each of their adult children and their aging parents.

Results

Models revealed that middle-aged individuals use more active strategies (e.g., discussing problems) with their adult children than their aging parents. In contrast, individuals used more passive strategies (e.g., avoidance) with aging parents than adult children. Further, passive strategies used with adult children are associated with greater depressive symptoms.

Discussion

Findings are consistent with the intraindividual stake hypothesis and imply that conflict strategies used with adult children may be more consequential for psychological well-being than those used with aging parents.

Depressive symptoms, Generation differences, Intergenerational conflict, Parent-child

Interpersonal tensions (i.e., irritations, upsets) in the parent–adult child tie predict poor relationship quality and reduced well-being (Birditt, Miller, Fingerman, & Lefkowitz, 2009). The strategies used in response to tensions have implications for parent–child relationship quality (Belsky, Jaffee, Hsieh, & Silva, 2001; Caughlin & Malis, 2004; Cichy, Lefkowitz, & Fingerman, 2013; Robin & Weiss, 1980). For example, middle-aged parents and their adult offspring are most likely to report constructive strategies (e.g., discussing problems), which are associated with better relationship quality (Birditt, Rott, & Fingerman, 2009). Conversely, destructive strategies (e.g., avoiding problems) appear to be harmful for the relationship. Earlier research also indicates that there are variations in the types of strategies used by each generation, with middle-aged parents reporting the use of more constructive strategies than their young adult offspring report (Birditt, Rott, et al., 2009). However, it is unclear whether middle-aged adults use different strategies with aging parents compared to adult children and whether the strategies used have implications for psychological well-being.

The present study contributes to the literature by examining conflict strategies middle-aged adults use with older and younger generations and assessing the associations between these strategies and psychological well-being. The broader literature on parent–child conflict has focused on the offspring’s childhood and adolescence (e.g., Bradford, Vaughn, & Barber, 2008; Caughlin & Malis, 2004; Caughlin & Ramey, 2005), neglecting an understanding of these processes later in the life course. This study focuses on the unique position of middle-aged individuals as they relate to both their aging parents and their young adult children. Midlife is a critical time to examine parent–child conflict because middle-aged individuals encounter new experiences and stressors associated with the launching of their young adult children and the increasing dependency of their aging parents (Fingerman, Pillemer, Silverstein, & Suitor, 2012) that may heighten intergenerational conflict and its adverse consequences for well-being.

Theoretical Framework

Overall, there are several theoretical perspectives in the literature that explain variations in the parent–child tie by generation, family role, and age. The intergenerational stake hypothesis suggests that there are generation differences in relationship quality and that parents are more emotionally invested in the parent–child relationship than are their children (Bengtson & Kuypers, 1971). In this case, generation is defined as family generational position (i.e., child vs. parent) and not birth cohort (Bengtson & Allen, 1993). Parents view their children as continuations of themselves and thus perceive more positive feelings and less negative feelings in this relationship, whereas children desire greater independence from parents and are more invested in enhancing differences. Older and middle-aged parents typically report greater investment in ties to children, greater closeness, and greater positive relationship quality regarding their children than their children report with them (Aquilino, 1999; Shapiro, 2004). Also consistent with the intergenerational stake hypothesis, children tend to report greater conflict and more negative relationship quality than do their parents (Aquilino, 1999; Fingerman, 2001). These negative qualities include the extent to which parents and children get on one another’s nerves, criticize the other person, and make too many demands on one another. Hence, the intergenerational stake hypothesis posits that parents and children tend to feel and behave differently toward one another because of different levels of commitment within the parent–child tie.

Building on the intergenerational stake hypothesis, Birditt and colleagues developed the concept of the intraindividual stake hypothesis (Birditt, Hartnett, Fingerman, Zarit, & Antonucci, 2015). This perspective extends the intergenerational stake hypothesis which refers to variations within a parent–child dyad by generation to a within-person context that considers one’s simultaneous roles as both a parent and an adult child. The intraindividual stake hypothesis suggests that individuals make distinctions within themselves, by role, in terms of how they feel about their offspring and their parents. Furthermore, this perspective suggests that individuals are more invested in their own children than they are in their own parents. These feelings of greater investment may influence the types of intergenerational conflict strategies that people use. Moreover, strategies used with children may have a stronger effect on well-being than strategies used with aging parents because of greater feelings of investment in ties with one’s own offspring.

The social input model (Fingerman & Charles, 2010) suggests that people use different strategies depending on the age of their social partners. In particular, when individuals of any age interact with older individuals, they use strategies that reduce tension and encourage positive emotions (Fingerman, Miller, & Charles, 2008). These preferential behaviors with older adults may stem from a variety of beliefs such as the perception that time remaining in these relationships may be limited, that older adults are unlikely to change to their annoying behaviors, or that older adults themselves will reciprocate with tension-reducing behaviors. Because individuals may be more likely to hold such beliefs about parents than children, they may use different types of strategies with their parents than with their children.

This study considers whether conflict strategies vary by role in terms of whether individuals are interacting with their children or with their parents (i.e., generational position of family member). Conflict strategies vary in terms of whether they are constructive or destructive and active or passive (Rusbult, Bissonnette, Arriaga, Cox, & Bradbury, 1998). Constructive strategies are intended to be positive for relationship maintenance, whereas destructive strategies may be detrimental to relationships. Active strategies are direct (e.g., confronting problems), whereas passive strategies are indirect (e.g., avoiding problems). We examine four types of strategies in the present study which are defined as active constructive, passive constructive, active destructive, and passive destructive. Active constructive strategies include discussion of the problem, whereas passive constructive strategies may involve turning the other cheek or accepting the problem as it is. Active destructive strategies often include arguing or yelling. In contrast, passive destructive strategies employ potentially harmful tactics such as avoiding talking about problems or withdrawing from the situation.

Differences in Conflict Strategies Across Generations

It remains unknown whether generation differences in investment within parent–adult child ties extend to conflict strategies. Recent research shows that individuals report that their ties with their own adult children are more important but also more negative than ties with their own aging parents (Birditt et al., 2015). It is unclear, however, how these findings apply to conflict strategies. Research examining two generations of parents and adult children showed that mothers and fathers reported using constructive strategies during conflict within these relationships more often than their adult children (Birditt, Rott, et al., 2009). Consistent with the intragenerational stake hypothesis, middle-aged individuals may use more constructive and less destructive strategies with their offspring than their parents because they are more invested in those ties than they are with their aging parents.

Middle-aged individuals may also report more passive strategies with parents and less passive strategies with adult children due variations in the amount of future time perceived remaining with adult children versus aging parents. Research shows that people are more likely to use active destructive strategies with younger social partners and more likely to use avoidant strategies with older social partners (Fingerman, Miller, & Charles, 2008). Similar findings have emerged in studies using vignettes (Miller, Charles, & Fingerman, 2009). On the basis of previous theoretical and empirical work, we posit that middle-aged adults may use more active conflict strategies with their adult children and more passive strategies with their aging parents.

Associations Between Conflict Strategies in Ties With Parents and Adult Children and Midlife Individuals’ Well-Being

The intraindividual stake hypothesis also suggests that because individuals are more invested in their children than their parents, conflict with adult children may have a stronger effect on psychological well-being than conflict with aging parents. We focus on depressive symptoms as a measure of well-being in this study as it is often highly linked with negative aspects of social relationships (e.g., Sacco, 1999; Steger & Kashdan, 2009; Teo, Choi, & Valenstein, 2013). Studies have examined links between relationship quality with parents and adult children and individuals’ well-being. Umberson (1992) assessed associations between negative relationship quality with mothers, fathers, and children older than 16 years and depressive symptoms. When relationships with parents and children were examined in the same model, greater negative relations with mothers and adult children predicted greater depressive symptoms. Birditt et al. (2015) examined relationship quality with both children and parents and their implications for middle-aged individuals’ depressive symptoms. Negative relationship quality with parents and not with adult children was associated with greater depressive symptoms among middle-aged individuals.

Reports of conflict or negative interactions with parents are associated with poorer well-being among young adult and middle-aged children (Barnett, Kibria, Baruch, & Pleck, 1991; Fingerman, Pitzer, Lefkowitz, Birditt, & Mroczek, 2008; Polenick, DePasquale, Eggebeen, Zarit, & Fingerman, 2016). Middle-aged children and their aging fathers, for example, report elevated depressive symptoms when they perceive more negative qualities (e.g., criticism, demands) in the father–child tie (Polenick et al, 2016). A less studied possibility is that the use of conflict strategies with aging parents and adult children is consequential for well-being. That is, how middle-aged adults handle conflict in their ties with older parents and grown children (e.g., active vs. passive strategies) may have positive or negative implications for their own well-being. Conflict strategies middle-aged adults use with aging parents and with adult children may have important implications for well-being; but may demonstrate different patterns across generations. Because individuals tend to be more invested in their relationships with their children than their parents, we predict that strategies used with adult children will have a greater impact on middle-aged individuals’ well-being than strategies used with parents.

Present Study

Taking a multigenerational approach, this study examined conflict strategies used by middle-aged adults in their relationships with aging parents and adult offspring. We also considered whether these strategies are linked to depressive symptoms at midlife. We addressed two main questions:

• 1)

  Do middle-aged adults use different conflict strategies with their adult children and their aging parents?

We predicted that middle-aged adults use more active and constructive strategies with their adult children but more passive strategies with their aging parents.

• 2)

  Do conflict strategies with adult children and aging parents have implications for middle-aged adults’ psychological well-being?

We predicted that conflict strategies used with adult children would be more strongly linked with middle-aged adults’ depressive symptoms than conflict strategies used with aging parents.

Method

Participants

Participants were middle-aged adults from the Family Exchanges Study (FES) Wave 2 conducted in 2013. The original FES sample of White and Black participants was recruited from the Philadelphia Metropolitan Statistical Area (PMSA) using a listed sample of 12,500 phone numbers purchased from Genesys. Phone numbers were randomly selected from the list and participants were eligible to participate if they were aged 40 to 60 years and had at least one living parent and one child older than 18 years. A total of 633 middle-aged individuals completed phone interviews in Wave 1 in 2008. Participants received 30 dollars for completing Wave 1. In Wave 2, which was conducted in 2013, interviewers attempted to recontact all respondents and a total of 488 of those middle-aged participants completed the phone interview. Participants received 30 dollars for completing the interview.

In this study, participants were included if they had both a living parent and at least one adult child in Wave 2. The sample included a total of 380 middle-aged adults (56% women). These respondents were asked about their mother and father as well as up to four children aged 18 and older. The respondents reported on a total of 490 parents and 895 adult children. We excluded 15 participants due to missing data on conflict variables for parents or children, depressive symptoms, or any of the covariates. The final sample included 365 participants who reported on 471 parents and 861 adult children (see Table 1 for sample description).

Table 1.

Background Characteristics of Middle-Aged Participants

Study variable	%	M (SD)
Gender (female)	57.5
Married	71.0
Minority	34.5
Coresiding with a parent or child	46.8
Age		55.48 (4.92)
Years of education		14.43 (1.98)
Family size		2.84 (1.48)
Neuroticism		2.60 (0.78)

Note: Middle-aged adult N = 365.

Measures

Conflict strategies

Middle-aged participants rated the extent to which they used the following strategies when they were upset with their mother, father, and each adult offspring: passive constructive (I accept there is nothing I can do about problem); active constructive (I calmly discuss the problem with (parent/child)); passive destructive (I avoid talking about the problem with (parent/child)); active destructive (I argue or fight with (parent/child)). Each item was rated on the following scale: 1 (not at all); 2 (a little); 3 (somewhat); 4 (quite a bit); 5 (a great deal).

Depressive symptoms

Participants completed five items from the depression scale of the Brief Symptom Inventory (Derogatis & Spencer, 1993). Participants reported the extent to which they experienced the following problems in the past 7 days: loneliness, feeling blue, having no interest in things, hopelessness about the future, and worthlessness. The answer choices included: 1 (not at all); 2 (a little); 3 (moderately); 4 (quite a bit); 5 (extremely). The items were averaged to create a mean (alpha = .87).

Covariates

We controlled for middle-aged participants’ gender, marital status, race, coresidence, education, neuroticism, and their total number of children. These variables are linked with relationship quality and conflict, and are often used as covariates in studies on the parent–child tie (Fingerman, 2001; Birditt, Miller, et al., 2009; Fingerman, 2001; Cooney & Dykstra, 2013; Gunthert, Cohen, & Armeli, 1999; Suitor, Sechrist, Gilligan, & Pillemer, 2011; Umberson, 1992). Gender was coded as 1 (man) or 0 (woman). Marital status was coded as 1 (married/remarried) or 0 (not married). Race was coded as 1 (racial/ethnic minority) or 0 (non-Hispanic White). Coresidence was coded as 1 (coresiding with a parent or offspring) or 0 (not coresiding with a parent or offspring). We included the number of years of education obtained. Four items assessed neuroticism (i.e., moody, worrying, nervous, calm; Lachman & Weaver, 1997). Participants rated how well each of these items described themselves using the scale: 1 (not at all), 2 (a little), 3 (some), 4 (a lot), averaged to create a neuroticism score (alpha = .70). Family size referred to the total number of children, which is similar to definitions of family size used in previous studies (e.g., Grundy & Read, 2012).

Analysis Strategy

We first estimated descriptive statistics and bivariate correlations. Next, to examine whether middle-aged people used different strategies with their adult children than their aging parents, we estimated three-level multilevel models using SAS proc mixed. Level 1 included family member, which was nested within generation (level 2), which was nested within the middle-aged participant (level 3). Generation was the predictor (1 = adult child; 0 = aging parent) and each conflict strategy type was the outcome. The covariates included middle-aged participants’ gender, marital status, education, race, coresidence, neuroticism, and number of children. There were a total of 1,323 family members with data on at least one conflict strategy (1,323 for argue, 1,320 for discuss, 1,321 for avoid, 1,316 for accept). A total of 365 middle-aged individuals reported a maximum of 6 (2 parents and 4 adult children) family members each. Thus, we included three levels to account for the fact that there might have been correlations within individuals (level 3) in how they felt about their family members (level 1) and that there may be a correlation within individuals in how they felt about their multiple children and their mother and father (level 2). Because previous research suggests that the levels of the multilevel model should include larger sample sizes than the current study (Bell et al., 2010), we estimated the model again with two levels (level 1 = family members; level 2 = participants) rather than three, but the fit of the two-level model was lower than the fit of the three-level model. Because the fit of the model was better with three levels, and the findings were the same for both models, we used the three-level model.

We then estimated a series of four linear regression models to assess whether the conflict strategies used with parents and children predicted middle-aged individuals’ depressive symptoms. We used linear regression equations because the outcome was at the level of the participant and not the family member. The predictors involved the average conflict strategy used with either adult children or parents. We examined each conflict strategy type separately. Thus, one model examined the use of active constructive strategies with parents and with adult children as predictors of depressive symptoms, followed by models that assessed the use of active destructive, passive constructive, and passive destructive strategies. The covariates included participant gender, marital status, education, race, coresidence, neuroticism, and family size.

We tested a series of post hoc models that examined whether middle-aged individuals used different conflict strategies with their mother, father, daughters, and sons. We did this by first estimating the multilevel models separately by generation and testing the effects of family members’ gender. We also estimated multilevel models to examine whether coresiding predicted conflict strategies by generation. Next, we estimated linear regression models assessing the links between conflict strategies and depressive symptoms again with average conflict strategies for mother, father, sons, and daughters separately to assess whether links with depressive symptoms varied by gender. Each model controlled for the same conflict strategy used with the other generation. For example, the model testing conflict strategies with mother included the rating of avoidance with mother and avoidance with children as predictors. Finally, we estimated the models predicting depressive symptoms in Wave 2 with depressive symptoms at Wave 1 as a covariate. Thus, we accounted for the variance in middle-aged adults’ depressive symptoms that was explained by their previous levels of depressive symptoms.

Results

Descriptives

An examination of the means of conflict strategies revealed that middle-aged individuals were most likely to report using active constructive strategies and least likely to report using active destructive strategies with both aging parents and adult offspring (Figure 1).

Figure 1.

Middle-aged adult reports on conflict strategies with their aging parents and adult children. *Estimated means are from multilevel model predicting conflict strategies as a function of generation.

Bivariate correlations between how middle-aged adults coped with problems with parents and adult children ranged from .22 for passive constructive strategies to .26 for active destructive strategies (Table 2). In terms of correlations between the types of conflict strategies used with adult children, the use of passive strategies (both constructive and destructive) was most highly associated with a correlation of .34. Similarly, in terms of the strategies used with parents, passive strategies were most highly correlated at .40.

Table 2.

Pearson Correlations Among Study Variables

Variable	1	2	3	4	5	6	7	8
1. Parent passive constructive
2. Parent active constructive	−.002
3. Parent passive destructive	.40***	−.24***
4. Parent active destructive	.14**	.10	.004
5. Child passive constructive	.22***	.05	.08	.04
6. Child active constructive	.12*	.24***	.01	.02	−.04
7. Child passive destructive	.11*	−.08	.24***	.01	.34***	−.26***
8. Child active destructive	.03	.03	−.01	.26***	.02	−.06	.23***
9. Depressive symptoms	.13*	.04	.15**	.04	.15**	−.10	.26***	.18***

Notes: Middle-aged adult N = 365.

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

Unrestricted multilevel models revealed that there was a significant amount of variance between and within middle-aged individuals in how they coped with problems across parents and adult offspring. For active destructive (argue), 19% of variance was between participants, 13% was between parents and children within participant, and 68% was within participant and error. For passive destructive (avoidance), 13% of the variance was between participants, 27% was within participants between parents and children, and 60% was within participants and error. For active constructive (discuss), 17% of the variance was between participants, 23% was within participant between parents and children, and 59% of the variance was within participant and error. For passive constructive (accept), 15% was between participants, 32% was within participants between parents and children, and 53% was between family members within participant and error. Overall, there was greater variance within individuals in how they reported coping with family members than there was between unrelated participants.

Variations in Conflict Strategies Used With Adult Children and Parents

We estimated multilevel models predicting each conflict strategy type with generation (coded 1 = adult child, 0 = aging parent) as a predictor (Table 3). The models showed that individuals were more likely to use active strategies (i.e., discuss, argue) with their adult children than their parents. In contrast, individuals were more likely to use passive strategies (i.e., accept, avoid) with their parents than their adult children.

Table 3.

Multilevel Models Examining Generation Differences in Conflict Strategies with Adult Children and Aging Parents

	Passive constructive	Active constructive	Passive destructive	Active destructive
Variable	B (SE)	B (SE)	B (SE)	B (SE)
Intercept	2.74 (.44)***	2.44 (.40)***	1.48 (.37)***	0.88 (.28)**
Generationa	−0.41 (.07)***	0.30 (.06)***	−0.42 (.06)***	0.23 (.05)***
Covariates
Education	0.01 (.02)	0.05 (.02)*	0.03 (.02)	0.00 (.01)
Coresidenceb	−0.06 (.09)	0.09 (.08)	−0.01 (.08)	0.17 (.06)**
Genderc	0.01 (.09)	0.07 (.08)	−0.15 (.08)*	−0.07 (.06)
Neuroticism	0.08 (.06)	−0.09 (.05)	0.13 (.05)*	0.22 (.04)***
Family size	−0.02 (.03)	−0.00 (.03)	0.00 (.03)	−0.04 (.02)*
Marital statusd	−0.07 (.11)	−0.02 (.10)	0.07 (.09)	0.03 (.07)
Racee	−0.07 (.11)	−0.06 (.10)	−0.04 (.09)	−0.05 (.07)
Covariance parameters
Between participants (level 3)	0.25 (.06)***	0.22 (.05)***	0.16 (.04)***	0.10 (.03)***
Within participants by generation (level 2)	0.41 (.07)***	0.25 (.06)***	0.24 (.06)***	0.07 (.03)*
Within participants by family member (level 1)	0.81 (.05)***	0.76 (.04)***	0.70 (.04)***	0.48 (.03)***
−2 log likelihood	4,072.3	3,889.3	3,773.9	3,117.4

Notes: Middle-aged adult N = 365. Covariates represent participant level characteristics.

^a1 = adult child, 0 = aging parent.

^b1 = coresiding with parent or offspring, 0 = not coresiding with parent or offspring.

^c1 = man, 0 = woman. ^d1 = married/remarried, 0 = not married.

^e1 = racial/ethnic minority, 0 = non-Hispanic White.

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

Associations Between Strategies Used With Parents and Adult Children and Middle-Aged Individual’s Depressive Symptoms

Next, we estimated a series of linear regression models to assess whether depressive symptoms varied by the strategies used with aging parents and adult children (Table 4). Overall, strategies used with adult children, but not with parents, were associated with depressive symptoms. In particular, the use of passive strategies (i.e., accept, avoid) with adult children was associated with greater depressive symptoms among middle-aged individuals. Passive strategies used with parents were not associated with depressive symptoms. In addition, active strategies (i.e., discuss, argue) used with adult children and parents were not associated with depressive symptoms.

Table 4.

Linear Regression Models Examining Middle-Aged Adults’ Depressive Symptoms as a Function of Conflict Strategies with Adult Children and Aging Parents

	Passive constructive	Active constructive	Passive destructive	Active destructive
	B (SE)	B (SE)	B (SE)	B (SE)
Conflict with adult children	0.08 (.04)*	−0.07 (.04)	0.20 (.05)***	0.06 (.05)
Conflict with aging parents	0.04 (.03)	0.06 (.03)	0.05 (.03)	−0.04 (.05)
Covariates
Education	−0.04 (.02)*	−0.03 (.02)	−0.04 (.02)*	−0.04 (.02)*
Coresidencea	0.14 (.07)*	0.12 (.07)	0.12 (.07)	0.11 (.07)
Genderb	0.11 (.07)	0.10 (.07)	0.15 (.07)*	0.11 (.07)
Neuroticism	0.33 (.05)***	0.34 (.05)***	0.30 (.05)***	0.32 (.05)***
Family size	−0.01 (.02)	−0.01 (.02)	−0.01 (.02)	−0.01 (.02)
Marital statusc	−0.16 (.08)	−0.17 (.08)*	−0.19 (.08)*	−0.17 (.08)*
Raced	−0.01 (.08)	−0.03 (.08)	−0.01 (.08)	−0.02 (.08)
R2	.20	.19	.24	.18
F test	9.98***	9.29***	12.21***	8.88***

Notes: Middle-aged adult N = 365. Columns refer to the specific strategy used in response to conflict.

^a1 = coresiding with parent or offspring, 0 = not coresiding with parent or offspring.

^b1 = man, 0 = woman.

^c1 = married/remarried, 0 = not married.

^d1 = racial/ethnic minority, 0 = non-Hispanic White.

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

Post Hoc Models

We first assessed whether conflict strategies varied by gender of the family member. Individuals used more active destructive strategies with mothers than with fathers (b = −.15, SE = .06, p = .018). There were no variations in strategies used with daughters and sons.

We also estimated multilevel models to examine whether coresiding predicted conflict by generation. Coresiding with adult offspring predicted greater use of both active constructive and active destructive strategies with those offspring (b = .19, SE = .09, p = .033; b = .36, SE = .08, p = .001).

Next, we examined whether links between conflict and depressive symptoms varied by gender of the family member. Conflict strategies with father did not predict depressive symptoms. Discussing problems with mother predicted greater depressive symptoms (b = .07, SE = .04, p = .037). There were no other significant associations between conflict strategies used with mother and depressive symptoms.

The models estimated separately for sons and daughters revealed that when conflict was with sons, accepting the problem (passive constructive; b = .10, SE = .04, p = .006) and avoiding talking about problems (passive destructive; b = .12, SE = .05, p = .009) predicted greater depressive symptoms. Avoiding talking about problems with daughters predicted greater depressive symptoms (b = .18, SE = .05, p = .001). There were no other significant associations between conflict strategies used with daughters and depressive symptoms.

We estimated the linear regression models predicting depressive symptoms with depressive symptoms in Wave 1 as a covariate and the same findings emerged. Middle-aged individuals who reported using passive constructive and passive destructive strategies with adult children reported greater depressive symptoms (b = .07, SE = .03, p < .040; b = .15, SE = .04, p = .001, respectively).

Discussion

According to the intraindividual stake hypothesis, middle-aged individuals may consider their relationships with children to be more important than those with aging parents. Previous studies have examined variations in how middle-aged individuals feel about these ties with regards to relationship quality and investment but have yet to consider conflict strategies which are key determinants of both relationship functioning and well-being. In this study, we build on the current literature by exploring the conflict strategies middle-aged adults use in relationships with their aging parents and their adult children. We considered whether there were generation differences in conflict strategies as well as their implications for the psychological well-being of middle-aged individuals. Consistent with the intraindividual stake hypothesis (Birditt et al., 2015), differences in conflict strategy use were observed on the basis of generation (i.e., generational position of family member). Furthermore, strategies used with adult children appear to be more consequential for well-being than those used with aging parents.

Differences in Conflict Strategies Across Generations

Regarding cross-generational differences in middle-aged adults’ use of conflict strategies in the parent–adult child relationship, middle-aged individuals reported using more active strategies (e.g., discussing or arguing about the problem) with their adult offspring than with their aging parents. Conversely, middle-aged individuals reported using more passive strategies (e.g., accepting or avoiding the problem) with aging parents than with adult offspring. There are at least three potential explanations for this pattern of findings.

First, in accord with the intragenerational stake hypothesis (Birditt et al., 2015), middle-aged adults may be more motivated to actively confront problems with their adult children than with their aging parents. In other words, individuals may use more active conflict strategies with their adult children because they are more heavily invested in these ties. Conversely, passive strategies may be more commonly used in relationships with parents because these ties involve relatively lower levels of investment. Bolstering this possibility, prior studies of married couples suggest that less emotionally invested partners tend to avoid or withdraw from conflict discussions (Eldridge & Christensen, 2002). Hence, cross-generational differences in active and passive conflict strategies may be partly attributed to levels of investment within parent–adult child ties. This generational difference may be due, in part, to parents’ focus on guiding and nurturing their offspring, which contrasts with children’s desire to establish autonomy and reduce parental control. Although parental control likely diminishes during the offspring’s adulthood, such dynamics may persist and carry over into the use of conflict strategies within parent–adult child ties.

Second, according to the social input model, individuals use different strategies depending on the age of social partners. In particular, individuals use more passive strategies with older social partners than with younger social partners. These variations may stem from a variety of beliefs such as the perception that time remaining in these relationships may be limited, that older adults are unlikely to change to their annoying behaviors, or that older adults themselves will reciprocate with tension-reducing behaviors. (Fingerman, Miller, et al., 2008). Thus, middle-aged people may use passive strategies with parents because they perceive time as more limited with parents or they believe that parents are unlikely to change their behavior. Indeed, research on adolescent children and their parents, for example, has found that parents engage in more active conflict strategies than their children, whereas children tend to use more passive strategies than their parents (Caughlin & Malis, 2004; Caughlin & Ramey, 2005).

Third, it is also possible that generation differences in strategy use may be partly driven by age differences in the other party’s response to intergenerational tensions. Aging parents, for instance, may be more likely to passively avoid conflict and less likely to engage in active strategies than adult children (Birditt & Fingerman, 2005; Birditt, Fingerman, & Almeida, 2005). Parent–adult child relationships are interdependent and reciprocal (Bengtson & Allen, 1993; Umberson, Pudrovska, & Reczek, 2010), and so the other party’s reactions to conflict may in turn shape the strategies middle-aged adults adopt in their interactions with each generation. Supporting this explanation, research examining young adults’ accounts of conflict with parents and other elder family members revealed that young adults reciprocated their elders’ conflict management styles (Wiebe & Zhang, 2016).

Associations Between Conflict Strategies in Parent–Adult Child Ties and Depressive Symptoms

With respect to the implications of parent–adult child conflict strategies for middle-aged adults’ well-being, we found that passive strategies used in relationships with adult children were linked to greater depressive symptoms. These associations were found for passive strategies (both constructive and destructive), suggesting that middle-aged individuals have elevated depressive symptoms when they either avoid discussing problems with their adult children or accept the problem as it stands. These strategies are counter to the active strategies that were used most frequently with adult children in this sample. As such, we speculate that the use of conflict strategies with adult children that are less normative may contribute to feelings of psychological distress. Using passive strategies in ties with adult children may also contribute to elevated depressive symptoms because these strategies are not aimed at solving problems in the long term. Research has shown that while avoidance may be initially satisfying, it predicts poorer well-being and increased cortisol levels the next day (Birditt et al., 2015). Thus, unresolved tensions with adult children may be an enduring source of stress that intensifies depressive symptoms. Alternatively, middle-aged adults who experience more depressive symptoms may be more likely to use passive conflict strategies with their adult children. Indeed, research examining married couples suggests that people who have greater depressive symptoms tend to disengage from conflict in their close relationships (Eldridge & Christensen, 2002).

In contrast, strategies used during conflict with aging parents were unrelated to middle-aged adults’ depressive symptoms. This generational difference implies that there are unique features of relationships with adult children that render conflict strategies within these ties more salient to well-being than relationships with aging parents. Consistent with the intra-intergenerational stake hypothesis (Birditt et al., 2015), such distinction may result in part from middle-aged individuals’ stronger investment in their adult children than in their parents.

Strengths and Limitations

Strengths of this study include the within-person, cross-generational approach to the examination of conflict strategy use in parent–adult child relationships, the assessment of conflict strategies used with multiple offspring, and the use of multilevel modeling to account for the interdependence among family members. We also considered similarities and differences in patterns of conflict strategy use across generations and their links with middle-aged adults’ psychological well-being. The findings lend support to the intragenerational stake hypothesis and indicate that the use of intergenerational conflict strategies during midlife is shaped in part by gender and generational position of family member.

We also acknowledge several limitations of this work that should be addressed in future research. First, the cross-sectional models precluded the determination of causal associations. For example, individuals who are depressed may be more likely to use passive conflict strategies with their children. Next, we examined middle-aged adult reports of the conflict strategies used and not the offspring or parent reports of strategies. We need to conduct further research to examine these issues across multiple generations of families. As it stands, we cannot differentiate whether the effects are due to the age of the social partner, cohort, or generation. We also need to understand conflict with parents and children on a more micro level to understand the interpersonal processes that are occurring in these relationships and the factors that account for generation differences and links with well-being. For example, are conflicts with parents or children viewed as more or less stressful? Do individuals have tensions about different types of topics with parents than adult children? Finally, we know little about how these strategies operate over time as parents and children age together. For instance, how do these strategies relate to the history of the relationship? Are there developmental trajectories of conflict strategies in the parent child tie that mirror changes found in other ties?

In sum, while studies have examined variations in how middle-aged individuals feel about their adult children and aging parents with regards to relationship quality and investment, studies had yet to consider conflict strategies used with adult children and aging parents in the same families. This study showed that middle-aged individuals are more likely to use passive strategies with parents and active strategies with children. Middle-aged individuals who used strategies that were inconsistent with the stake and the social input model (i.e., using passive strategies with adult children) reported greater depressive symptoms. We hope that this study highlights the importance of examining within family variation in conflict strategies and that it inspires more in depth studies of the implications of conflict strategies for well-being in the parent and adult child tie.

Funding

This study was supported by grants from the National Institute on Aging under award number R01AG027769 the Family Exchanges Study II, the Psychology of Intergenerational Transfers (K. L. Fingerman, Principal Investigator). The MacArthur Network on an Aging Society (John W. Rowe, Network Director) also provided funds. This research also was supported by Grant 5 R24 HD042849 awarded to the Population Research Center at The University of Texas at Austin by the Eunice Kennedy Shriver National Institute of Child Health and Human Development. C. A. Polenick was supported by training grant T32 MH073553-11 from the National Institute of Mental Health (Stephen J. Bartels, Principal Investigator).

Conflict of Interest

The authors declared no conflict of interest.

Acknowledgment

These data have been analyzed extensively and a comprehensive list of publications and presentations are provided on the website (http://sites.utexas.edu/adultfamilyproject/family-exchanges-study/). This paper has not been published or presented elsewhere. We thank Angela Turkelson for her assistance with the analyses.