Violated Wishes About Division of Childcare Labor Predict Early Coparenting Process During Stressful and Nonstressful Family Evaluations

Abstract

Prior research has indicated that expectant parents overestimate the extent to which fathers will take part in the “work” of parenting, with mothers often becoming disenchanted when these expectations are violated following the baby’s arrival. In this study, we examine the role of violated wishes concerning childcare involvement in accounting for variability in maternal and paternal marital satisfaction, and in early coparenting behavior as assessed during family-interaction sessions. The results indicate possible negative effects of violated wishes on the enacted family process and confirm previous findings regarding the effects of marital satisfaction. In addition, we uncovered differences in the way that violated maternal wishes are related to coparenting during playful and mildly stressful family interactions.

A small, but consistent, body of research has reliably documented the relevance of first-time parents’ prenatal wishes and expectancies about family life after the birth of the baby in predicting adjustment quality following the baby’s arrival (e.g., Delmore-Ko, Pancer, Hunsberger, & Pratt, 2000; Goldberg & Perry-Jenkins, 2004; Hackel & Ruble, 1992; Kach & McGhee, 1982; Kalmuss, Davidson, & Cushman, 1992; McHale et al., 2004; Ruble, Fleming, Hackel, & Stangor, 1988; Van Egeren, 2004). Most studies indicate that parents-to-be expect a far more egalitarian division of household and childcare labor than will typically ever eventuate (Cowan & Cowan, 1992). In fact, more than one third of all women in a sample reported on by Ruble et al. (1988) expected a fully egalitarian division of labor, and so not surprisingly, many had to confront a significant violation of their prepartum expectations after the baby’s arrival. Men, too, are overly optimistic; Belsky (1985) found that the bias anticipating that fathers will be equitably involved in childcare is shown by both parents.

Most studies of violated expectancies document consistent negative effects on postpartum marital adjustment. For example, Belsky (1985) reported that especially unrealistic expectations about division of labor in childcare tasks or about the effect that the baby would have on the couple’s own relationship and on their relationships with family and friends predicted greater marital dissatisfaction during the early postpartum months. Ruble et al. (1988) found that women assuming significantly more of the house- and childcare burden than anticipated reported the most negative feelings about their husbands’ involvement in these tasks, and were more negative when evaluating the effect of the baby on their marital relationship. Hackel and Ruble (1992) showed that women highly wed to prebirth expectations about division of labor reported more negative change in marital satisfaction after transitioning to new parenthood than did women with less prenatal commitment to their expectations.

Beyond predicting general declines in marital satisfaction, violated expectations regarding division of childcare-related tasks may have more insidious effects in the emergent family process. Mothers whose expectations have been violated may not only be less content in the marriage but also less inclined to cooperate actively with their husbands in co-creating a supportive coparental alliance. The construct of coparenting, central in the writings of structural family theorists such as S. Minuchin (1974), is a critically important one that has taken shape over the past 10 years in family research studies (Belsky, Crnic & Gable, 1995; McHale, 1995; for more extended theoretical discussion, see McHale & Sullivan, 2007). Coparenting has been defined as the quality of coordination between adults in their roles as parents (McHale, 1997; McHale, Lauretti, Talbot, & Pouquette, 2002). While related to marital adjustment, marital and coparenting adaptations also differ from one another in several important respects (McHale, 1995). In general, collaborative coparenting partners are those who develop genuine trust and confidence in one another’s parenting judgments and interventions, and who support and validate one another’s parenting efforts (Abidin & Brunner, 1995; McHale, 1997).

To date, no published studies have examined whether violations of parents’ wishes about their eventual division of childcare labor are tied to features of the emergent coparenting process during the early postpartum months. In one relevant study, Van Egeren (2004) did link such expectancy violations to both mothers’ and fathers’ experience of coparenting. Specifically, in Van Egeren’s sample, mothers and fathers whose expectancies regarding division of labor were violated most severely felt the least positive about their coparenting relationship at 3 months postpartum whereas those whose prenatal expectancies came closer to the actual division of labor that had come to materialize at 3 months postpartum felt more positively about their emerging coparental partnership.

In Van Egeren’s (2004) study, the link uncovered was between violated expectations regarding division of labor and parents’ reports of their coparenting experiences—an important and informative finding in that subjective experience is often tied to enacted behavior. However, it is equally important to examine directly the relationship between violated wishes about division of labor and observed coparenting process, as over a dozen studies have shown that observed coparenting can be linked, both contemporaneously and longitudinally, to important social and emotional outcomes in infants and young children (e.g., Belsky, Putnam, & Crnic, 1996; Caldera & Lindsey, 2006; Fivaz-Depeursinge & Favez, 2006; Katz & Low, 2004; Keren, Feldman, & Namdari-Weinbaum, 2005; Leary & Katz, 2004; Lindsey, Caldera, & Colwell, 2005; McConnell & Kerig, 2002; McHale, Johnson, & Sinclair, 1999; McHale & Rasmussen, 1998; McHale & Rotman, 2007; Schoppe, Mangelsdorf, & Frosch, 2001; Stright & Bales, 2003).

What exactly is the relationship between division of childcare labor and coparenting collaboration? Although the two are often treated as synonymous, there is little empirical basis for this equation. In fact, empirical evidence has sometimes contradicted this view. For example, Carleton, Rotman, Fish, and McHale (2001) reported that at 3 months postpartum, parents’ reports of equity in their division of labor are not strongly connected with observed coparental collaboration during playful family interactions. This finding makes some sense, given that prior studies have established that in certain families, mothers respond negatively when fathers take on greater parenting involvement with young infants than mothers had envisioned (e.g., Goldberg & Perry-Jenkins, 2004; Hackel & Ruble, 1992). Hence, we saw the value in directly examining the relationship between the extent to which parents’ prenatal wishes about childcare involvement were violated, and the nature and quality of observed coparenting dynamics that had begun to consolidate in their family.

Although there have been some interesting perspectives on men’s expectancies about and involvement in childcare labor, no consensus has emerged as to whether expectancy violations among men hold the same psychological salience as do expectancy violations for women (e.g., Blair & Hardesty, 1994; Cowan & Cowan, 1988; Croghan, 1991; Grote & Clark, 2001; Helms-Erikson, 2001; Wicki, 1999). Perhaps this is because mothers are the ones who generally anticipate assuming less childcare-related work than they ultimately end up taking on while fathers generally end up doing less than they had predicted prenatally (e.g., Belsky, 1985; Ruble et al., 1988). Prenatal expectancies about future labor inequity also appear to be more closely tied to maternal pessimism about future family process than to paternal pessimism (McHale et al., 2004). For these reasons, we expected stronger connections between wish violations and family process for mothers than for fathers.

Another issue is whether violated wishes about childcare labor should be expected to correlate with enacted family process in the same fashion in any and all circumstances. For example, if parents already are displeased about their childcare roles and responsibilities, coparenting conflict may be more likely to brew when the couple has to deal with an ill-tempered or inconsolable infant. By contrast, during routine playful family interactions when the couple is not confronting any acute stressor, there is less reason to expect that more substantial wish violations would be responsible for any coparental miscoordination that might be in evidence. To our knowledge, no prior study has asked whether similar patterns of linkage are evinced across different parenting contexts. Hence, in this report, we also examine whether the associations between wish violations and coparenting dynamics are equally apparent during family play and during a mildly stressful family interaction.

In summary, we set out to establish whether coparenting and family process at 3 months postpartum is more dissonant or less collaborative when parents have experienced greater violations of their prenatal wishes concerning sharing of childcare labor. We speculated that a pattern of associations between wish violations and coparenting process would be more likely to show up for women, and that the pattern would be most readily apparent when coparenting was assessed during mildly stressful family interactions rather than when it was assessed during playful family interactions.

METHOD

Participants

Expectant couples (N = 119) were recruited from childbirth classes offered at two hospitals in an urban region of the Northeastern United States. All were participants in a federally funded, longitudinal investigation of the transition to new coparenthood, eligible for the study if married or cohabiting and expecting their first child together at the time of recruitment. Eighty-eight percent of these individuals were Caucasian, and 12% identified themselves as African American, Asian American, or Hispanic. Average age of mothers was 31.7 (SD = 5.1, range = 22–47) years and fathers 33.3 (SD = 6.0, range = 21–49) years. This was an advantaged, middle-class sample; median family income in 2002 U.S. dollars fell in the 70,000 to 75,000 range (range = $30,000–35,000 to > 100,000). Further, nearly all couples reported at least one partner with some postsecondary education, and over 70% included a parent who had earned at least a college degree.

Procedure

During a third-trimester assessment, both mothers and fathers completed a prenatal version of Cowan and Cowan’s (1988) “Who Does What” (WDW) survey (described later). At 3 months postpartum, couples completed the standard WDW survey and a measure of current marital satisfaction during a home visit. During the visit, they also took part in two videotaped, semistructured interactions designed to highlight coparenting process: a triadic Family Play interaction, patterned after Fivaz-Depeursinge and Corboz-Warnery’s (1999) Lausanne Trilogue Play (LTP), and a three-person Still-Face Posing and Reunion task, following Als, Tronick, and Brazelton’s (1979) usually dyadic protocol.

For the adapted LTP interaction, family members’ seats were arranged to form an equilateral triangle. The infant was seated in an orthopedically crafted infant seat that accommodated the baby’s contour and girth and supported the baby’s head and neck. The parents took seats equidistant from one another and from the baby at the child’s eye level. The baby seat could be easily maneuvered on a swivel towards either parent or to a fixed location between the two of them. The standard LTP protocol was followed, with an important exception: Each of the four segments of the assessment were limited to approximately 2 min rather than being left open to parents, as is standard. For readers unfamiliar with the LTP, it is divided into four parts capturing all possible configurations of interaction within a 3-member family. First, one parent played with the baby while the other parent was just present. Second, the parents switched, so that the parent initially in the “Third Party” position now played actively with the baby. Third, both parents played with the baby together. Finally, the parents took a turn actively engaging with one another as the baby assumed a “Third Party” position. This procedure was videotaped and later rated along several process dimensions.

In the adapted Still-Face procedure, the seating arrangements were the same as that during the LTP. The Still-Face Posing and Reunion was divided into three segments, again each lasting for 2 min. First, both parents played together with the baby. Second, the parents were asked to terminate the interaction and pose motionless, expressionless faces. Finally, both parents reengaged with the baby and worked together to “repair” the family connection with the child. On occasions when babies became unduly upset during the still face part of the task, either or both parents were permitted to call an end to that part prematurely to reengage with the baby. Six mothers and 2 fathers asked to end the still face segment early. Another 31 mothers and 24 fathers showed marked difficulties posing the motionless face for durations of longer than 1 min, although all parents made genuine efforts to carry out the procedure (which also was videotaped).

Measures

Division of Childcare Labor: Materialized Realities and Wish Violations

The WDW surveys were used to assess both the actual “Who does what childcare?” at 3 months and the degree of discrepancy between prenatally wished-for and actual WDW proportions at 3 months. The standard WDW asks respondents to describe who is responsible in their family for a variety of childcare-related activities (e.g., feeding the baby, doing laundry, and responding to the baby’s cries). Two sets of ratings are obtained. For the first set, parents answer 12 questions about how they divide up specific tasks (e.g., deciding about the baby’s feedings, changing diapers, choosing toys for the baby). For the second set, they answer 12 questions about how they delegate “on-duty” times (e.g., from 9 a.m. to 1 p.m. on weekdays, from 1 p.m. to 5 p.m. on weekdays; dinner/playtime/bedtime on weekdays, 9 a.m. to 1 p.m. on weekends, etc.). After answering these 24 “how it is” questions, parents answer a parallel set of 24 questions indicating how they would like to ideally divvy up childcare duty (“how I would like it to be”). The scale for each WDW item ranges from 1 to 9. Scores of 1 signify that the woman does all the work, scores of 5 indicate that both partners equally shoulder responsibility, and scores of 9 indicate that the man does all the work for that particular task or time period.

For the prebirth WDW assessment, since there was no baby as of yet, parents were asked to do two things: Describe how they expected the childcare work might be allocated after the baby’s arrival and then how they wished these tasks would be divvied up, in the ideal. At 3 months, they provided the standard WDW “how it is” and “how I would like it to be” ratings. These two sets of surveys (prenatal and 3 months) were used to create the following indices.

Actual Division of Labor: Fathers’ Shouldering of Childcare Responsibility at 3 Months

As a measure of the degree of weight that fathers were actually pulling at 3 months postpartum, we did a straightforward averaging of the parents’ ratings on the 12 specific tasks and the 12 “on-duty” items. In this sample, the average score for mothers’ reports was 3.47 and for fathers’ reports 3.61 (see Table 1), and fathers and mothers concurred to a remarkable degree about how much the father was doing (r = .75, p < .001). In five families, father was a stay-at-home parent at 3 months while mother was working; these were the only families in the study where the division of labor score exceeded 5 (signifying that father shouldered more child care work than mother). But for all families, higher scores on this index connoted relatively greater father involvement, if not equity, in the sharing of duties. Lower scores signified the opposite: greater inequity and overreliance on the mother.

Table 1.

Descriptive Data for “Who Does What” Measures

	Mothers			Fathers
	M	SD	Range	M	SD	Range
Actual Division of Labor (Extent of 3-month father care)	3.47	0.95	1.5/6.0	3.61	0.88	1.5/6.0
Wish Violations for Specific Tasksa
Deciding meals	1.03	2.03	−7/+6	1.53	1.70	−4/+5
Mealtimes	1.75	1.62	−2/+4	1.73	1.87	−4/+6
Diapering	1.14	1.47	−3/+4	0.67	1.64	−6/+5
Bathtimes	1.45	2.10	−5/+6	1.17	2.13	−4/+6
Handling crying (days)	0.64	1.46	−3/+4	0.61	1.41	−4/+4
Handling crying (mid-of-night)	1.63	2.27	−6/+6	1.49	2.08	−5/+7
Taking baby on excursions	0.99	1.39	−3/+4	0.76	1.53	−4/+5
Choosing toys	1.07	1.57	−2/+4	1.04	1.59	−4/+5
Playtimes	0.91	1.37	−2/+5	1.05	1.14	−1/+3
Laundering baby’s clothes	1.50	2.37	−4/+8	1.28	2.11	−4/+8
Arranging sitters/childcare	0.97	1.90	−4/+4	0.32	1.83	−4/+5
Dealing with doctors	1.28	1.84	−3/+5	1.15	1.36	−2/+4
Sum of Wish Violations for “On Duty” Items	5.71	7.93	−9/+38	4.84	7.19	−3/+25
Fathers’ “on-duty” Responsibility

^aFor both the specific task Wish Violation scores and for the overall “On Duty” Wish Violation scores, values above zero signify that the mother was doing more (and the father doing less) than the respondent had wished whereas values below zero signify that the father was doing more (and the mother less) than the respondent had wished.

Degree of Pre- to Postpartum Wish Violations

As a means of evaluating the violation of parents’ prenatal wishes, we calculated discrepancy scores for each of the 12 specific WDW tasks (e.g., diapering, doing laundry) and for the 12 “on-duty” items (weekdays 9–1; weekend evenings–midnight, etc.). We did so by subtracting each 3-month “how it is now” score from its parallel prebirth “how I would like it to be” score. We chose to employ a wished-for/actual discrepancy index rather than using an “expected/actual” discrepancy score principally for conceptual reasons, anticipating that there would be greater discontent associated with the gap between parents’ wished-for ideal and the actual materialized circumstance.

Given the nature of the WDW’s 1 to 9 scale point definitions and how we calculated individual item-discrepancy scores, discrepancy scores higher than zero signified that the mother was shouldering more work than the respondent had reportedly hoped she would. For example, the item-discrepancy score for a parent who had said during the pregnancy “I would like it if we responded to the baby’s middle-of-the-night needs 50/50” (a score of 5), but who reported at 3 months that mother was shouldering 100% of the middle-of-the-night needs (a score of 1 on the 9-point scale, signifying “she does it all”), would be +4 (the prenatal rating of “5” minus the postnatal rating of “1,” yielding a score of +4). By contrast, scores below zero signified that fathers were doing more than the respondent had ideally hoped. So the item-discrepancy score of a parent who said during the pregnancy that he or she hoped mother would handle all of the middle-of-the-night needs (a score of 1 on the 9-point scale), but reported at 3 months that father had actually done half the work (i.e., handling middle-of-the-night needs as much as mother had, or a score of 5), would be −4 (1 – 5 = −4). Means and ranges for the 12 specific WDW task discrepancy scores are reported in Table 1.

Marital Satisfaction

Parents’ marital satisfaction was assessed using the Locke-Wallace Marital Adjustment Test (MAT; Locke & Wallace, 1959), a widely used measure with demonstrated reliability and validity (Gottman, Markman, & Notarius, 1977). On the MAT, scores below 100 are generally taken as an index of clinical distress. Overall, parents in this sample were adjusting relatively well in their marriages. At 3 months postpartum, the average MAT score of mothers was 124.5 (range = 29–157), and among fathers, 119.6 (range = 65–153); however, 10% of mothers and 20% of fathers reported clinically significant marital dissatisfaction 3 months after the baby’s arrival.

Coparenting Behavior at 3 Months

Coparenting dynamics were assessed both during the LTP and the Still-Face procedure. An adapted version of the Coparenting and Family Rating System (CFRS; McHale, Kuersten-Hogan, & Lauretti, 2000) was used to code these interactions, incorporating a number of minor adjustments in subscale exemplars to accommodate behaviors evinced during the different interaction tasks (McHale, Berkman, Kavanaugh, Carleton, & Alberts, in press). Overall scores for the LTP interactions were generated after viewings of all four parts while parallel scores for the Still-Face task were based on coparenting behavior observed during the Reunion episode only.

Quality of the coparenting relationship was evaluated for each of the four dimensions summarized below. Rating scales of 1 to 5 were used, with 1 indicating no evidence of the target behavior and 5 indicating frequent recurrence of that behavior. Ratings for the LTP and Still-Face interactions were provided by separate teams of blind, trained coders (graduate and postdoctoral students in clinical and developmental psychology). Descriptive data are provided in Table 2; interrater reliabilities for all dimensions were acceptable, with intraclass correlations ranging from .66 to .89.

Table 2.

Descriptive Data for Coparenting Assessed During Lausanne Trilogue Play (LTP) and Still-Face Procedures

	LTP			Still-Face Interaction
	M	SD	Range	M	SD	Range
Cooperation	3.56	0.82	2–5	2.66	1.11	1–5
Warmth	3.47	0.97	1–5	2.40	1.06	1–5
Competition	2.08	0.89	1–5	1.57	0.94	1–5
Verbal Sparring	1.95	0.98	1–5	1.20	0.77	1–5

LT P = Lausanne Trilogue Play.

Active Competition Between Parents

In the context of the LTP and Still-Face interactions, competition took the form of nonverbal behavior by one or both parents that deflected the infant’s attention away from the other parent. Such behaviors included orienting the child towards oneself and away from the partner in Parts 3 and 4 of the LTP or during the Still-Face reunion, interrupting a game the child appeared to be enjoying with the other parent, and stimulating or soothing the baby in a manner at variance with the partner (appearing to intentionally draw the baby’s attention from the partner and to the self, and/or to disqualify or “outparent” the other parent).

Verbal Sparring

This was an index of the frequency of sarcastic and disqualifying comments parents directed toward one another during the flow of interaction with the baby. Lower end scores (of 2 and 3) were assigned when such comments were infrequent and of uncertain valence (i.e., said playfully, but conveying some underlying disapproval) while scores of 4 and 5 connoted clearer disapproval and sanctioning.

Cooperation Between Parents

Cooperation ratings captured the parents’ propensity to work together as a team, actively respecting and supporting each other’s initiatives and activities with the child. During the LTP, parents who showed primarily benign support (e.g., simply staying out of the partner’s way without seeking to address the baby him- or herself) received midrange scores; the highest scores were given to parents who mirrored and built upon each other’s efforts rather than simply tolerating them. By point of contrast, high Cooperation scores could be attained in the Still-Face if one parent perceived that the other was effectively soothing the baby and gave that parent due leeway, so long as he or she remained empathically in touch with the child.

Warmth

Warmth scores reflected the parents’ affect both towards one another and towards the baby. Interadult warmth was typically expressed facially and sometimes verbally while warmth towards the baby also was conveyed via touch. The highest scores went to families in which convincing levels of warmth were present between the adults as well as with the baby.

Conflict and Collaboration Scores

The aforementioned four sets of scores were then combined to form composite variables on conceptual grounds. Parallel coparental Conflict scores were constituted for each of the two assessments by adding scores for competition and verbal sparring (Cohen’s αs: LTP = .60, Still Face = .52). Parallel coparental Collaboration scores were constituted by adding scores for cooperation and warmth (Cohen’s αs: LTP = .87, Still Face = .86).

RESULTS

Data Reduction

As the WDW was designed for a range of respondents, from families with young babies to families of older toddlers, we expected certain WDW items to be especially significant for new parents while others would be of little consequence. For example, we imagined that at 3 months, unfulfilled wishes about diaper changes and dealing with crying or middle-of-the-night needs might challenge the emergent family process more markedly than unfulfilled wishes about playtime or toy choices. We also suspected that violated wishes about the amount of “on-duty” time taken on by fathers would be far more unsettling than unfulfilled wishes about most specific chores.

To address this issue, the 24 wished-for/actual discrepancy scores were subjected to principal components factor analyses with varimax rotation, conducted separately for mothers and for fathers. While we found no clearly defined, conceptually meaningful “expectancy violation clusters” for the first 12 items concerned with specific tasks, the analyses indicated that 6 of the second 12 “on-duty time” items did comprise a strong, interpretable factor that replicated for both parents. This factor, summarized in Table 3, accounted for 24% of the variance for mothers and 21% for fathers. It was constituted of wish violations concerning (a) the extent of father involvement on weekday evenings (at dinner/playtime/bedtime and from evenings–midnight), and b) the extent of father involvement on the weekend (from 9 a.m. to 1 p.m., 1 p.m. to 5 p.m., at dinner/playtime/bedtime, and from evenings–midnight). Violated wishes concerning the amount of on-duty time by fathers on weekdays and during middle-of-the-night periods did not contribute to this intuitively meaningful discrepancy score factor.

Table 3.

“On Duty” Factor Loadings for Mothers and Fathers

“Who Does What” Item	Factor Loading Mothers	Factor Loading Fathers
Weekday dinner/playtime/bedtime	.59	.57
Weekday evenings to midnight	.40	.54
Weekends 9 a.m.–1 p.m.	.65	.80
Weekends 1 p.m.–5 p.m.	.86	.86
Weekend dinner/playtime/bedtime	.82	.81
Weekend evenings to midnight	.58	.64
%Variance	24%	21%

Note. Items were used to create a composite score reflecting extent to which the parent’s prenatal wish about the father’s overall “on-duty” responsibility was violated.

Given this factor’s salience for both men and women, we created an index reflecting the extent to which the parent’s prenatal wish about the family’s overall “on-duty” responsibilities had been violated. We formed this score for both men and women by summing the six discrepancy items loading highly on this factor (Descriptive data are shown in Table 1.) By virtue of the manner in which individual discrepancy scores were calculated (such that scores above zero signified that mothers were doing more and fathers less than had been hoped for during the pregnancy), higher scores on this index reflect more substantially violated wishes in the direction of lower father involvement. This is the primary wish-violation index we use in regression analyses later that examine the salience of wish violations for 3-month marital and family outcomes. For heuristic purposes, and to provide a fully interpretable reading of the data on violated wishes about specific tasks, correlations between the 12 specific task-discrepancy scores and marital and family outcomes are presented in Table 4.

Table 4.

Significant Correlations Between Magnitude of Wish Violations About Specific Childcare Tasks and Family Adaptation at 3 Months Postpartum

	Marital Satisfaction		LTP		Still Face
	Wife	Husband	Collaboration	Conflict	Collaboration	Conflict
Violated Wish of Mom About:
Deciding meals/feedings
Mealtimes/feedings
Diapering	−.21		−.20
Bathtimes
Handling crying (days)	−.20
Middle-of-night needs	−.34
Taking baby on excursions	−.34	−.20
Choosing toys
Playtimes
Laundering baby’s clothes	−.24
Arranging sitters/childcare		−.20
Dealing with doctors	−.20
Violated Wish of Dad About:
Deciding meals/feedings
Mealtimes/feedings
Diapering
Bathtimes
Handling crying (days)					.21
Middle-of-night needs	−.32
Taking baby on excursions
Choosing toys
Playtimes
Laundering baby’s clothes	−.20
Arranging sitters/childcare
Dealing with doctors

Note. All correlations shown are significant at p <.05.

Analyses: Expectancy Violations and 3-Month Adjustment

Were parents whose pre-baby childcare wishes failed to materialize more likely to be dissatisfied with the marriage or to have co-created a less collaborative, more antagonistic coparenting alliance by 3 months postpartum? Were patterns of associations between expectancy violations and marital or coparenting adjustment especially pronounced for women and less prominent or consistent for men?

Pearson Correlations

Turning first to the 12 specific task-discrepancy scores, we found that half of these items, including violated expectancies about diapering, responding to infants’ crying, handling laundry, and making babysitter and doctor’s arrangements, significantly predicted mothers’ 3-month marital dissatisfaction. These associations can be found in Table 4 (which details significant associations only). As can be seen in Table 4, no expectancy violation had yet soured the couple’s observed coparenting process, save for one. In families where maternal wishes about the joint sharing of diaper changes were more substantially violated, LTP play interactions were characterized by lower coparenting collaboration (Table 4). In other words, when mothers were changing markedly more diapers than they hoped to, we did find subtle evidence that the couple had already, by 3 months postpartum, begun co-creating a less collaborative coparenting process. This said, the diapering-violation index did not yet appear to be associated with observable coparenting conflict during the LTP or with family process during the stressful Still-Face interaction.

For fathers, as anticipated, wish violations about specific tasks were largely inconsequential. In two instances, men’s wish violations mirrored women’s in predicting maternal marital dissatisfaction—men responding less to crying and doing less laundry than they had indicated they wished to had unhappier wives, although the men’ s marital satisfaction was not so affected. Besides these two associations, there was only one linkage between men’s wish violations and observed coparenting process. Coders noted greater collaboration during the Still-Face repair phase in families where mothers took greater responsibility for addressing the baby’s crying (and fathers less) than the men had hoped might be the case during the pregnancy (Table 4). Though this is a mildly perplexing finding, recall that couples could receive a high coparenting cooperation score in the Still-Face if one acceded to the other in soothing the child, so long as both remained affectively in touch. Perhaps fathers who had grown more accustomed to taking a back seat to mothers when babies cried found entrusting Still-Face soothing responsibility to mothers more comfortable and familiar.

Fathers’ On-duty Involvement: Regression Analyses

To determine whether violated wishes about the overall extent of father’s on-duty involvement affected 3-month marital satisfaction or coparenting process, we conducted three sets of multiple regression analyses in which we sought to predict (a) marital satisfaction, (b) coparental process during the Still Face, and (c) coparental process during the LTP. In these analyses, the parents’ reports of their “actual” division of labor (i.e., the extent to which fathers were currently shouldering childcare responsibility) were entered first, followed by the index signifying the degree to which parents’ wishes about fathers’ on-duty involvement had been violated. Framing the analyses in this way allowed us to assess whether it was the violated wish, per se, or simply the actual amount of the load fathers were carrying at 3 months that made a difference. Table 5 presents a summary of these analyses.

Table 5.

Predicting 3-Month Outcomes From Division of Labor and Wish Violations

Mother’s Marital Satisfaction		δR2	F	β
Block 1		.04	4.05**
Actual Labor (Mom Report)				.19*
Actual Labor (Dad Report)				.11
Block 2		.07	8.98***
Wish Violation (Dad)				−.28**
Wish Violation (Mom)				−.22**
Father’s Marital Satisfaction	R2	δR2	F	β
Block 1	.01	.01	0.63
Actual Labor (Dad Report)				.07
Actual Labor (Mom Report)				.05
Block 2	.04	.03	1.75
Wish Violation (Dad)				−.13
Wish Violation (Mom)				−.11
Still-Face Conflict
Block 1		.03	2.03
Actual Labor (Dad Report)				.22
Actual Labor (Mom Report)				−.27**
Block 2		.06	3.57**
Wish Violation (Dad)				−.14
Wish Violation (Mom)				.32***
Still-Face Collaboration		δR2	F	β
Block 1		.04	2.17
Actual Labor (Dad Report)				−.26**
Actual Labor (Mom Report)				.25*
Block 2		.02	1.59
Wish Violation (Dad)				−.19*
Wish Violation (Mom)				.02
LTP Conflict		δR2	F	β
Block 1		.02	0.88
Actual Labor (Dad Report)				−.01
Actual Labor (Mom Report)				.13
Block 2		.04	2.63*
Wish Violation (Dad)				−.22*
Wish Violation (Mom)				.23*
LTP Collaboration		δR2	F	β
Block 1		.00	0.21
Actual Labor (Dad Report)				−.08
Actual Labor (Mom Report)				.08
Block 2		.01	0.09
Wish Violation (Dad)				−.04
Wish Violation (Mom)				−.02

Note. LT P = Lausanne Trilogue Play.

^*p < .10.

^**p = .05.

^***p < .01.

As seen in Table 5, more sizeable violations of parents’ wishes about fathers’ on-duty involvement explained unique variance both in mothers’ marital satisfaction and in the extent to which the couple exhibited coparenting conflict during the Still-Face interaction—and this was after taking stock of father involvement in the couple’s actual division of labor (Marital Satisfaction: δR² = .07, F = 8.98, p < .01; Still-Face Conflict: δR² = .06, F = 3.57, p < .05). These data support, in part, our speculation that the effects of wish violations were likely to be stronger during the mildly stressful Still-Face assessment than during the LTP assessment. This said, however, mothers who took on more responsibility than they had hoped tended to have more conflict in the LTP task (δR² = .04, F = 2.63, p < .10). The degree of parents’ wish violations did not explain a significant increment in overall variance either in fathers’ marital satisfaction or LTP or Still-Face collaboration (Table 5).

Although our results were in line with the anticipated gender pattern (It was mothers’ wish violations rather than fathers’ that significantly predicted Still-Face conflict; β = −.32, p < .01.), we found two instances where fathers’ expectancy violations tended to be associated with family coparenting. When fathers reported shouldering more on-duty responsibility at 3 months than they hoped to during the pregnancy, LTP conflict tended to be higher (β = −.22, p < .10), and Still-Face collaboration tended to be lower (β = −.19, p < .10). Note that this latter association is consistent with one previously reported correlational finding: that Still-Face Collaboration during the reunion was actually greater when mothers had taken on greater responsibility for handling the baby’s crying than fathers had said they hoped mothers would during the pregnancy.

Finally, we underscore that the extent to which respondents described fathers as actually involved in childcare labor failed to explain significant variance in any measure of observed coparenting process, despite these scores having been entered in first in the regression analyses—although, as others have found, these scores did correlate with mothers’ marital satisfaction (R² = .04, F = 4.05, p < 05; see Table 5).

DISCUSSION

We have known for a number of years that when women’s expectations about caregiving support from their partner are significantly violated during the early postpartum months, they are unhappy with this state of affairs. As summarized earlier, some studies have documented higher rates of depression, and most have chronicled lower levels of marital satisfaction for these disgruntled mothers. This study is one of the first to extend beyond these important findings to document that the more mothers’ wishes are unfulfilled, the greater the possibility that there also will be negative ramifications for the enacted family process. In families where women witnessed a greater discrepancy between their wished-for family ideal and the one that actually materialized, they were not only more likely to report lower levels of marital satisfaction but also to be partners in coparenting alliances showing some subtle, beginning signs of poorer coordination and greater conflict.

Our data further indicated that the assessment context used to evoke coparenting process may prove to be an important consideration in studies of this nature. Few reliable and valid paradigms for studying early coparenting process currently exist. However, there has been growing awareness of the promise and clinical utility of the LTP paradigm for evaluating early family process, especially for families showing some form of alliance disturbance. In the current study, despite the fact that none of the families we saw had been preselected on the basis of clinical parent or family pathology, we did document interconnections between the extent to which partners worked cooperatively and supportively during the LTP interactions and the magnitude of mothers’ wish violations in one realm. Specifically, in families where mothers’ wishes concerning father participation in the “grunt” work of diaper changing had been more severely violated, coparenting interactions were less warm and less cooperative—although they were not marked by greater hostility or competitiveness.

By contrast, in families where women’s wishes concerning fathers’ on-duty involvement with the baby had been violated more dramatically, the emotionally stimulating Still-Face assessment did reveal evidence of greater antagonism between parents. In those families where fathers were on-duty less of the time than mothers had hoped they would be during the pregnancy, parents were more likely to work at cross-purposes, behave in excluding or disqualifying ways, or exchange verbal barbs as they worked to comfort their distressed baby during the reunion. If this proved to be a replicable finding, it would suggest that enacted antagonism between disenchanted parents can be meaningfully assessed in situations where the parents are coparenting under duress, and this antagonism might potentially be foreseeable from prenatal assessments. In more routine family-play interactions, the only significant sign of strain was the lower likelihood of active coparenting collaboration if mothers’ wishes about fathers’ participation in changing diapers had not become a reality. At minimum, these findings suggest that researchers concerned with early coparenting dynamics may find added value in observing coparenting during both playful and stressful family interactions. Had we chosen to evaluate interadult commerce only during playful family interactions, as have most previous studies, we would have missed the important cross-time linkage with early coparenting conflict.

We also underscore another meaningful and provocative finding: Overall, parents’ perceptions of their family’s actual division of labor were unrelated to coparenting process during the LTP and Still Face. In other words, families who looked the best during coparenting observations were not necessarily the ones reporting greater equity in their division of childcare labor. This confirms that the quality of early coparenting alliances is based on considerations beyond the simple division of childcare labor. Before closing the door prematurely on this important theme, we do note that at the level of individual parent reports, when mothers portrayed fathers as shouldering a more equitable share of childcare duties there was a lower likelihood that there would be active signs of conflict between parents in their efforts to soothe.

We emphasize that our labor-division findings may reflect age-sensitive effects. Van Egeren (2004) reported that by the time children in her sample had reached 12 months of age, the family’s division of labor had inched closer to the parents’ prenatal expectations, with fathers shouldering more of the burden and mothers less than they had at 3 months. More importantly, the shrinking discrepancy between expected and actual labor division positively influenced partners’ experience of coparenting. If wishes and expectancies do indeed continue shaping coparenting experience beyond the early postpartum months, there would be value in future studies examining whether coparenting dynamics follow lock-step with shifts in coparenting experience, or whether early emerging behavioral patterns take on momentum of their own and continue progressing on the path the family began charting early on. In such work, researchers should attend both to the role of parents’ personally held experiences and wish violations concerning caregiving and to the role of enacted coparental dynamics during routine and stressful interactions as contributors to infants’ attachment security with the parents.

Finally, returning to the topic of gendered findings, we found, as anticipated, that it was mothers’ and not fathers’ violated wishes that significantly predicted observed coparenting conflict during the Still-Face procedure. However, both fathers’ and mothers’ violated wishes predicted mothers’ marital satisfaction. Mothers were significantly more likely to be maritally satisfied when their own and their partners’ prenatal wishes were not far off from reality. Fathers violated wishes came into play, albeit at a trend level, in two cases—and both are worth mentioning. When fathers said they were carrying more on-duty responsibility than they had wished for during the pregnancy, collaboration during the Still Face tended to be lower, and conflict during the LTP tended to be higher. We highlight these trends primarily to caution researchers not to “throw the baby out with the bathwater” and dismiss the wish violations of men as inconsequential. If fathers feel more put upon by the amount of time they are devoting to babies (whether they actually are carrying a more equitable share of the load), it may be possible to note subtle distress signs in the early family process.

We acknowledge that this study has several limitations. First, there are certainly a great many other factors in families besides those examined here that could potentially play a role in the relationships we were concerned with. For example, parents’ internal working models of attachment and the couple’s marital dynamics likely combine in setting the stage both for parents’ wishes about family dynamics and for the nature of their responses to wish violations (Cowan & Cowan, 2005; Talbot & McHale, 2003). While we know a fair amount about how individual and marital forces directly and indirectly affect parenting and attachment (Belsky, 2006; Belsky & Jaffee, 2006), we know much less about how these same forces organize coparenting in families. This said, given the complexities of the wish-violation data and our aim to chart the terrain in a manner that was both practically useful and interpretable, our goal in this report was a modest one—outlining “big-picture” links between different expectation violations and different manifestations of the coparenting process. However, we recognize that charting these associations in a relatively homogenous community sample enrolling relatively few clinically distressed families (at least according to the marital adjustment data) leaves us unclear whether similar or even more dramatic findings would have surfaced had we undertaken this same study with a more clinically diverse population.

Second, we made a conscious choice to study the discrepancy between parents’ wished-for family ideals and the ones that materialized rather than between expected and materialized realities. Most studies of expectancy violations do not ask parents about their prenatal ideal wishes, sticking just with “how I expect things will be.” We realize that in asking parents not just about what they expected to happen but what they ideally wanted to happen, we could have inadvertently intimated that what they hoped for may not actually come to pass. But our rationale was that explicitly inviting parents to distinguish their closely held wishes from their arguably more pragmatic expectancies tapped a more psychologically salient wellspring, encouraging a broader range of responses and ultimately yielding “violation” scores that would be uniquely meaningful and less likely to be voiced through other means. Nonetheless, interpretation of the findings of this report must keep in mind that it is unclear whether our results would have been different had we too limited parents only to telling us about their “expectancies” and not their “wished-for” outcomes.

With these important caveats and qualifications in mind, we believe that the sensible pattern of associations uncovered in this study calls for further investigation of how violated wishes may come to influence early coparenting and triadic family process. Given the stability of family dynamics from early infancy forward (Fivaz-Depeursinge & Corboz-Warnery, 1999; McHale & Rotman, 2007), and the capacity of coparenting and family dynamics during infancy to significantly predict later adjustment as early as the toddler years (Belsky et al., 1996; McHale & Rasmussen, 1998; Schoppe et al., 2001), there is clinical value in understanding the radiating effect that wish violations may have in the nascent, emergent family system.