Abstract
This study examined the degree to which 205 full-time employed mothers of infants endorsed intensive mothering beliefs (IMB), the stability of IMB, and contextual correlates of IMB. Results suggested that full-time employed mothers in this study did not endorse IMB, on average, but that endorsement varied for specific domains of IMB for the total sample and by mothers’ education. Global IMB scores did not change from 4 to 16 months postpartum, with the exception of single mothers whose scores declined over time. Endorsement of IMB was associated with multiple socioeconomic and demographic characteristics, and endorsement of IMB also varied based on unique intersections of race, education, and marital status.
Keywords
Introduction
Although current employment statistics suggest that many mothers of young children are constructing their mothering identities within the context of full-time employment (U.S. Bureau of Labor Statistics, 2010), a recent large-scale survey suggests that most full-time employed mothers would prefer working fewer hours (Pew Research Center, 2007). Although employment is generally protective for women’s well-being (Buehler & O’Brien, 2011; Turner, 2007), the positive effects of full-time employment are not uniform across mothers and are shaped by the meanings mothers ascribe to their work and family roles (Helms-Erikson, Tanner, Crouter, & McHale, 2000). Researchers have drawn attention to the importance of considering mothers’ beliefs and attitudes about their work and family roles (Arendell, 2000; Elgar & Chester, 2007; Perry-Jenkins, Repetti, & Crouter, 2000). However, limited studies have explicitly examined the mothering beliefs of full-time employed mothers of infants, a growing segment of the population facing a unique set of life demands. The present study examined intensive mothering beliefs (IMB), a specific set of mothering beliefs that has been theorized to be prevalent and influential in the lives of contemporary mothers (Hays, 1996). Informed by a theoretically grounded yet understudied literature on this specific set of mothering beliefs, we examined the extent to which 205 full-time employed mothers of infants endorsed IMB, the stability of IMB across 1 year, and contextual correlates of IMB.
Using qualitative data from a purposive sample of 20 employed and 18 nonemployed mothers, Hays (1996) was the first to describe intensive mothering ideology as a set of beliefs about appropriate caregiving for children and corresponding maternal behaviors. According to this set of beliefs, children are innately good or innocent, mothers should be self-sacrificing and primary caregivers, and correct childrearing requires all of mothers’ time, energy, and resources (Hays, 1996). Hays and others (Thurer, 1994, Warner, 2005) have asserted that the selfless, child-absorbed mother (i.e., one that performs intensive mothering) has been and continues to be a golden standard by which many women judge their success as mothers. These same scholars have suggested that IMB are particularly problematic for full-time employed mothers of non-school-aged children because of their potential to elicit feelings of anxiety, depression, and self-blame for failing to conform to the intensive mothering ideal. Studies that have explicitly examined IMB (or aspects of IMB) have been mostly qualitative, cross-sectional, and some have been conducted with smaller samples of primarily privileged mothers with children of varying ages, some of whom were not employed (e.g., Garey, 1999; Hattery, 2001). Findings from these qualitative studies suggest that some full-time employed mothers support conventional views of motherhood, like those reflected in an intensive mothering ideology; however, the nature of the samples (i.e., with variation in employment status and work hours) used in past work have limited scholars’ ability to capture variation and examine changes in IMB among full-time employed mothers.
In contrast to what has been proposed by Hays (1996), the 2008 National Study of the Changing Workforce reported that a minority of women endorsed the idea that mothers should be primary caregivers for their children (Galinsky, Aumann, & Bond, 2009). Other national studies (e.g., The National Survey of Families and Households) have examined beliefs and attitudes about childrearing and maternal employment, but their measures only reflected pieces of IMB as conceptualized by Hays (Kroska & Elman, 2009). Thus, the information gained through these studies may be missing key aspects of mothering beliefs that have been identified as salient among employed mothers in the qualitative literature. To date, no studies have examined IMB in community samples of full-time employed mothers, examined IMB over time, or described the contextual correlates of IMB.
Background
Intensive Mothering Defined
Intensive mothering ideology, as described by Hays (1996), consists of a set of beliefs about children and maternal behaviors. Hays organized these beliefs into three domains of focus: (a) sacred children/sacred mothering, (b) the responsibility of individual mothers, and (c) intensive methods of childrearing. Part of an intensive mothering ideology is the belief that children are innately good, that their innocence is sacred and something that must be protected (by mothers), and being a mother is the most important role a woman can ever have. Because IMB place a high degree of responsibility for children’s welfare on mothers, the importance of mothers’ commitment to and effect on children is intensified. Intensive methods of childrearing are, according to Hays, child-centered and require all of mothers’ time, energy, and resources. Accordingly, children’s needs and desires should be mothers’ first priority, and it is assumed that this can only be accomplished when mothers deprioritize their own needs and personal goals and are readily available to meet children’s physical and emotional needs. Implied in Hays’ theorizing is that the time required to effectively meet the expectations of intensive mothering does not leave room for full-time employment because it limits mothers’ availability to their children and places the responsibility for childrearing in the hands of another, presumably less skilled, caregiver. In other words, intensive methods of childrearing stem from the belief that good mothers are not employed. Also implicit in Hays’ theorizing is that IMB are particularly salient for mothers of young children because of heightened social pressures for mothers to assume primary caregiving responsibilities while children are not yet in school and viewed as most vulnerable.
Theoretical Foundations
Peplau’s (1983) theorizing expanded traditional conceptualizations of roles by identifying behavioral, affective, and cognitive dimensions of roles. Peplau defined roles as consistent individual behavioral patterns that exist within the context of close relationships, but she acknowledged the possibility that beliefs and behaviors about a given role may be incompatible. Hays (1996) applied this line of thinking to her study of mothering beliefs by highlighting how some employed mothers in her study experienced ambivalence about their employment as a result of the psychological pull they felt to conform to cultural ideals of the nonemployed (intensive) mother. As such, we expected some variability in endorsement of IMB among the full-time employed mothers in our sample. Although Peplau’s conceptualization of roles underscores potential variability in mothering beliefs among full-time employed mothers, it does not explain why we might expect some full-time employed mothers to endorse a set of mothering beliefs that clearly undermine their involvement in full-time employment or how mothers’ endorsement of IMB might change over time. We find that gender perspectives and identity development theories are useful in filling this gap.
A feminist approach to the study of motherhood adopts a gender perspective, which argues that mothering is socially constructed and closely tied to social and political agendas (Baber & Allen, 1992; Ferree, 1990, 2010). Feminist scholars have highlighted the ways in which mothering beliefs are tied to oppressive social structures that work to sustain inequities based on gender, class, and race (Allen, 2001; Flax, 1979). This perspective is useful for understanding why conventional beliefs about mothering, such as IMB, which place primary responsibility for childrearing on mothers, might persist despite the prevalence of employed mothers. Through a feminist lens, IMB are likely to persist in American culture as long as they favor socially and historically privileged groups. Hays (1996) adopted a feminist perspective by acknowledging how IMB limit women’s opportunities in the workplace, primarily in terms of time constraints (by suggesting mothers should be 100% available to their children) and level of workplace commitment (by suggesting that only primary caregivers are fully dedicated to mothering). Drawing from the feminist literature, we hypothesized that some of the mothers in our sample would endorse IMB, or aspects of IMB, even though these beliefs may not be supportive of their involvement in full-time employment.
Feminist scholarship also highlights how mothering beliefs are influenced by “multiple-intersecting historical forces,” which necessitates a consideration for how intersecting social and economic contexts afford mothers different resources (Ferree, 2010, p. 423). We therefore anticipated variation in endorsement of IMB by characteristics of mothers that place them in different social locations, such as their race, education, and marital status. For example, IMB, by definition, suggest an ideal of mothering that is more attainable by the economically privileged. Moreover, feminist scholars have argued that IMB do not reflect the lives of women of color, who have a longer history of integrating paid work and motherhood (Collins, 1994, 2000; Glenn, 1994). We expected that Black mothers and those with lower income and education levels would be less likely to endorse IMB than mothers with social or economic advantage because IMB are not culturally relevant to them or realistically attainable. We also expected differences in mothers’ endorsement of IMB by unique intersections of race, education, and marital status. For example, we expected to find similarities among single employed mothers’ IMB scores, regardless of race or education, because they share the responsibility of sole economic provider.
Theories of gender identity development in the family studies literature assume that individuals will seek to minimize discrepancies between their experiences and belief systems by adjusting aspects of their lives or their beliefs (Robinson, 2007). According to these perspectives, we would expect full-time employed mothers’ endorsement of IMB across the first year postpartum to either (a) be low and remain low (because “traditional” mothers would self-select themselves out of employment) or (b) decrease over time because of the apparent discrepancy between full-time employment and the ideological requirements of IMB (i.e., mothers as primary caregivers). We also examined several conditions under which IMB scores might change for employed mothers. Mothers raising their second or third child have already done the cognitive work needed to reconcile their employment status and mothering beliefs. Thus, we expected that the hypothesized decline in IMB scores would be more dramatic for first-time mothers, who were negotiating full-time employment and motherhood for the first time, than for multiparous mothers. Married mothers might have an easier time achieving their mothering ideals than single mothers because they have another adult with whom to share parenting and economic responsibilities. Thus, we expected to see more of a decline in IMB scores for single mothers, compared with married mothers. We also expected that less educated mothers would have higher IMB scores than more educated mothers at Time 1, and as a result their scores would decrease more over time in an effort to align their mothering beliefs with full-time employment.
Empirical Foundations
Hays (1996) described IMB using in-depth interviews and survey questions with a snowball sample of 38 mothers, purposefully stratified by social class and employment status (10 middle-class employed, 10 working-class and poor employed, 9 middle-class homemakers, 9 working-class and poor homemakers). The sample was mostly White (71%), married (87%), in their mid-30s. Although the purpose of Hays’s work was mainly theoretical and descriptive, she argued for the pervasiveness of IMB as exemplified in the words of study participants. However, the notable absence of quotes she provided from ethnic minority and working-class mothers who endorsed IMB raises the question of how much women from non-White or lower socioeconomic backgrounds endorsed this set of beliefs. Hays and others have contended that intensive mothering is a standard that only married, middle-class women can achieve (Fox, 2006). Yet the relationships between employed mothers’ endorsement of IMB and their marital and socioeconomic backgrounds have not been explicitly examined.
The majority of studies that are directly tied to the work of Hays (1996) have employed qualitative methods to examine the mothering beliefs of employed and nonemployed mothers (Christopher, 2012; Elvin-Nowak & Thomsson, 2001; Garey, 1999; Hattery, 2001; Johnston & Swanson, 2006, 2007; Sutherland, 2006). All these studies included at least some mothers who were working full-time hours, with convenience or snowball samples ranging from 15 to 98 mothers. Mothers in these studies were primarily White (with the exception of Christopher, 2012, whose sample was 50% White and 50% African American or Latina), married, and college educated, with some variability in occupation and the age of their children. It is notable that variability among full-time employed mothers in their beliefs about mothering was found in several studies, with some full-time employed mothers who endorsed mothering beliefs akin to IMB or reported feelings of guilt for not spending enough time with their children (Hattery, 2001; Johnston & Swanson, 2007). When background characteristics were considered, Black and single mothers were less likely to endorse IMB than White and married mothers (Christopher, 2012). Other studies found that full-time employed mothers attempted to reconcile competing tensions between employment and motherhood by reframing their mothering beliefs in a way that better accommodated their employment status (Christopher, 2012; Garey, 1999; Johnston & Swanson, 2006, 2007). For example, via in-depth interviews with 40 employed mothers of young children, Christopher (2012) found that employed mothers endorsed “extensive” mothering, which allowed mothers to remain child-centered (a key facet of intensive mothering) without being home with children full-time by assuming primary responsibility for organizing children’s care in their absence. These findings underscore variability among full-time employed mothers in their beliefs about mothering, but do not tell us how common these beliefs are among full-time employed mothers in a systematic way.
Quantitative studies based on the work of Hays (1996) are few, and limited to group comparisons of mothering beliefs by mothers’ employment status. Hattery (2001) used an 8-item measure of IMB and found that nonemployed mothers in her sample had higher mean levels of IMB than part-time and full-time employed mothers, and full-time employed mothers scored lowest on IMB. With an Australian sample of 112 college-educated and predominantly married mothers of young children, McDonald, Bradley, and Guthrie (2005) found that greater work hours were associated with more positive views of nonmaternal child care. These findings echo results from larger national surveys of gender role attitudes that suggest the majority of adult women endorse egalitarian beliefs about gender roles (e.g., Galinsky et al., 2009). However, previous large-scale studies have not focused specifically on full-time employed mothers of infants and have relied on single item measures of gender role attitudes, which do not assess all aspects of IMB, as conceptualized by Hays.
A related body of research has explored the meanings mothers’ ascribe to paid employment and parenting roles in community samples of middle- and working-class women. The degree to which dual-earner wives viewed themselves as economic providers and sole versus shared caretakers of their children, also referred to as provider-role attitudes, has been found to vary among employed mothers (Helms, Walls, Crouter, & McHale, 2010; Loscocco & Spitze, 2007; Perry-Jenkins, Seery, & Crouter, 1992). Although distinct from IMB, the provider-role literature underscores the importance of understanding role-specific sets of beliefs, in addition to global indicators of gender role attitudes. Yet a parallel set of role-specific beliefs about mothering, such as IMB, has not been examined for full-time employed mothers of infants.
The Current Study
The work of Hays and studies related to her work have yet to substantiate the extent to which full-time employed mothers of infants endorse IMB, the stability of IMB, and the contextual correlates of IMB. The current study offers an initial step in the quantitative literature to clarify somewhat contradictory conclusions of this earlier body of work and (a) describes the extent to which a community sample of full-time employed mothers of infants endorse global and specific aspects of IMB, (b) examines the stability of IMB across 1 year postpartum and for specific subgroups of employed mothers, and (c) examines contextual correlates of and mean differences in IMB.
Method
Procedures
Data were drawn from the “Weaving Work and Family: Implications for Mother and Child” study, a short-term longitudinal study of full-time employed mothers and their infants. To construct the sample frame, mothers were identified between May 2007 and November 2007 at a hospital in the southeastern region of the United States. A physician, and co-investigator on the research project, approached mothers (N = 704) after giving birth whose medical records indicated they were employed during pregnancy and provided a brief description of the study. Mothers were asked if they could be contacted at a later date to participate in the study if they met study criteria and lived within 30 minutes of the city in which the targeted hospital was located. The final sample frame consisted of 630 mothers (or 89% of potentially eligible mothers) who agreed to be contacted.
At 3 months postpartum, eligible mothers were contacted by phone and invited to participate if they were (a) currently employed full-time or were planning to be employed full-time (defined as 30 hours or more per week) at 4 months postpartum, (b) fluent in English, and (c) their child was free from any health conditions that would be considered severe. One hundred and four mothers were unable to be reached by phone, 10 mothers refused to participate, and 116 mothers were ineligible. Of the 288 eligible mothers, 217 agreed to participate and were interviewed within 10 days of their infant’s 4-month birthday. During face-to-face interviews, mothers responded to fixed-response questions about their work and family experiences as well as their own and their infants’ health when their infants were 4 (Time 1) and 16 (Time 2) months of age. Participants received $75 on completion of the study.
Participants
Sample sizes were 217 at Time 1 and 195 at Time 2. Mothers who dropped out of the study by Time 2 (n = 24) were, on average, younger, less educated, worked fewer hours per week, earned less annually, experienced greater economic hardship, and had greater depressive symptoms than mothers who remained in the study from Time 1 to Time 2. Because the focus of this study was on full-time employed mothers, those working less than 30 hours per week were excluded from all analyses. In addition, the 4 mothers who did not identify as non-Hispanic White or Black were excluded from all analyses. Five mothers who had unemployed husbands were dropped from analyses as well. The final analytic sample at Time 1 consisted of 205 full-time employed mothers and 181 full-time employed mothers at Time 2. Seventy-three percent of mothers self-identified as White and 27% as Black. Mothers were in their mid-30s, on average, and the majority of mothers (82%) had some college education or a 4-year degree. Mothers’ work hours ranged from 30 to 65 (M = 39.84 hours, SD = 5.23). Among mothers who reported having a spouse or partner (n = 159), 100% were dual-earner couples. At Time 1, 68% (n = 137) of mothers were working daytime shifts and 32% (n = 65) were working nonstandard shifts. Mothers’ mean yearly income was $40,050 (SD = 26,052). Mean annual household income was $77,293 (SD = 58,360). Descriptive statistics are summarized in Table 1.
Descriptive Statistics for Baseline Sample at 4 Months Postpartum (N = 205).
N = 184 for mother’s income.
Measures
Intensive Mothering Beliefs
A 21-item measure of IMB, the Intensive Mothering Beliefs Scale, was developed for this study. The measure was developed to assess specific dimensions of IMB identified by Hays (1996) and Hattery (2001), including maternal employment, childrearing, self-sacrificing, and mothering as a natural talent. Mothers completed the Intensive Mothering Beliefs Scale at 4 and 16 months postpartum (Time 1 and Time 2, respectively) and rated their agreement with statements such as, “Mothers should stay at home to care for their children,” “Childcare is the responsibility of the mother,” and “Mothers should always place children’s needs before their own” on a scale of 1 (strongly disagree) to 5 (strongly agree), with 3 = neither agree nor disagree. Three items were dropped due to negative correlations with the overall scale to improve reliability. Although not an explicit part of her description, Hays (1996) discussed intensive mothering beliefs as a continuous construct, something that mothers could endorse to a greater or lesser degree. As such, the remaining 18 items were averaged, with higher scores indicating greater endorsement of intensive mothering beliefs. Cronbach’s alpha for intensive mothering beliefs was .75 at Time 1 and .74 and Time 2.
Demographic Characteristics
Mothers reported their own age in years, highest level of education, and marital status (married = 1, not married = 0). Mothers self-identified as belonging to one of the following racial groups: White, Black/African American, America Indian or Alaskan Native, Asian, or Other. They also classified themselves as Hispanic/Latino or not. These categories were collapsed into non-Hispanic White (coded as 0) and non-Hispanic Black (coded as 1).
Economic Characteristics
Mothers reported their personal annual earnings as well as their total family income for the previous year. Conger et al.’s (1990) measure of economic hardship was used to assess mothers’ perceptions of economic strain and financial need in the past 12 months. Specifically, mothers indicated how much difficulty they had paying their bills (1 = a great deal of difficulty, 3 = some difficulty, 5 = no difficulty at all) and at the end of each month if they had 1 = more than enough money left over, 2 = some money left over, 3 = just enough to make ends meet, and 4 = not enough to make ends meet. In addition, mothers were asked if they had engaged in 16 different behaviors that indicate financial need (e.g., “used savings to meet daily living expenses”) to which they could respond yes (coded as 1) or no (coded as 0). Economic hardship items were standardized and summed to form a composite measure with higher scores indicating greater economic hardship.
Work Schedule
Mothers reported whether they worked days, nights, evenings, or variable shifts. These categories were collapsed into standard/daytime shifts (coded as 0) and nonstandard shifts (coded as 1).
Parity
Mothers indicated the number of dependent children living at home. Mothers with one child were coded as 1 (“first-time mothers”); mothers with more than one child were coded as 0 (“multiparous mothers”).
Results
Our first goal was to examine the extent to which full-time employed mothers of infants endorsed global and domain specific IMB. To address this goal, we first examined the mean and standard deviation for average IMB scores (across all IMB items) at 4 months postpartum, followed by a descriptive analysis of individual items on the IMB scale. Because responses to individual IMB items were averaged, IMB scores had a possible range from 1 to 5, with higher scores indicating greater endorsement of IMB. At Time 1, mothers’ scores on the IMB Scale ranged from 1.50 to 4.22, with an average score of 2.62 (SD = 0.48). These scores reflect that, on average, most mothers in our sample reported between a 2 (“moderately disagree”) and a 3 (“neither agree nor disagree”) on the Intensive Mothering Beliefs Scale. To understand what percentage of mothers endorsed IMB, on average, we analyzed frequencies for the mean global IMB scores. We found that nearly 17% of the mothers endorsed IMB (i.e., had mean IMB scores greater than 3, our neutral response option).
To further address our first research goal concerning mothers’ endorsement of specific domains of IMB, we also examined descriptive statistics and frequencies for individual items on the Intensive Mothering Beliefs Scale at Time 1 (Table 2). First, we examined the means and standard deviations for each item. Descriptive statistics revealed variation in mothers’ endorsement of specific aspects of IMB, with items such as “Child care is women’s work” having a mean score of 1.37 (SD = 0.93) and other items such as “Mothers should always place children’s needs before their own” having a mean score of 4.36 (SD = 1.04). These results suggest that, on average, mothers strongly endorsed some aspects of IMB, but not others. To determine if mothers endorsed, did not endorse, or felt indifferent about specific aspects of IMB, responses for “strongly agree” and “moderately agree” were combined, and responses for “strongly disagree” and “moderately disagree” were combined. This resulted in three response categories: agree, neither agree nor disagree, and disagree. Results suggested that mothers’ endorsement of IMB varied at the item level (Table 2); for some items nearly all mothers felt similarly and for other items responses were relatively equally divided. For example, the vast majority of mothers agreed with the statement, “Childcare should be shared by men and women,” and disagreed with the statement, “Child care is solely the responsibility of the mother.” However, mothers were almost equally divided in their responses to statements such as, “Mothers are ultimately responsible for how children turn out” (45.8% agreed, 48.6% disagreed). The degree to which mothers “neither agreed nor disagreed” also varied at the item level. For example, 100% of mothers had an opinion (i.e., either agreed or disagreed) about the statement “Men should leave the childrearing to women,” whereas nearly 14% of mothers responded neutrally (i.e., neither agreed nor disagreed) to the statement “Mothers should stay at home to care for their children.”
Endorsement of Intensive Mothering Beliefs at Time 1 (N = 205).
Note. IMB = intensive mothering beliefs.
Item was reverse coded when computing average IMB scores.
Our second goal was to examine the stability of mothers’ IMB over time, for the entire sample and separately by parity, marital status, and education. To address this goal, we conducted a 2 (Parity) × 2 (Time) repeated-measures ANOVA, treating time as the within-subjects factor and parity as the between-subjects factor. The results of this analysis allowed us to determine if mothers’ IMB changed from Time 1 (4 months postpartum) to Time 2 (16 months postpartum) and whether mothers’ parity was related to changes in IMB across this 12-month period of time. The within-subjects effect of time was nonsignificant, F(1, 179) = 0.27, p = .61, indicating that mothers’ average IMB scores did not change from 4 to 16 months postpartum. The main effect for parity was nonsignificant, F(1, 179) = 1.04, p = .31. The interaction effect for Time × Parity also was nonsignificant, F(1, 179) = 0.68, p = .41 (M = 2.63 and 2.64 for first-time mothers at Time 1 and Time 2; M = 2.58 and 2.55 for multiparous mothers at Time 1 and Time 2), indicating that patterns of IMB over time did not vary based on parity. The same analyses were conducted separately for education and marital status as the between-subjects factors. Although the main effect for education was significant, F(1, 179) = 21.48, p = .00, the interaction effect for Time × Education was nonsignificant, F(1, 179) = 3.31, p = .07 (M = 2.76 and 2.70 for low educated mothers at Time 1 and Time 2; M = 2.44 and 2.47 for highly educated mothers at Time 1 and Time 2), indicating that patterns of IMB over time did not vary based on education. The main effect for marital status was significant, F(1, 179) = 4.20, p = .04, and the interaction effect for Time × Marital Status was significant, F(1, 179) = 7.08, p = .01 (M = 2.78 and 2.65 for single mothers at Time 1 and Time 2; M = 2.55 and 2.57 for married mothers at Time 1 and Time 2), indicating a decline in IMB over time for single, but not married mothers.
Our third goal was to examine the contextual correlates of IMB for full-time employed mothers of infants. To accomplish this goal, we examined correlations between continuous variables and IMB, mean differences in IMB for categorical variables, two-way interactions between categorical variables and IMB, a regression model including all contextual variables, and descriptive statistics for individual IMB items by education group. As a first step, we computed Pearson correlations between mothers’ global IMB scores and the continuous demographic, economic, and employment related indicators including mothers’ age, education, personal income, total household income, economic hardship, employment hours, and number of children at Time 1. Mother’s age, personal income, total household income, and education were associated negatively with IMB (r = −.22, −.30, −.31, and −.32, p = .00, respectively). Mother’s economic hardship was associated positively with IMB (r = .22, p = .00). Thus, younger, lower income, and less educated mothers were, the higher their global IMB scores.
Next, we conducted a series of one-way ANOVAs to examine mean level differences in the endorsement of IMB by mother’s race, marital status, work schedule, and parity. Results showed that Black mothers endorsed IMB more than White mothers (F[1] = 10.37, p = .00, M = 2.78, SD = 0.46, for Black mothers; M = 2.56, SD = 0.47, for White mothers). Single/never married mothers endorsed IMB more than married/cohabiting mothers (F[2] = 7.37, p = .00; M = 2.88, SD = 0.41, for single mothers; M = 2.57, SD = 0.46, for married/ cohabiting mothers). Mothers working nonstandard schedules endorsed IMB more than mothers working standard/daytime schedules (F[1] = 16.55, p = .00, M = 2.82, SD = 0.44, for nonstandard; M = 2.53, SD = 0.47, for standard/daytime). First-time and multiparous mothers did not differ in average IMB scores.
To further accomplish our third goal and describe IMB in terms of mothers’ background variables, we conducted a series of exploratory analyses to determine which variables were most strongly related to IMB. We began with preliminary analyses to determine if there were differences in education and marital status by race. Results of one-way ANOVAs suggested that White mothers had higher education levels, on average, than the Black mothers (M = 6.20, SD = 1.49, for White mothers; M = 5.42, SD = 2.09, for Black mothers), F(1, 203) = 8.83, p = .00. White mothers were more likely to be married than Black mothers (M = 0.92, SD = 0.27, for White mothers; M = 0.38, SD = 0.49, for Black mothers), F(1, 203) = 98.50, p = .00.
Based on the results of our preliminary analyses, we examined mean level differences in IMB by intersections of race and education, and race and marital status. Results of two-way ANOVA with race and education as fixed factors and IMB as the dependent variable suggested a significant main effect for race, F(1, 197) = 5.91, p = .02, and education, F(3, 197) = 5.91, p = .00. The interaction term for race and education was nonsignificant, F(3, 197) = .081, p = .49. Results of two-way ANOVA with race and marital status as fixed factors and IMB as the dependent variable suggested the main effects for race and marital status, as well as the interaction term for race and marital status were nonsignificant. To examine unique intersections of race, marital status, and education, we computed a new variable that divided mothers into one of eight social status groups based on two levels of race (White/Black), education (high/low), and marital status (married, not married). A one-way ANOVA was conducted with social status group as the independent variable and average IMB at Time 1 as the dependent variable. The overall model was significant, F(7, 197) = 5.69, p = .00. Post hoc Tukey HSD comparisons revealed that married, White, low educated mothers had higher average IMB scores than married, White, highly educated mothers (Mdiff = 0.27, p = .01); Single, Black, low educated mothers had higher average IMB scores than married, White, highly educated mothers (Mdiff = 0.47, p = .00); Single, Black, low educated mothers had higher IMB scores than single, White, highly educated mothers (Mdiff = −1.06, p = .03).
Last, to determine which factors were the most important in predicting IMB, we conducted a multiple linear regression with mothers’ age, number of children, marital status, race, and education as predictors of IMB at Time 1. Results suggested that the overall model was significant, F(5) = 6.22, p = .00; however, education emerged as the only significant predictor of IMB (Table 3). To get a better sense of the relationship between education and IMB, we examined frequencies and compared means for individual IMB items by education group. Results suggested significant mean differences by education group for 10 out of 18 items on the IMB measure (Table 4). In general, the low education group was consistently higher in their endorsement of IMB than the high education group. Significant differences were related to maternal employment (e.g., “Mothers should work outside the home only if their families need the money”), childrearing (e.g., “Nurturing children is something that comes naturally to women”), and self-sacrifice (e.g., “Mothers should always place children’s needs before their own”).
Summary of Multiple Linear Regression Analyses for Variables Associated With Mothers’ Intensive Mothering Beliefs Scores at 4 Months Postpartum (N = 205).
p < .01.
Endorsement of Intensive Mothering Beliefs at Time 1 by Education Level (n = 109 High School Degree, GED, or Some College; n = 96 4-Year or Advanced Degree) a .
Data in parentheses are for the higher education group.
One-way ANOVA between groups comparisons with 1 degree of freedom.
p < .05. **p < .01.
Discussion and Conclusion
The importance of women’s beliefs about their work and family roles has been emphasized in the literature, yet few studies have directly examined intensive mothering beliefs among full-time employed mothers of infants. The current study provided a closer look at the construct of IMB by examining the extent to which full-time employed mothers of infants endorsed IMB, the stability of IMB, and contextual correlates of IMB. Our findings suggest that IMB are not consistently endorsed by full-time employed mothers of infants. More specifically, mothers’ average endorsement of IMB when their infants were 4 months old ranged from “moderately disagree” to “neither agree nor disagree” on the IMB scale developed for the study. This finding contrasts the conclusions reached by scholars using qualitative methods with samples of primarily White employed and nonemployed women (Elvin-Nowak & Thomsson, 2001; Garey, 1999; Hattery, 2001; Johnston & Swanson, 2006; Sutherland, 2006), but is similar to some findings from studies using nationally representative samples who reported more egalitarian beliefs by adults, on average (Galinsky et al., 2009). Similarly, our findings suggest that most employed mothers hold beliefs about mothering that are congruent with their employment status. Thus, statements by Hays and others concerning how common IMB are among full-time employed mothers are not supported by the current findings.
It is important to consider, however, that some of the full-time employed mothers (albeit a minority of our sample) endorsed IMB to some degree. More specifically, 17% of our sample had mean IMB scores greater than 3, which suggests endorsement of IMB. Theories of attitude development purport that individuals will eventually alter their beliefs or behaviors as a way of minimizing internal discrepancies (e.g., Klein, 1984). Yet, in strained economic times, mothers of infants may feel pressure to remain employed full-time, despite their beliefs that good mothers should not be employed. Notably, in our sample, the more economic hardship mothers reported, the higher their IMB scores. In situations where families rely on women’s income to make ends meet, mothers may continue to embrace intensive mothering ideals despite the economic necessity of their employment.
Closer examination of individual items on the IMB scale revealed that mothers’ endorsement of IMB varied at the item level. The distribution of scores was highly skewed for some items (with almost 100% endorsement or nonendorsement), but relatively evenly distributed for other items. For example, nearly all mothers (97%) disagreed with the statement, “Child care is solely the responsibility of the mother.” In contrast, responses to the statement “Mothers are ultimately responsible for how children turn out” were relatively evenly distributed between agree and disagree response categories (i.e., 46% agreed, 48% disagreed, 6% neither agreed nor disagreed). In examining IMB at the item level, we were able to observe that beliefs associated with self-sacrificing were endorsed more, on average, among the full-time employed mothers of infants in this sample than beliefs associated with other aspects of intensive mothering. Perhaps, as a way of resolving any cognitive dissonance that might have resulted from combining full-time employment and mothering (Higgins, 1987), the mothers in our sample rejected certain aspects of intensive mothering that necessarily placed them in the home (e.g., being primary caregivers), whereas they tended to endorse aspects of intensive mothering that could be accomplished within the context of full-time employment (e.g., believing that mothers are naturally nurturing). This explanation is consistent with research by Johnston and Swanson (2006, 2007), who noted that the full-time employed mothers in their sample defined availability (one key facet of intensive mothering) in terms of being emotionally and psychologically available to children, instead of physically available, which allowed them to be good mothers in the context of full-time employment. Alternatively, mothers may have different views of what behaviors symbolize the prioritizing of children’s needs. For example, although some mothers may forego employment or tailor their work schedule as a means of prioritizing their children’s needs (e.g., Garey, 1999), mothers who are employed full-time may view their financial contributions to the family as accomplishing the same goal. Based on the patterning of mothers’ average responses to IMB items, it appears as if the mothers in our sample defined good mothering in terms of self-sacrifice, shared parenting responsibilities, quality childcare, and mothers as economic coproviders. The endorsement of the value of self-sacrifice alongside economic coprovision, quality childcare, and shared parenting may suggest that contemporary full-time employed mothers construct an ideology of good mothering that recognizes multiple ways to prioritize and care for children including providing for them economically and securing high-quality substitute and shared care (with partners) for children. Johnston and Swanson (2007) similarly noted that some employed mothers in their sample sacrificed their personal needs as a way of making more time for children and fulfilling their perceived mothering responsibilities. Our findings contrast the writings of Hays (1996) who suggested that IMB are inherently at odds with full-time employment, particularly when specific domains of IMB are considered separately.
We expected that endorsement of IMB would decline from 4 and 16 months postpartum, but our results showed that IMB were stable for full-time employed mothers of infants. Exposure to employment, for women, is associated with less “traditional” beliefs about family roles (Kroska & Elman, 2009). Because the mothers in our sample were employed prior to the birth of their children, it could be that they already endorsed less traditional mothering beliefs as a result of exposure to egalitarian beliefs in their workplaces or as a cognitive strategy when anticipating work–family issues. Thus, any cognitive “work” to align their mothering beliefs with their experiences might have been accomplished prior to this study. We also hypothesized that the decline in IMB would be more dramatic for first-time mothers, single mothers, and less educated mothers. Contrary to our hypothesis, general nonendorsement of IMB across the first year postpartum appears to be the case for full-time employed mothers regardless of parity and education. That being a first-time mother was not highly influential on changes in mothers’ endorsement of IMB over time was surprising to us, particularly given studies that have demonstrated the numerous cognitive changes that occur as women adapt to motherhood (Cowan & Cowan, 1992). Likewise, it was surprising that changes in IMB over time did not differ by mothers’ education, despite the contemporaneous association between education and IMB at Time 1. However, we did find that single mothers’ endorsement of IMB declined, whereas married mothers’ IMB scores remained stable over time. This finding supports our hypothesis that single mothers would be higher on IMB at Time 1 and would have a difficult time carrying out the intensive mothering ideal as sole economic provider and without the support of a partner/spouse. Yet it should be noted that single mothers’ average IMB scores declined from 2.78 to 2.65, both in between “neither agree nor disagree” and “moderately disagree” on our scale. Thus, the observed decline in IMB for single mothers was not dramatic in terms of its practical interpretations. Nonetheless, that single mothers made cognitive adjustments around their mothering beliefs across the first year postpartum and married mothers did not is noteworthy.
Last, we examined the contextual correlates of and mean-level differences in IMB. Our findings suggested that full-time employed mothers were more likely to endorse IMB when they were younger, had lower incomes, experienced greater economic hardship, had lower levels of education, were Black, single/never married, and worked nonstandard shifts. These findings are somewhat counterintuitive because, taken together, these factors describe women who are less likely to have the opportunities and resources to adhere to intensive mothering standards. Particularly with regard to race, we expected that, given their longer history of involvement in paid work compared with White mothers, Black mothers would be less likely to endorse a set of beliefs that generally does not support involvement in paid work outside the home (Collins, 1994, 2000). It should be noted, however, that mean-level comparisons by race, marital status, and work schedule showed that although our comparisons were significant, the average higher scores equated to “neither agree nor disagree” (i.e., a score of 3, on a scale from 1 to 5). Based on these average scores, it would be inappropriate to conclude that Black mothers endorsed IMB and White mothers did not.
In addition to main effects, we examined interactions among key demographic variables (race, education, and marital status) and their relationship with IMB. Our findings suggested that although race and education were independently associated with IMB, the interactions between race and education or race and marital status were not associated with IMB. However, when we examined unique configurations of race, education, and marital status together, we found that certain subgroups of mothers differed from each other on IMB. Specifically, White and highly educated mothers were consistently lower on IMB than mothers in other social locations, particularly those with lower education levels. Mothers from social locations that afford them fewer opportunities and resources (e.g., single, Black, low education) had higher IMB scores than mothers from more privileged backgrounds (e.g., married, White, higher education). Taken together with the results of our regression analyses, it appears as if education plays a major role in differentiating mothers on IMB. Through the experience of higher education, mothers are exposed to new ways of thinking and by virtue of their educational status are afforded greater social status as well. As a result, highly educated mothers might see more possibilities for themselves in society, in addition to caregiving. Because IMB deemphasizes mothers’ personal goals and careers and emphasizes their responsibilities to children, highly educated mothers might perceive this set of beliefs as limiting their potential for success. Women in Blair-Loy’s (2003) study, for example, were able to maintain professional careers while raising children by endorsing mothering schemas that emphasized the benefits of their employment for children’s development. Similarly, higher educated mothers in our sample were more likely than low educated mothers to endorse statements such as “Working outside the home can help women to be better mothers” and less likely to endorse statements such as “Mothers should stay at home to care for their children.”
Limitations and Future Directions
Although our study advanced earlier work by addressing some notable gaps in the literature, analyses for this study were conducted with Black and White mothers only, which limit the generalizability of our findings. Mothers from other ethnic groups were omitted because there were too few in our sample to make meaningful comparisons. Given the growing Latino and Asian immigrant populations in the United States, the mothering beliefs of women from these backgrounds are important to consider in future research, particularly because some scholars have noted variability in mothering beliefs among employed Latino mothers (Segura, 1994). Notably absent from previous work in this area (as well as in the present study) is an examination of fathers’ and spouses’/partners’ beliefs about mothering. For example, to what extent are fathers’ and mothers’ beliefs about mothering congruent? Given Peplau’s (1983) emphasis on the relational nature of roles (both in terms of how they are formed and enacted), fathers’ and spouses’/partners’ endorsement of IMB may have important consequences for mothers’ personal well-being and family relationships.
Although our sample size was adequate for analyses involving the entire sample (and for some between-group comparisons), our cell sizes dropped dramatically when examining intersections of race, marital status, and education in relation to IMB. For example, there were only six mothers who fell into the “single, Black, high education” group. Although some significant findings emerged, our nonsignificant findings should be interpreted with caution and considered preliminary at this point. Future studies employing larger samples would be better able to examine differences in IMB between groups of mothers from specific backgrounds.
Intensive mothering ideology has been theorized as generally detrimental for women’s well-being because it places primary responsibility for caregiving on mothers and suggests a set of behaviors that place high demands on mothers’ time, energy, and psychological resources. Despite the average nonendorsement of IMB among the mothers in this study, we did find that approximately 17% of our sample endorsed IMB, which suggests that some full-time employed mothers continue to endorse “traditional” mothering beliefs, such as IMB, and do not always select themselves out of full-time employment. Sixty-six percent of mothers in our “low education” group, for example, agreed with the statement “Mothers should stay at home to care for their children.” This highlights that mothers’ beliefs do not always align with their employment status, a finding that has been underscored in recent studies regarding beliefs about breadwinning and family roles (Helms et al., 2010; Helms-Erikson et al., 2000). This also begs the question of how easily can full-time employed mothers make cognitive shifts around their mothering beliefs, particularly when their children are small. For full-time employed mothers of infants who endorse IMB, what effects, if any, does endorsing this set of beliefs have on them? Role and discrepancy theorists (Higgins, 1987; Peplau, 1983; Rogers, 1959), as well as scholars in this area (Hays, 1996; Johnston & Swanson, 2006, 2007; Sutherland, 2006), have suggested that tensions between IMB and full-time employment can result in negative psychological outcomes for mothers, specifically feelings of depression and guilt related to their employment. Further research is needed to examine these linkages as they may have implications for maternal well-being, and subsequent child and marital outcomes.
Prior to this study, the concept of IMB existed primarily in the qualitative literature on mothering and theoretical writings of feminist scholars. This study was a first attempt to capture the construct of IMB quantitatively and with a community sample of full-time employed mothers of infants. Findings from this study provide preliminary evidence that full-time employed mothers, on average, do not endorse IMB. What this study does not tell us, however, is how full-time employed mothers are defining “good” mothering. For the majority of our sample who did not endorse IMB, what mothering beliefs do they endorse and how do these beliefs align with their full-time employment status? Our item-level examination of the IMB Scale suggests that certain dimensions of IMB may be more relevant for full-time employed mothers of infants, and mothers’ education is a key factor in understanding their endorsement of IMB. Our results also suggest that intersectionality is important in understanding IMB for full-time employed mothers. Through our examination of intersections of key background variables, we uncovered a complex pattern of associations that has not yet been captured by previous studies. We believe that the present study has scratched the surface of a larger set of questions that will be fruitful for scholars to examine, as women continue to construct their mothering identities within the context of full-time employment and an ever-changing cultural and economic climate.
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Eunice Kennedy Shriver National Institute for Child Health and Development (R21 HD48601).