Review of Economics of the Household

, Volume 14, Issue 3, pp 507–527

Are pregnant women happier? Racial and ethnic differences in the relationship between pregnancy and life satisfaction in the United States

Authors

    • Hamilton College
  • Stephen Wu
    • Hamilton College
Article

DOI: 10.1007/s11150-014-9239-8

Cite this article as:
Hagstrom, P. & Wu, S. Rev Econ Household (2016) 14: 507. doi:10.1007/s11150-014-9239-8

Abstract

This paper examines the relationship between pregnancy and life satisfaction for US women of childbearing age using a large sample from the 2005 to 2009 waves of the Behavioral Risk Factor Surveillance System. The results show strong differences by race and ethnicity. Pregnancy has a significant positive correlation with happiness for Whites and Hispanics, but no relationship for Blacks. This differential in the marginal effect of pregnancy is in addition to a general decrease in satisfaction for Black women, independent of being pregnant. The results cannot be explained by differences in other demographics such age, income, education, or physical health status. Within each racial/ethnic group, the results are consistent across different categories for all these characteristics. Racial and ethnic differences in the effects of pregnancy on support from others can partly explain this result. For Whites and Hispanic women, pregnancy increases their feelings of social and emotional support from others, while pregnant Black women report lower levels of social and emotional support than non-pregnant Black women.

Keywords

Life satisfactionWell-beingPregnancyRaceEthnicity

JEL Classification

D1I00Z1

A growing literature addresses the circumstances and life events that influence life satisfaction or happiness: being healthy, being obese, or having a fulfilling job, for example. Spurred on by research such as Oswald (1997) and Easterlin (2001), which relate happiness to incomes and economic performance, other studies have sought to measure the impact of life events such as marriage, divorce, winning the lottery, military combat, or losing a job on overall life satisfaction (Stellman et al. 2000; Clark and Oswald 2002; Lucas et al. 2003; Gardner and Oswald 2006, 2007; Ballas and Dorlin 2007). Recent papers by Myrskylä and Margolis (2012) and Clark and Georgellis (2013) use longitudinal data to study the impact of children on parental happiness. That work suggests that maternal happiness tends to decline post birth, with age and marital status being key moderators. We contribute to this body of research while addressing a topic with significant public health implications, the relationship between pregnancy and life satisfaction, a measure commonly used in the happiness literature.

Previous research establishes that happiness during pregnancy is positively associated with positive birth outcomes and negatively related to behaviors considered to be risky during pregnancy. Blake et al. (2007) find that being unhappy during pregnancy increases the likelihood of smoking, drug use, depression, and becoming the victim of partner violence. Such outcomes may all result in adverse health infant outcomes. Related research addressing the higher infant death rates among black families finds disparities that cannot be explained by levels of income, employment status, education, or genetic theories (Jackson 2007). In response, we ask how a woman’s race or ethnicity in the United States affects the impact of the pregnancy on life satisfaction. Related research suggests that race may influence the intensity of the pregnancy-life satisfaction relationships through differentials in health insurance or payment systems, social support systems, physical burdens of pregnancy, socio-economic status, pregnancy intendedness, the degree of paternal support and the residential status of the father. In connection with the preceding literature, our findings suggest the need for improved social emotional screening during pregnancy.

For many women, announcing a pregnancy is met with congratulations and compliments, suggesting that pregnancy constitutes a happy event. In such a case, the value placed on the extra attention and emotional support, in addition to the anticipation of a child, could well lead to an increase in life satisfaction prior to the child’s birth. To the contrary, the expected costs of parenting a child, the strain on relationships, the physical discomfort, the potential loss of earnings, the possible social stigma, or the anticipation of a changed lifestyle may outweigh the positives and lead to a diminished level of life satisfaction.

Within the broader research on subjective well-being, numerous studies focus on the relationship between having children and happiness. Research in sociology consistently finds that having children decreases well-being along a number of dimensions (Glenn and McLanahan 1981, 1982; McLanahan and Adams 1987, 1989; Evenson and Simon 2005). Parents report lower levels of marital happiness and overall life satisfaction and higher levels of anxiety and depression than non-parents. Blanchflower (2009a) summarizes the economics literature on this issue and reports that well-being is generally lower among families with children. Kahneman et al. (2004) use the day reconstruction method to show that although interaction with kids is listed as one of the most enjoyable daily experiences, taking care of one’s children is one of the least enjoyable activities. A few studies, however, show more mixed evidence. Kohler et al. (2005) find that the results depend on the number of children in the household as well as the gender of the parent. Having a first child increases well-being of parents, but additional children beyond the first child decrease well-being for females, and have no effect for males. Herbst and Ifcher (2012) show the effect is sensitive to the covariates included in the regressions and in some cases parents are happier than comparable non-parents. Yamauchi (2010) finds that parental life satisfaction also varies with the availability of high quality child care. These results, however, do not speak to the relationship between happiness and pregnancy, a period shown to have significant implications for child health outcomes. Furthermore, studies of women post-pregnancy cannot capture this within pregnancy effect; women with children who are less happy now may have been overjoyed while anticipating a child and subsequently overwhelmed by the reality of parenting. Our data allow us to focus specifically on the time period during which a woman is pregnant while controlling for a variety of factors that may affect life satisfaction while pregnant, including prior parenting experience or the ability to provide financially.

1 The relationship between pregnancy and life satisfaction

The economics literature provides no clear theories or predictions about the impact of pregnancy on life satisfaction. For some, becoming pregnant may bring personal value or respect from others. Edin and Kefalas (2005) suggest that having children may provide low-income youth with a sense of purpose and provide a turning point toward a better life. Deutsch (2001) argues that having children, particularly for married women, encourages the family to resolve differences. Alternatively, the “focusing illusion,” predicts that when people consider one particular aspect of their lives, they tend to overestimate its impact on their well-being. If this theory holds, women may overestimate the positive impact of having a child. Evidence of the focusing illusion has been found in the anticipation of marriage (Lucas et al. 2003), moving to a sunnier location (Schkade and Kahneman 1998), and increasing income (Kahneman et al. 2006). Powdthavee (2009, 2010) discusses the focusing illusion with respect to having children, while Clark et al. (2008) find that men and women report higher levels of life satisfaction the year before and the year after having a first child. Although pregnancy and life satisfaction would seem to be positively correlated for many women, for others the relationship may go in the opposite direction. For example, pregnancy and happiness may be negatively related if the pregnancy is unintended or if the timing of the pregnancy is problematic, for example while one is unemployed or in school. During pregnancy, a woman may also experience physical or emotional stress or anticipate financial difficulties in raising a child. The degree of such difficulties may depend on one’s relationship status or the depth of one’s emotional support network.

Racial disparities in birth outcomes, stress, depression, and risky behaviors during pregnancy may be due to the pregnancy itself or may be explained by other socioeconomic factors such as income levels and education. In our empirical analysis, we specifically examine the degree to which race and ethnicity play a significant role in the relationship between pregnancy and well-being. For example, race and ethnicity are correlated with income. Those who are in lower income groups may be less able to afford the financial costs of an additional child, and the anticipation of this economic strain may result in diminished life satisfaction. Also, the average maternal age at births and at first births varies across racial and ethnic groups and younger mothers may be less equipped to care for their children. Furthermore, out-of-wedlock and unintended pregnancies are more common for minorities than for whites (Kost et al. 2012; Delhendorf et al. 2013). Limited social support from the father may decrease the probability that pregnancy increases life satisfaction. In a clinical study of black-white difference in the quality of life during early pregnancy, Liu et al. (2013) find that black mothers reported a lesser degree of social support by a spouse, boyfriend or significant other, a factor that significantly lowered measures of health related quality of life outcomes relative to white mothers. Similarly, O’Hara (2009) reports strained marital relationships, childcare-related stressors, insufficient social support, and lower socioeconomic status, all of which were reported during pregnancy, contribute to postpartum depression.

2 Related literature

Although there is a lack of research directly focusing on the impact of pregnancy on women’s life satisfaction, there is significant work on a number of psychological effects of pregnancy on both mothers and their children. Most studies focus on negative measures of well-being such as depression and anxiety. Other research seeks to explain the effect of such anxiety on child outcomes such as low birth weight, pre-term births, and infant mortality.

Orr et al. (2002) find associations between antenatal depression and the probability of pre-term births for black women, but relatively few studies pursue the theme of racial and ethnic differences in the occurrence of antenatal depression. Orr et al. (2006) find greater risk of antenatal depressive symptoms among black relative to white pregnant women, though they do not compare their findings to a non-pregnant sample. Specifically, they find black pregnant women to be 1.5 times more likely to suffer either mild or severe levels of depression than white pregnant women. In contrast, Williams et al. (2007) find lower levels of depression risk during pregnancy for black women than for white women. Canady et al. (2008) use data on 2,731 women in the Pregnancy Outcomes and Community Health Survey to study the role of race, socioeconomic status, and reported measures of discrimination on antenatal depression. They posit that it is the degree of discrimination encountered rather than race that influences the onset of depression. While their results show a positive association between several measures of race discrimination and depression, the results lose their statistical significance in a model with a full vector of socioeconomic characteristics.

The above literature finds conflicting results regarding the occurrence of anxiety and depression during pregnancy and the impact of race and ethnicity on depression during pregnancy. In an effort to explain negative birth outcomes, another branch of research explores the racial and ethnic differences in the sources of stress experienced during pregnancy. Such stresses include higher poverty rates for black and Hispanic women, racism and discrimination, the lack of emotional support, and the lack of prenatal and general medical care (Jackson, 2007). Data from the 2004–2007 Oklahoma Pregnancy Risk Assessment Monitoring System (PRAMS) show that in the 12 months prior to delivering a child, black women are more likely to be homeless, lose a job, or experience relationship problems. In the subsequent analysis of low birth weight, however, the only statistically significant factors affecting low birth weights for black women are: living in an urban area, having had a previous low birth weight baby, and the number of medical complications during pregnancy.

For this paper, we use survey data from the Behavioral Risk Factor Surveillance System (BRFSS), an annual survey conducted by the Center for Disease Control (CDC) in the United States, to study self-reported life satisfaction among women of childbearing age. We find that pregnancy is associated with increased life satisfaction for white and Hispanic women but has little or no association for black women. This differential in the marginal effect of pregnancy is in addition to a general decrease in satisfaction for black women, independent of being pregnant. These results hold at all income, education, and age categories, and they are not due to differences in self-reported physical health. Instead, we find that differences in the level of social support such as those documented by Liu et al. (2013), Wiemann et al. (2006), Koniak-Griffin et al. (1993), and Sagrestano et al. (1999) can partly explain these racial and ethnic differences in life satisfaction. We discuss this idea later in the paper.

3 Data and descriptive statistics

Our data are taken from several waves of the BRFSS, a nationally representative survey conducted by the US CDC. The BRFSS is the world’s largest on-going telephone health survey, with several hundred thousand respondents each year. The survey has been conducted annually since its inception in 1984; our analysis uses five waves of the survey (2005–2009). The BRFSS collects information on health risk behaviors, preventive health practices, and health care access primarily related to chronic disease and injury. Beginning in 2005, the survey started asking the following question about subjective well-being: “Overall, how satisfied are you with your life?” Respondents are able to answer one of the following: very satisfied, satisfied, dissatisfied, or very dissatisfied. Some recent research on subjective well-being has utilized the answers to these questions in the BRFSS (Blanchflower 2009b; Oswald and Wu 2010).

Importantly, for our purposes, the survey also asks the following question, “To your knowledge, are you now pregnant?” The survey does not contain any additional information about the pregnancy such as how far along the pregnancy is, or how much difficulty a woman has had during this pregnancy. However, the survey does contain demographic variables such as age, income, educational attainment, and employment and marital status. This allows us to account for the fact that the correlation between pregnancy and overall well-being may vary according to these characteristics. While the level of detail and the large sample size provide significant advantages, the cross-section nature of the data precludes a longitudinal analysis. Because we do not observe life-satisfaction in prior periods we cannot refute the possibility of reverse causality in which pregnancy is caused by happiness. Even so, our results shed new light on the moderators of the pregnancy-happiness relationship. To measure race, we use the respondent’s self-reported answer to the survey question, “which one of these groups would you say best represents your race?” The possible categories include white, black/African American, Asian, native Hawaiian/pacific islander, Native American/Alaskan native, and other. To simplify, we collapse the Asian and native Hawaiian/pacific islander groups into one category, “Asian”. A separate question asks whether or not the respondent is Hispanic. We code our variable “Hispanic” to be equal to one for all respondents that report yes to this question. In order to have mutually exclusive categories for race/ethnicity, we replace the original race variables (white, black, Asian, Native American, other) to be equal to zero if an individual reports being Hispanic. Therefore, our variable “white” is equal to one for all non-Hispanic whites. Likewise, those in the “black” category are all non-Hispanic blacks, and we use the analogous coding for Asians, Native Americans and other races.1

Because our focus is on the relationship between pregnancy and life satisfaction, we restrict our sample to women between the ages of 18 and 45 with non-missing information. The remaining sample contains 367,339 respondents. To give an overall sense of the data, we present means and standard deviations for the entire sample in columns 1–2 of Table 1. Mean life satisfaction is 3.38 (based on a 1–4 scale, where 4 represents “Very satisfied” and 1 represents “Very dissatisfied”). Slightly over half the sample has a household income below $50,000, while more than a quarter has over $75,000 in annual income. African Americans and Hispanics each comprise about 11 % of the sample, while 3 % of the sample are Asians and 2 % of the sample are Native Americans. A little over a third of the individuals have a college degree, while nearly 60 % are married. For this age group of women, 4 % report being currently pregnant.2 Columns 3–4, 5–6, and 7–8 show the summary statistics for whites, blacks and Hispanics, respectively. Overall, the samples of blacks and Hispanics have lower levels of income, education, and life satisfaction. The percentage of Hispanics that are currently married is somewhat lower than for whites (53 and 65 %, respectively), while the percentage of blacks that are married is only 28 %.
Table 1

Summary statistics

 

(1–2)

(3–4)

(5–6)

(7–8)

All

Whites

Blacks

Hispanics

Mean

SD

Mean

SD

Mean

SD

Mean

SD

Life Sat (1–4)

3.38

0.62

3.42

0.62

3.22

0.65

3.31

0.60

Inc < 10 K

0.05

0.23

0.04

0.18

0.12

0.32

0.11

0.30

10 K ≤ Inc < 15 K

0.05

0.21

0.03

0.18

0.08

0.27

0.10

0.30

15 K ≤ Inc < 20 K

0.07

0.26

0.05

0.22

0.13

0.33

0.15

0.36

20 K ≤ Inc < 25 K

0.09

0.28

0.07

0.26

0.13

0.33

0.15

0.36

25 K ≤ Inc < 35 K

0.12

0.32

0.11

0.31

0.16

0.36

0.14

0.35

35 K ≤ Inc < 50 K

0.16

0.37

0.17

0.37

0.15

0.36

0.13

0.33

50 K ≤ Inc < 75 K

0.19

0.39

0.21

0.41

0.12

0.32

0.10

0.30

Inc ≥ 75 K

0.28

0.45

0.33

0.47

0.12

0.32

0.12

0.33

Age

34.66

7.29

35.11

7.19

33.82

7.45

33.08

7.29

White

0.73

0.44

Black

0.11

0.31

Asian

0.03

0.16

Hispanic

0.11

0.31

Native Am.

0.02

0.14

Other race

0.01

0.08

Less than HS

0.03

0.15

0.01

0.08

0.01

0.09

0.15

0.35

Some HS

0.06

0.24

0.04

0.20

0.09

0.28

0.15

0.36

HS Grad

0.25

0.44

0.24

0.43

0.33

0.47

0.30

0.46

Some college

0.29

0.45

0.30

0.46

0.31

0.46

0.22

0.42

College grad

0.37

0.48

0.41

0.49

0.26

0.44

0.18

0.38

Married

0.59

0.49

0.65

0.48

0.28

0.45

0.53

0.50

Divorced

0.11

0.31

0.11

0.31

0.11

0.32

0.09

0.28

Separated

0.03

0.19

0.03

0.16

0.07

0.25

0.07

0.25

Widowed

0.01

0.10

0.01

0.10

0.01

0.11

0.01

0.10

Partner

0.05

0.21

0.04

0.20

0.03

0.18

0.10

0.30

Never married

0.21

0.41

0.16

0.37

0.49

0.50

0.20

0.40

Self employed

0.07

0.26

0.08

0.27

0.04

0.21

0.06

0.23

Unemployed

0.06

0.24

0.05

0.21

0.11

0.32

0.09

0.28

Homemaker

0.16

0.37

0.17

0.37

0.05

0.23

0.26

0.44

Student

0.05

0.22

0.05

0.21

0.07

0.26

0.05

0.22

Retired/disable

0.04

0.20

0.04

0.19

0.07

0.25

0.04

0.19

Paid employ

0.61

0.49

0.62

0.48

0.64

0.48

0.50

0.50

Pregnant

0.04

0.19

0.04

0.19

0.03

0.18

0.05

0.21

Observations

367,339

264,195

37,393

38,567

Data taken from the 2005–2009 waves of the Behavioral Risk Factor Surveillance System (BRFSS). Sample is all women between the ages of 18–45

4 Documenting the relationship between pregnancy and life satisfaction

In order to examine the relationship between pregnancy and life satisfaction, we begin our analysis by estimating the following regression equation:
$$LifeSat = \beta_{0} + \beta_{ 1} {\text{X}} + \beta_{ 2} Pregnant + \varepsilon$$
where X represents a vector of demographic and background characteristics including income categories, age, education, race, number of children in the household, and marital and employment status. We use ordinary least squares to estimate these models, but we have also re-estimated these regressions using ordered probit and logit models, and the substantive results are unchanged. Therefore, for ease in interpreting marginal effects, we present OLS estimates in our main tables.

This first specification compares pregnant women with non-pregnant women, but in subsequent specifications, we use more refined comparison groups to account for the different reference groups with which people may compare themselves. In later analysis, we look at differences in life satisfaction between pregnant and non-pregnant women after dividing the sample in the following ways: race/ethnicity, current number of children, income levels, age range, education level, and marital status.

Column 1 of Table 2 shows results for a regression using the entire sample of women between the ages of 18–45 for the 2005–2009 waves of the BRFSS. Consistent with much of the literature on the economics of happiness, we find that life satisfaction is positively related to income, education, being married, and being employed.3 The data also show the broadly confirmed U-shape relationship between life satisfaction and age, with a negative linear coefficient and a positive squared coefficient.4 There also exist differences in life satisfaction by race and ethnicity even after controlling for the aforementioned demographic variables. Hispanic and Native American women report higher average life satisfaction than whites, while the coefficients for blacks, Asians, and those indicating their race as “other” are all negative and statistically significant.
Table 2

Pregnancy and life satisfaction by race

Variables

(1)

(2)

(3)

(4)

(5)

All

All

Whites

Blacks

Hispanics

10 K ≤ Inc < 15 K

0.05***

(0.01)

0.05***

(0.01)

0.06***

(0.01)

0.04**

(0.02)

0.04**

(0.01)

15 K ≤ Inc < 20 K

0.11***

(0.01)

0.11***

(0.01)

0.13***

(0.01)

0.10***

(0.01)

0.08***

(0.01)

20 K ≤ Inc < 25 K

0.16***

(0.01)

0.16***

(0.01)

0.20***

(0.01)

0.17***

(0.01)

0.10***

(0.01)

25 K ≤ Inc < 35 K

0.22***

(0.01)

0.22***

(0.01)

0.27***

(0.01)

0.20***

(0.01)

0.14***

(0.01)

35 K ≤ Inc < 50 K

0.29***

(0.01)

0.29***

(0.01)

0.34***

(0.01)

0.26***

(0.01)

0.20***

(0.01)

50 K ≤ Inc < 75 K

0.36***

(0.01)

0.36***

(0.01)

0.40***

(0.01)

0.30***

(0.02)

0.26***

(0.02)

Inc ≥ 75 K

0.45***

(0.01)

0.45***

(0.01)

0.50***

(0.01)

0.39***

(0.02)

0.34***

(0.02)

Age

−0.03***

(0.00)

−0.03***

(0.00)

−0.04***

(0.00)

−0.03***

(0.00)

−0.02***

(0.00)

Age squared/1,000

0.39***

(0.02)

0.39***

(0.02)

0.46***

(0.03)

0.37***

(0.07)

0.21***

(0.06)

Black

−0.01***

(0.00)

−0.01***

(0.00)

   

Asian

−0.07***

(0.01)

−0.06***

(0.01)

   

Hispanic

0.03***

(0.00)

0.04***

(0.00)

   

Native Am.

0.03***

(0.01)

0.04***

(0.01)

   

Other race

−0.07***

(0.01)

−0.07***

(0.01)

   

Some high school

−0.03***

(0.01)

−0.03***

(0.01)

0.02

(0.02)

0.08*

(0.04)

−0.01

(0.01)

High school grad

0.03***

(0.01)

0.03***

(0.01)

0.09***

(0.02)

0.17***

(0.04)

0.01

(0.01)

Some college

0.03***

(0.01)

0.03***

(0.01)

0.11***

(0.02)

0.12***

(0.04)

0.02*

(0.01)

College grad

0.11***

(0.01)

0.11***

(0.01)

0.19***

(0.02)

0.20***

(0.04)

0.10***

(0.01)

Married

0.17***

(0.00)

0.17***

(0.00)

0.18***

(0.00)

0.10***

(0.01)

0.14***

(0.01)

Divorced

−0.01

(0.00)

−0.01

(0.00)

0.01**

(0.00)

−0.01

(0.01)

−0.00

(0.01)

Separated

−0.10***

(0.01)

−0.10***

(0.01)

−0.15***

(0.01)

−0.02

(0.01)

−0.03**

(0.01)

Widowed

−0.04***

(0.01)

−0.04***

(0.01)

−0.04***

(0.01)

0.00

(0.03)

−0.00

(0.03)

Partner

0.02***

(0.01)

0.02***

(0.01)

0.04***

(0.01)

−0.07***

(0.02)

0.04***

(0.01)

# Children in HH

0.01***

(0.00)

0.01***

(0.00)

0.01***

(0.00)

−0.01**

(0.00)

0.01**

(0.00)

Self employed

0.06***

(0.00)

0.06***

(0.00)

0.07***

(0.00)

0.04**

(0.02)

0.07***

(0.01)

Unemployed

−0.14***

(0.00)

−0.14***

(0.00)

−0.17***

(0.01)

−0.12***

(0.01)

−0.09***

(0.01)

Homemaker

0.07***

(0.00)

0.07***

(0.00)

0.07***

(0.00)

0.07***

(0.02)

0.05***

(0.01)

Student

0.05***

(0.01)

0.05***

(0.01)

0.07***

(0.01)

0.05***

(0.01)

−0.02

(0.02)

Pregnant × White

 

0.12***

(0.01)

   

Pregnant × Black

 

−0.02

(0.02)

   

Pregnant × Asian

 

0.04

(0.03)

   

Pregnant × Hispanic

 

0.07***

(0.02)

   

Pregnant × Native Am.

 

0.06

(0.04)

   

Pregnant × Other race

 

0.01

(0.06)

   

Pregnant

0.10***

(0.01)

 

0.12***

(0.01)

0.01

(0.02)

0.08***

(0.02)

Constant

3.51***

(0.03)

3.52***

(0.03)

3.51***

(0.03)

3.35***

(0.09)

3.37***

(0.07)

Observations

323,903

323,903

240,654

33,937

33,189

R-squared

0.130

0.130

0.140

0.070

0.079

Life satisfaction is measured on a 1–4 scale (4 = very satisfied, 3 = satisfied, 2 = dissatisfied, 1 = very dissatisfied). Indicator variables for month and year of survey included in all regressions. Omitted category for household income is <$10,000. Omitted category for education is no high school. Omitted category for employment status is employed for wages. Omitted category for marital status is never married. Standard errors are in parentheses

*** Significant at 1 %; ** Significant at 5 %; * Significant at 10 %

We are particularly interested in the relationship between pregnancy and life satisfaction. Column 1 shows that in the overall sample, we find that being pregnant is associated with increased average life satisfaction of approximately 0.1 points on a 1–4 scale. The pregnant coefficient is significant at the 1 % level. The magnitude of this estimated pregnancy effect on life satisfaction is nearly as large as the effect of being employed (relative to being unemployed) and more than half of the size of the effect of being married.

When we look at the relationship between pregnancy and happiness by race/ethnicity, we find that the results are not consistent across all categories. In column 2, we look at separate interaction terms between pregnancy and six different racial and ethnic groups (whites, blacks, Hispanics, Asians, Native Americans, and others). The results show that the magnitude of the relationship between being pregnant and life satisfaction is greatest for whites: the coefficient of 0.12 is slightly larger than the coefficient in column 1. For Hispanics, pregnancy is also positively and significantly related to happiness, though the coefficient is about half of the size (0.07). However, none of the interaction terms for the other four groups are statistically different from zero. Note that the pregnancy-race interaction effects are in addition to the general race effects that we have included in the life satisfaction regressions. We are cautious about placing too much emphasis on the results for Asians, Native Americans, and those categorized as “Other” because the sample sizes are quite small. These three groups comprise 3, 2, and 1 % of the sample, respectively, and for the remainder of our analysis, we limit our sample to all whites, blacks, and Hispanics. On the other hand, while there are a large number of blacks in our sample, pregnancy does not have any noticeable effect on life satisfaction for this group. In fact, the coefficient on the interaction between black and pregnant is negative, though not statistically distinguishable from zero. In columns 3–5, we conduct separate analyses for each of these groups to allow the effects of all other variables to vary by race/ethnicity. We find the same results—pregnancy increases life satisfaction for white and Hispanic women, but not for black women.

What might account for the difference in these effects? One issue to explore is whether there are differences between first and subsequent pregnancies, as first pregnancies may be more significant events than later ones. If there are differences in family size and number of pregnancies across racial and ethnic lines, then this may partly explain why pregnancy and life satisfaction show a strong positive correlation for some groups, but not others. In Table 3, we run separate regressions by the number of other children in the household. While this is not a perfect measure for the number of prior pregnancies a woman has had, it is likely strongly correlated. For those with zero children currently living in the household (column 1), the likelihood that this is a first pregnancy is probably greater than the analogous likelihood for someone who currently has children living in the house (columns 2 and 3). The results show that for whites and Hispanics, women without any children living in the household receive a larger life satisfaction boost from pregnancy than those who already have children in the household, a result consistent with Myrskylä and Margolis (2012). For whites, the coefficient on pregnancy is 0.18 for those with zero children living at home, 0.10 for those with one child, and 0.11 for those with two or more children living at home. While the magnitude is greatest for those with no children in the household, all three of the coefficients are statistically significant. For Hispanics, the analogous coefficients are 0.16, 0.05, and 0.05, though the second coefficient is only significant at the 10 % level. Interestingly, there is a large difference in the magnitude of the coefficients for those without children at home and the coefficients for those with at least one child at home, but the coefficients for those with one child and those with at least two children are indistinguishable. For blacks, none of the coefficients are statistically significant. Regardless of the number of children that are currently living the household, there is no relationship between pregnancy and life satisfaction of black women.
Table 3

Pregnancy and life satisfaction by number of children in household

Variables

(1)

0 children

Life satisfaction

(2)

1 child

Life satisfaction

(3)

2+ children

Life satisfaction

Pregnant × White

0.18***

(0.01)

0.10***

(0.01)

0.11***

(0.01)

Pregnant × Black

−0.00

(0.04)

−0.05

(0.03)

0.00

(0.03)

Pregnant × Hispanic

0.16***

(0.04)

0.05*

(0.03)

0.05***

(0.02)

Observations

81,909

72,700

152,518

R-squared

0.120

0.135

0.135

Life satisfaction is measured on a 1–4 scale (4 = very satisfied, 3 = satisfied, 2 = dissatisfied, 1 = very dissatisfied). All regressions include variables for age, age squared, race, income categories, employment status, marital status, and month and year of survey. Standard errors are in parentheses

*** Significant at 1 %; ** Significant at 5 %; * Significant at 10 %

There are a number of other possible reasons why the effect of pregnancy on well-being may vary across different racial and ethnic groups of the population. Those who are in lower income groups may be less able to afford an additional child, and it may be the correlation between race/ethnicity and income that is driving this result, as discussed by Rich-Edwards et al. (2006). To check on this possibility, we re-estimate the regressions for those in different income categories. Table 4 shows the results of these regressions. Column 1 uses the entire sample of whites, blacks, and Hispanics and interacts eight income categories with the indicator variable for being pregnant (but does not distinguish between racial or ethnic categories). It is striking to see that each of the eight coefficients is positive and statistically significant, and none is statistically different from any of the others. Note that the coefficients for the two lowest income categories (0.10 and 0.10) are nearly identical to the coefficients for the two highest income categories (0.12 and 0.11). In Columns 2–4, we replicate the analysis for three separate income categories, those with annual family incomes less than 25 K, those with incomes between 25 and 50 K and those with incomes over 50 K. Once again, pregnancy is positively associated with life satisfaction for white and Hispanic women in each of the income categories, but is not correlated with life satisfaction of black women for any income group. Even for blacks with household income greater than 50 K, the effect is not statistically different from zero; in fact, the point estimate is negative (−0.03). The results for other covariates such as age, education, and marital and employment status are similar to those in Table 2.
Table 4

Pregnancy and life satisfaction by income

Variables

(1)

All

Life satisfaction

(2)

Inc < 25 K

Life satisfaction

(3)

25 K ≤ Inc < 50 K

Life satisfaction

(4)

Inc ≥ 50 K

Life satisfaction

Preg × Inc < 10 K

0.10***

(0.02)

   

Preg × 10 K ≤ Inc < 15 K

0.10***

(0.03)

   

Preg × 15 K ≤ Inc < 20 K

0.08***

(0.02)

   

Preg × 20 K ≤ Inc < 25 K

0.11***

(0.02)

   

Preg × 25 K ≤ Inc < 35 K

0.11***

(0.02)

   

Preg × 35 K ≤ Inc < 50 K

0.08***

(0.01)

   

Preg × 50 K ≤ Inc < 75 K

0.12***

(0.01)

   

Preg × Inc ≥ 75 K

0.11***

(0.01)

   

Pregnant × White

 

0.14***

(0.02)

0.11***

(0.01)

0.12***

(0.01)

Pregnant × Black

 

−0.04

(0.03)

0.00

(0.03)

−0.03

(0.04)

Pregnant × Hispanic

 

0.05**

(0.02)

0.06**

(0.03)

0.10***

(0.03)

Observations

307,780

76,793

85,334

145,653

R-squared

0.132

0.054

0.045

0.040

Life satisfaction is measured on a 1–4 scale (4 = very satisfied, 3 = satisfied, 2 = dissatisfied, 1 = very dissatisfied). All regressions include variables for age, age squared, race, number of children in household, employment status, marital status, and month and year of survey. Standard errors are in parentheses. “All” indicates whites, blacks, and Hispanics

*** Significant at 1 %; ** Significant at 5 %; * Significant at 10 %

Tables 2, 3, 4 show that there are strong racial and ethnic differences in the effects of pregnancy on life satisfaction and these differences cannot be explained by family income or number of prior pregnancies. What other possibilities might explain these results? Data from a number of different sources show that black women tend to have children at earlier ages than whites and Hispanics.5 In the BRFSS sample studied here, we also see this to be the case. The average age of black women that are pregnant is 27.7, while the average ages of white and Hispanic pregnant women are 29.8 and 28.7, respectively. Perhaps those that become pregnant at earlier ages are less equipped to handle the stressors of pregnancy and parenthood. In Table 5, we examine this idea by running separate regressions for four age categories: women under 25, women between 25 and 30, women between 30 and 35, and women over 35. After controlling for the same demographic characteristics as in the earlier tables, we continue to observe positive correlations between pregnancy and life satisfaction for white and Hispanic women, but not for blacks. For whites, each of the coefficients is statistically significant at the 1 % level; for Hispanic women, the standard errors on the pregnancy coefficients are a bit higher for those between ages 30 and 35 and those over 35. However, pregnancy does not increase life satisfaction for black women in any of the age categories.
Table 5

Pregnancy and life satisfaction by age

Variables

(1)

Age < 25

Life satisfaction

(2)

25 ≤ Age < 30

Life satisfaction

(3)

30 ≤ Age < 35

Life satisfaction

(4)

Age ≥ 35

Life satisfaction

Pregnant × White

0.08***

(0.02)

0.12***

(0.01)

0.13***

(0.01)

0.12***

(0.01)

Pregnant × Black

−0.04

(0.03)

−0.00

(0.03)

0.06

(0.04)

−0.03

(0.04)

Pregnant × Hispanic

0.09***

(0.03)

0.12***

(0.03)

0.05*

(0.03)

0.04

(0.03)

Observations

28,308

45,014

60,750

173,708

R-squared

0.079

0.134

0.145

0.139

Life satisfaction is measured on a 1–4 scale (4 = very satisfied, 3 = satisfied, 2 = dissatisfied, 1 = very dissatisfied). All regressions include variables income categories, race, number of children in household, employment status, marital status, and month and year of survey. Sample includes whites, blacks, and Hispanics. Standard errors are in parentheses

*** Significant at 1 %; ** Significant at 5 %; * Significant at 10 %

Our results cannot be explained by differences in income, nor can they be explained by differences in age at pregnancy. It may be the case that differences in educational attainment or marital status explain the different effects of pregnancy on life satisfaction. In Table 6, we split the sample by educational attainment. Once again, the results are consistent across all columns of the table: pregnancy makes white and Hispanic women happier, but does not have a demonstrable effect on the happiness of black women.
Table 6

Pregnancy and life satisfaction by education

Variables

(1)

<HS grad

Life satisfaction

(2)

HS grad

Life satisfaction

(3)

>HS grad

Life satisfaction

Pregnant × White

0.13***

(0.04)

0.11***

(0.02)

0.13***

(0.01)

Pregnant × Black

0.00

(0.06)

0.00

(0.03)

−0.03

(0.02)

Pregnant × Hispanic

0.09***

(0.03)

0.07**

(0.03)

0.07***

(0.02)

Observations

22,027

75,260

210,493

R-squared

0.076

0.100

0.120

Life satisfaction is measured on a 1–4 scale (4 = very satisfied, 3 = satisfied, 2 = dissatisfied, 1 = very dissatisfied). All regressions include variables for age, age squared, income categories, race, number of children in household, employment status, and month and year of survey. Sample includes all whites, blacks, and Hispanics. Standard errors are in parentheses

*** Significant at 1 %; ** Significant at 5 %; * Significant at 10 %

In columns (1) and (2) in Table 7 we conduct separate analysis by marital/partner status. While the pregnancy effect for married women in column 1 is stronger than the cohabiting partner effect (column 2) for white women, the opposite pattern occurs for black women. Living with an unmarried partner is the only situation in which the average pregnancy effect for black women is positive, and this effect is marginally significant at the 10 % level. The Hispanic effect is identical across those who are married versus cohabiting. In columns (3–5) we distinguish between those who are single and divorced, single and separated, and those who are single and have never married. Among these three categories, only white women experience an increased level of satisfaction due to pregnancy, and this effect is positive across all three categories of singleness. Black and Hispanic women without a spouse or residential partner experience no statistically discernible pregnancy effect. Given these differences by relationship status, particularly the importance of having a spouse or partner present, we look for further evidence that social support from others acts as a moderator in the pregnancy and overall well-being relationship.
Table 7

Pregnancy and life satisfaction by marital status

Variables

(1)

Married

Life satisfaction

(2)

Partner

Life satisfaction

(3)

Divorced

Life satisfaction

(4)

Separated

Life satisfaction

(5)

Never married

Life satisfaction

Pregnant × White

0.12***

(0.01)

0.08**

(0.03)

0.13***

(0.04)

0.17***

(0.06)

0.08***

(0.02)

Pregnant × Black

−0.01

(0.03)

0.14*

(0.07)

−0.11

(0.09)

0.07

(0.10)

−0.03

(0.03)

Pregnant × Hispanic

0.09***

(0.02)

0.09**

(0.04)

0.01

(0.08)

−0.05

(0.08)

0.06

(0.04)

Observations

186,010

14,152

34,571

10,808

59,375

R-squared

0.071

0.076

0.096

0.053

0.071

Life satisfaction is measured on a 1–4 scale (4 = very satisfied, 3 = satisfied, 2 = dissatisfied, 1 = very dissatisfied). Partner in column (2) indicates cohabitating partner. All regressions include variables for age, age squared, income categories, race, number of children in household, employment status, and month and year of survey. Sample includes all whites, blacks, and Hispanics. Standard errors are in parentheses

*** Significant at 1 %; ** Significant at 5 %; * Significant at 10 %

5 Pregnancy, physical and emotional health, and social support

Thus far, we have shown that there are racial and ethnic differences in the relationship between pregnancy and life satisfaction. These differences cannot be explained by differences in household income or level of education. For Hispanic women, the pregnancy effects are somewhat greater for younger women and first time mothers. For black women, pregnancy is generally not related to increased life satisfaction with one exception: pregnant women have higher life satisfaction during pregnancy when they have a cohabiting partner. What else might account for the reasons that pregnancy increases life satisfaction for white and Hispanic women but not for black women? Earlier, we discussed research that finds some evidence of differences in the effects of pregnancy on various physical and mental health outcomes for women of different racial and ethnic groups. In Table 8, we examine how measures of general physical and emotional health are affected by pregnancy and whether there are any disparities by race or ethnicity. In column 1, the dependent variable is a self-reported measure of general health status, where the answers range on a scale from 1–5, with 5 indicating “excellent” health and 1 indicating “poor” health. The positive coefficients indicate that pregnant women feel better about their overall physical health than otherwise similar women who are not pregnant. We see that the effects of pregnancy on general physical health are roughly the same between whites, blacks, and Hispanics, with no statistical difference between any of the three interaction terms. In fact, the coefficient for blacks is slightly larger than the coefficient for whites and Hispanics. In column 2, the dependent variable is the answer to the following survey question, “Now thinking about your physical health, which includes physical illness and injury, for how many days during the past 30 days was your physical health not good?” Here, we see that the coefficients for the interaction terms pregnant × white and pregnant × black are not statistically significant, though the pregnant × Hispanic coefficient is negative and significant at the 1 % level. Hispanic women who are pregnant have fewer days in the last month with poor physical health than non-pregnant Hispanic women.
Table 8

Pregnancy and physical/mental health and emotional support

Variables

(1)

All

General health

(2)

All

Poor physical health days

(3)

All

Poor mental health days

(4)

All

Emotional and social support

Pregnant × White

0.14***

(0.01)

−0.08

(0.07)

−1.23***

(0.09)

0.10***

(0.01)

Pregnant × Black

0.16***

(0.03)

−0.22

(0.19)

−0.29

(0.23)

−0.06**

(0.03)

Pregnant × Hispanic

0.11***

(0.02)

−0.60***

(0.17)

−1.31***

(0.20)

0.13***

(0.02)

Observations

319,316

317,651

317,269

307,972

R-squared

0.169

0.063

0.067

0.081

General health is measured on a 1–5 scale (5 = excellent, 4 = very good, 3 = good, 2 = fair, 1 = poor). Poor physical (mental) health days are the number of days in the last 30 where physical (mental) health was not good. Emotional and social support is measured on a 1–5 scale, based on how often one gets the emotional and social support needed (5 = always, 4 = usually, 3 = sometimes, 2 = rarely, 1 = never). All regressions include variables for age, age squared, income categories, race, number of children in household, employment status, marital status, and month and year of survey. “All” indicates all whites, blacks, and Hispanics. Standard errors are in parentheses

*** Significant at 1 %; ** Significant at 5 %; * Significant at 10 %

Next we turn to assessing the effects of pregnancy on mental health. The dependent variable in column 3 is the answer to the following question, “Now thinking about your mental health, which includes stress, depression, and problems with emotions, for how many days during the past 30 days was your mental health not good?” The results show a stark contrast in the effects of pregnancy across racial and ethnic groups. For whites and Hispanics, pregnancy is associated with significantly fewer days of poor mental health over the course of the past month. However, for blacks, there is no effect of pregnancy on the number of poor mental health days.

Why might pregnancy improve mental well-being of expectant white and Hispanic women, but not expectant black women? One possibility is that there is a difference between the social support received by pregnant women of different racial and ethnic groups. Recent research has shown strong relationships between levels of social support and well-being. Seiger and Wiese (2011) use a longitudinal survey of European mothers and find that emotional support strongly impacts affective well-being. Rook (1987) provides evidence that companionship support increases happiness and decreases feelings of loneliness. Of particular interest for our analysis, Liu et al. (2013) and Elsenbruch et al. (2006) conclude that lack of social support is an important risk factor for maternal well-being during pregnancy. In column 4 of Table 8, we analyze the answer to a question that asks, “How often do you get the social and emotional support that you need?” The possible answers include “Never, Rarely, Sometimes, Usually, and Always”. Now, we see an even greater racial and ethnic disparity in the effects of pregnancy on this variable. For whites and Hispanics, pregnancy increases the degree to which women feel they receive social and emotional support from others. However, holding the other variables in the model constant, pregnant black women receive, on average, lower levels of social and emotional support than black women who are not pregnant. This coefficient is statistically significant at the 5 % level.

To test the robustness of this finding, we repeat the analysis of column 4 of Table 8 for different subsamples of the data, dividing the sample by marital status and by income. The results, shown in Table 9, are qualitatively similar. For white and Hispanic women, pregnancy generally increases the level of perceived social and emotional support, while for black women, pregnancy is negatively associated with this variable for every category except those living with an unmarried partner, where the coefficient is positive (though statistically insignificant for this sample).
Table 9

Pregnancy and emotional support

Variables

(1)

Married

Emotional and social support

(2)

Partner

Emotional and social support

(3)

Single

Emotional and social support

(4)

Inc < 25 K

Emotional and social support

(5)

Inc ≥25 K

Emotional and social support

Pregnant × White

0.12***

(0.01)

0.05

(0.05)

0.15***

(0.03)

0.16***

(0.03)

0.09***

(0.01)

Pregnant × Black

−0.08*

(0.04)

0.02

(0.11)

−0.06

(0.04)

−0.04

(0.05)

−0.09**

(0.04)

Pregnant × Hispanic

0.11***

(0.03)

0.10

(0.06)

0.19***

(0.05)

0.16***

(0.04)

0.07**

(0.03)

Observations

186,006

14,990

122,177

76,944

231,028

R-squared

0.036

0.073

0.071

0.041

0.027

All regressions include variables for age, age squared, race, number of children in household, employment status, and month and year of survey. Regressions (1) and (2) also include controls for income categories, while regressions (3) and (4) also include controls for marital status. Sample includes all whites, blacks, and Hispanics. Standard errors are in parentheses

*** Significant at 1 %; ** Significant at 5 %; * Significant at 10 %

We find strong evidence that there are stark racial and ethnic differences in the effects of pregnancy on emotional and social support from others. However, we acknowledge that the BRFSS has only one general question on this topic, so we are unable to address how pregnancy affects each of the specific types of social support. Seiger and Wiese (2011) distinguish between four types of social support: emotional support; instrumental support such as help with household chores and childcare; informational support such as giving advice on decision making; and companionship support. There is corroborating research from a number of small sample surveys that show that social support from fathers during and after pregnancy varies by race and ethnicity. The Los Angeles Mommy and Baby Survey (LAMBS, 2007), which samples 726 new mothers, asks detailed questions about the father’s involvement and emotional support during and after pregnancy. Relative to whites, African American husbands and partners were less likely to provide financial support, help with chores, put their name on the birth certificate or help with childrearing. Although Hispanic partners were not more likely than African American partners to put their name on the birth certificate, they did provide more support in the other categories. Koniak-Griffin et al. (1993) analyze a sample of 161 pregnant teenagers and show that blacks have significantly lower levels of social support than whites, and slightly lower support than Hispanics. Sagrestano et al. (1999) also find differences in levels of social support during pregnancy by racial and ethnic groups.

6 Conclusion

In this paper, we use a large sample of American women and find that there is a differential impact of pregnancy on the life satisfaction of white, Hispanic and black women. Further, we demonstrate that these differences in impacts are not explained by income levels, age, education, or marital status. In other results not shown here, we also dismiss health insurance status, and access to medical care as explanations for the observed differences by race and ethnicity.

So why does there fail to be a happiness effect of pregnancy only for black women? One plausible explanation supported by our findings concerns the impact of pregnancy on the level of social and emotional support received during pregnancy. White and Hispanic women report increased levels of emotional support during pregnancy, but pregnant black women experience lower levels of emotional and social support relative to non-pregnant black women. This result holds across different categories of marital status. Black women, regardless of whether they are married, experience lower levels of social support during pregnancy, while the analogous correlation is positive for white and Hispanic women, regardless of marital status. Existing research lends support to this explanation. The LAMBS survey shows that black husbands and partners were less likely to provide financial support, and help with child care responsibility than whites and Hispanics. Other research also shows racial and ethnic differences in the support received from partners, relatives, and friends during pregnancy. Given the established connections between happiness and infant outcomes, our findings suggest that social-emotional screenings may improve birth outcomes for African American women.

Our analysis here, based on a very large sample of both pregnant and non-pregnant women, shows strong results that are consistent with other related literature and robust to dividing the sample along many dimensions. Nonetheless, a few caveats are worth mentioning. First, as discussed earlier, our data are not longitudinal, so it is difficult to get at the issue of causality. It is possible that happier women become are more likely to become pregnant, rather than the other way around. However, it is still an interesting finding that the correlation between life satisfaction and pregnancy exists for white and Hispanic women, but not for black women, regardless of the direction of causality. Another issue worth mentioning is that the BRFSS does not give much detail about a women’s pregnancy. In particular, we do not have any information on the intendedness or wantedness of the pregnancy. If intendedness of pregnancy is correlated with race and life satisfaction, it may provide an alternative explanation. Some research has shown that black and Hispanic women have higher rates of unwanted pregnancies than white women (Aquilino and Losch 2005; Keeton and Hayward 2007). These differences may be an explanation for the differences in family and social support (Lifflander et al. 2007). Another piece of information that we do not know from the data is whether or not the pregnancy will eventually be terminated. It has been documented that abortion rates for blacks are several times higher than that for whites and Hispanics (U.S. Department of Health and Human Services 2009). Perhaps the negative effects of pregnancy on the support from others for black women can partially explain these relatively high rates of abortion. Future work may be able to uncover the reasons for these relationships.

Footnotes
1

An alternative would be to have the following mutually exclusive groups: white non-Hispanic, white Hispanic, blacks, Asians, Native Americans, and “other” (with each of the non-white groups including both Hispanics and non-Hispanics). Because the overwhelming majority of Hispanics in this sample categorize themselves as being either white or “other race”, the choice of categorization does not alter the results of the paper.

 
2

This is comparable with pregnancy rates for women between 18 and 45 in the 2006–2010 waves of the National Survey of Family Growth.

 
3

See Clark and Oswald (2002), Easterlin (2001), and Gardner and Oswald (2007).

 
4

See Di Tella et al. (2001) and Blanchflower and Oswald (2008).

 
5

In 2006, the average age of mother for first birth was 22.7 years for black mothers, compared to 26.0 and 23.1 for white and Hispanic mothers, respectively (U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, National vital Statistics Report, Volume 57, Number 7, January 2009).

 

Copyright information

© Springer Science+Business Media New York 2014