Content uploaded by Tomas Frejka
Author content
All content in this area was uploaded by Tomas Frejka on Dec 03, 2016
Content may be subject to copyright.
Tomas Frejka ()
e-Mail: tfrejka@aol.com
Childlessness in the United States 1
Tomas Frejka 2
3 4 5 6 7 8 9 Abstract This paper uses cohort fertility tables and periodic surveys to provide an overview of the 10 development of childlessness in the United States. Estimates of the temporarily, involuntarily, and 11 voluntarily childless are available. Childlessness has attracted considerable attention because it doubled 12 between the mid-1970s and the mid-2000s, from 10 to about 20 per cent. The U.S. has experienced 13 considerable variation in levels of childlessness. The challenging living conditions during the Great 14 Depression of the 1930s were the principal cause of high childlessness. The contrasting favorable 15 living standards and enlightened public policies of the late 1940s-1960s were instrumental in 16 maintaining low childlessness. For much of the 20th century, childlessness was higher among blacks 17 than among whites mainly because their living conditions were more difficult. Subsequently, black 18 childlessness declined below that of whites possibly due to rapid improvements in health and living 19 conditions which nonetheless remained inferior. Numerous additional factors have been shaping 20 childlessness in the U.S., including high female employment rates, the conflict between work and 21 family responsibilities, separation of spouses due to wars or incarceration, high costs of childrearing, 22 inadequate childcare infrastructure, insecurity of employment and income, uncertainty of spousal 23 relationships, and concern for the wellbeing of children. Because the interactions of factors which 24 shape childlessness are not well understood, it is difficult to make predictions. Future trends will 25 depend on the extent to which material conditions will facilitate or obstruct family formation, and 26 cultural norms and personal attitudes change. 27 28 Keywords Childlessness, Fertility, Fertility Intentions, United States 29 30 31 32 1 Introduction 33
In recent decades, childlessness among women in the United States has attracted a considerable amount 34 of attention in the professional literature, and is frequently discussed in newspapers and on radio and 35 television talk shows. This does not come as a surprise, as the percentage of women who do not have 36 any children by the end of their reproductive years doubled between the mid-1970s and the mid-2000s, 37 from about 10 to 20 per cent. Since then, however, the share of women who remain childless has been 38 declining: in 2010-12, the share was around 15 per cent (Table 1)1. While establishing the levels of and 39 the trends in childlessness is relatively simple, determining the circumstances and reasons which lead 40 women and couples to remain childless is more complex. 41 Three different sources of statistical data on childlessness are available in the U.S. This wealth of 42 data is almost as much a curse as it is a blessing. However, using data from all three sources one can 43 obtain a good approximate idea of the levels of and the trends in childlessness. Yet because each source 44 provides somewhat different data, it is difficult to determine which one most closely reflects reality. On 45 balance the positive aspect of good approximate information prevails. Moreover, the overall perception 46 provided by the three sources of data is consistent. Not only that. The available sources offer various 47 types of information, including some kinds which are relatively rare. One of the sources contains a time 48 series spanning an entire century, which is also broken down by race. Another source provides data not 49 only by race, but also for Hispanics. A third source contains data on whether women are temporarily, 50 voluntarily, or non-voluntarily childless, as well as information about women’s personal characteristics 51 and selected attitudes to work and family. These data are available for a span of close to four decades. 52 Knowledge which can be gleaned from all three sources of data is likely to be expanded in the future. 53
1 The levels of and trends in childlessness among women are based primarily on data from the Current Population
Surveys in Table 1, which is generally corroborated by data from the cohort fertility tables (Figure 2, 1970 cohort) and
from the National Surveys of Family Growth (Table 2, latest years)
Following this introduction, the sources of data are discussed. In section three levels of and trends 54 in childlessness are outlined. Section four deals with motivations and reasons for childlessness. Section 55 five discusses trends and circumstances of black childlessness. The chapter concludes with an epilogue. 56 57 58 2 Sources of Data 59
The three sources of statistical data on childlessness are cohort fertility tables (National Center for 60 Health Statistics of the Centers for Disease Control and Prevention), the biannual supplements on 61 fertility of the Current Population Survey (Census Bureau and Bureau of Labor Statistics), and the 62 National Survey of Family Growth (National Center for Health Statistics of the Centers for Disease 63 Control and Prevention [NCHS]). 64 65 66 2.1 The Cohort Fertility Tables consist of two sets. The first set is based on recorded period 67 fertility data for the years 1917-1973, and was prepared by Robert L. Heuser (1976). It provides 68 information on childbearing of complete and incomplete birth cohorts of 1868-1959. The second set 69 uses period data for 1960-2005, and was prepared by Brady E. Hamilton in collaboration with Candace 70 M. Cosgrove (2010). Hamilton and Cosgrove updated this set with period fertility data for 2006-2009. 71 It provides information on childbearing of complete and incomplete birth cohorts of 1911-1995. The 72 Heuser tables can be linked with the Hamilton and Cosgrove tables to create a series of data on 73 childlessness for 93 consecutive birth cohorts. 74 75 76 2.2 The Fertility Supplement of the Current Population Survey is one of 20 77 supplements sometimes included in the Current Population Survey (CPS), a monthly survey of 78 households conducted by the U.S. Census Bureau for the Bureau of Labor Statistics. The CPS collects 79 and maintains a comprehensive body of labor force data, including information on employment, 80 unemployment, hours of work, earnings, and other demographic and labor force characteristics. The 81 periodic fertility supplement provides data on the number of children women aged 15-50 have ever 82 had, and their characteristics. It is usually conducted every two years, but the intervals have varied 83 from one to four years (see Table 1 and Figure 3). Since the mid-1990s data on the U.S. Hispanic 84 population2 have been provided (Bachu 1995). 85 86 87 2.3 The National Survey of Family Growth (NSFG) gathers information on family life, 88 marriage and divorce, pregnancy, infertility, use of contraception, and men's and women's health; i.e. 89 data on fertility and on the intermediate factors that explain fertility. The NSFG was conducted by the 90 National Center for Health Statistics (NCHS) in 1973, 1976, 1982, 1988, 1995, and 2002. The most 91 recent NSFG covered the years 2006-2010 (Martinez et al. 2012). In these surveys childless women are 92 comprised of three categories defined as follows (Abma and Martinez 2006; Martinez et al. 2012): 93 Temporarily childless women are those who have not had any live births and expect a birth in the 94 future. 95 Involuntarily childless women are those with a fecundity impairment who reported to be sterile for 96 non-contraceptive reasons; subfecund, i.e. they reported difficulty conceiving or delivering a baby or 97 difficulty for partner to father a baby; or a doctor advised the woman never to become pregnant 98 because of a medical danger to her, her fetus or both; married or cohabiting women that have had a 99 three year period of unprotected sexual intercourse with no pregnancy. 100 Voluntarily childless women are those who do not expect to have any children, and are either fecund 101 or surgically sterile for contraceptive reasons. 102 Note that the cohort fertility tables are based on data from administrative birth records, whereas the 103 other two data sources are based on sample surveys. The sample surveys provide information on the 104 characteristics of mothers and their children which are not available in birth records. However, the 105 estimates of common measures based on the sample surveys are not precisely the same as those based 106 on administrative birth records. 107
2 The Office of Management and Budget (OMB) defines Hispanic or Latino as “a person of Cuban, Mexican, Puerto
Rican, South or Central American, or other Spanish culture or origin, regardless of race.” In data collection and
presentation, federal agencies are required to use a minimum of two ethnicities: “Hispanic or Latino” and “Not Hispanic
or Latino.”
108 109 3 Levels of and Trends in Childlessness 110
3.1 Cohort Fertility Tables 111
In any given birth cohort, the youngest women bear few children. With each passing year, these women 112 will have borne more children, and the share of women who remain childless declines. To ensure the 113 comparability of the rates of childlessness between cohorts, the data on the proportion of childless 114 women at the end of their childbearing years are assembled for each cohort. Figure 1 depicts the shares 115 of all U.S. childless women, and of white and black women at age 50 in the Heuser (1976) and in the 116 Hamilton and Cosgrove (2010) cohort fertility tables. 117 118 Fig. 1 Shares of childless women at age 50, all, white and black women, birth cohorts 1867-1960, 119 United States 120
121 Sources: Heuser (1976); Hamilton and Cosgrove (2010) 122 123 Among the 40 cohorts born between the late 1860s and the early 1910s, around 20 per cent of white 124 women remained childless. Women who lived through the main years of their childbearing period 125 during the core years of the historic economic depression of the 1930s—cohorts born between 1906 126 and 1911—experienced relatively high rates of childlessness, about 21 per cent. However, this was not 127 dramatically more than most of the preceding 40 cohorts. A rapid decline in the share of childless 128 women started with the 1913 birth cohort and lasted through the 1925 cohort that reached a childless 129 rate of nine per cent. A low share of childlessness among white women fluctuating between eight and 130 10 per cent was retained for almost 20 cohorts from the 1925 through the 1943 birth cohort. A 131 pronounced increase in the shares of childless women ensued, from 10 per cent among the 1943 cohort 132 to 18 per cent among the 1953 cohort. The childless rate at age 50 was close to 18 per cent for a few 133 cohorts and then started to decline to around 17 per cent in the 1959 and 1960 cohorts (Figure 1). 134 The long-term trends in the shares of childless black women differed from those of white women. 135 For about 60 cohorts, starting with those of the mid-1880s through those of the mid-1940s, black 136 women experienced higher rates of childlessness than white women. Notably, almost one-third of black 137 women who were in their most fertile years during the Great Depression of the 1930s remained 138 childless. With a time lag of about five cohorts shares of childless black women declined from 29 per 139 cent among women born in 1916 for more than 30 cohorts to a low of six per cent in the 1948 birth 140 cohort. Thereafter, the share of childless black women increased reaching a share of 11 per cent in the 141 1960 cohort (Figure 1). 142
Although numbers of births after age 40 have increased in recent years (Sobotka 2009), these still 143 tend to be relatively small. Consequently, trends in the shares of childless women at age 40 are 144 essentially the same as trends in the shares of childless women at age 50 (Figure 2). Thus the 145 delineation of trends can be extended for 10 additional cohorts, namely for U.S. women trends of 146 childless women can be obtained by observing trends of shares at age 40 for the 1960s birth cohorts. 147 These women concluded their childbearing during the 2010s, and their principal period of childbearing 148 was during the mid- to late 1980s and early 1990s. 149 Among white women the declining trend of childless women extended into the 1960s cohorts. The 150 share of childless women in the 1960 birth cohort at age 40 was 17.7 per cent and declined to 14.6 per 151 cent in the 1970 birth cohort (Figure 2). This implies that around 13 per cent of white women in the 152 1970 cohort will be childless at age 50. The rising trend in childlessness among black women of the 153 1950s cohorts stalled among the 1960s cohorts. The share of women who were childless at age 40 was 154 11.9 per cent among the 1960 birth cohort, and 12.1 per cent among the 1970 birth cohort (Figure 2). 155 This implies that around 11 per cent of black women in the 1970 cohort will be childless. 156 157 Fig. 2 Shares of childless women at age 40 (in per cent), all, white and black women, birth cohorts 158 1877-1970, United States 159
160
Sources: Heuser (1976); Hamilton and Cosgrove (2010) 161 162 It appears that shares of white and black childless women in the 1970 cohort will be quite similar. The 163 difference in the shares of white and black childless women in the 1950 cohort at age 40 was 10.2 164 percentage points which declined to 5.8 points in the 1960 cohort and to 2.5 points in the 1970 birth 165 cohort. 166 Levels and trends of overall shares of childless women follow the levels and trends of white women 167 quite closely. This is not surprising, as the majority of the U.S. population was and still is white, 168 although the percentage of whites has been declining. In 1900 about 88 per cent of the U.S. population 169 was white and 12 per cent was black (U.S. Bureau of the Census 1975). These percentages were 170 essentially maintained through 1970. As of 2000, whites comprised about 82 per cent and blacks 13 per 171 cent of the population (U.S. Census Bureau 2012). The effect of black childlessness on the overall 172 levels and trends is nonetheless discernable. When black childlessness is high the overall curve is 173 above the white one, and vice versa. 174 The share of all childless women at age 50 in the 1960 cohort was 15.5 per cent and at age 40 – 175 16.5 per cent, a difference of exactly 1.0 percentage point. The share of all childless women at age 40 176 in the 1970 cohort was 13.8 per cent. Thus it is virtually assured that the overall share of childless 177 women in the 1970 cohort at age 50 will be below 13 per cent, because the difference in the ten years 178
younger cohort was 1.0 percentage point and this difference of childlessness between ages 40 and 50 in 179 a particular birth cohort was growing. 180 181 182 3.2 Fertility Supplements of the Current Population Survey 183
In the fertility supplements of the Current Population Surveys parity distributions—and thus also the 184 shares of childless women—are provided for five-year age groups. Until recently the oldest age group 185 for whom these data were available was 40-44. Since 2012 the age group 45-50 has been added. Table 186 1 and Figure 3 are based on data for the 40-44 age group. Although childbearing does not end at age 187 44, this cut off was necessary to obtain long-term time series. 188 189 Table 1 Shares of childless women at ages 40-44, all, white, white non-Hispanic, black, and Hispanic 190 women, 1976-2012, United States 191
Survey year
Percent of women childless
Effect of
Hispanic on
White
childlessness (in
% points)
All
White
White Non-
Hispanic
Black
Hispanic
1976
10.2
n.a.
n.a.
n.a.
n.a.
n.a.
1980
10.1
n.a.
n.a.
n.a.
n.a.
n.a.
1985
11.4
n.a.
n.a.
n.a.
n.a.
n.a.
1990
16.0
n.a.
n.a.
n.a.
n.a.
n.a.
1994
17.5
18.0
n.a.
14.3
13.0
n.a.
1995
17.5
18.1
n.a.
15.1
10.1
n.a.
1998
19.0
19.5
20.1
17.0
14.5
-0.6
2000
19.0
19.2
20.3
17.7
10.9
-1.1
2002
17.9
17.9
18.5
19.2
13.1
-0.6
2004
19.3
19.1
20.0
21.3
13.8
-0.9
2006
20.4
21.2
22.5
16.4
14.4
-1.3
2008
17.8
18.0
17.9
18.0
18.9
0.1
2010
18.8
19.1
20.6
17.2
12.4
-1.5
2012
15.1
15.3
16.4
15.4
10.9
-1.1
Source: U.S. Census Bureau, Current Population Survey for selected years, June 1976 to June 2012. 192 193 According to these data the average share of all childless women aged 40-44 in the United States 194 increased from 10 per cent around 1980 to almost 20 per cent in the 2000s, i.e. the proportion of 195 childless women increased almost two-fold within 20 years. Toward the end of the 2000s and the early 196 2010s, childlessness declined (Table 1). In the mid-1990s the shares of white childless women were 197 almost 10 per cent higher than those of black women. By 2008-2012 the differences between white and 198 black women in the rates of childlessness had diminished (Figure 3 and Table 1). 199 When comparing childlessness of Hispanic women with childlessness among white and black 200 women one has to keep in mind that in U.S. statistics Hispanics are included in the categories of 201
“white” and “black.” Hispanics are considered an ethnic minority, not a race. It is nonetheless possible 202 to get an idea of the effect of Hispanic childlessness on overall levels of childlessness of the race 203 categories. Even though the Hispanic childlessness rate (5th numerical column in Table 1) is on average 204 about 30 per cent lower than childlessness of non-Hispanic white women (3rd col.), the difference 205 between the shares of all white childless women (2nd col. which includes white Hispanic women) and 206 non-Hispanic white women is relatively small, on average this difference is only 0.9 percentage points 207 (last col. in Table 1). The reason for such a small difference is that in 2010, for instance, Hispanic 208 women constituted only about 18 per cent of white women, although the share of Hispanics in the 209 population was increasing (U.S. Census Bureau 2011: Table 6). The effect of Hispanic black 210 childlessness on total black childlessness was even smaller as the proportion of Hispanics among 211 blacks was only about five per cent in 2010. 212 213 Fig. 3 Shares of childless women ages 40-44, white, black, and Hispanic women, 1976-2012, United 214 States 215
216
Source: U.S. Census Bureau, Current Population Survey for selected years, June 1976 to June 2012. 217 218 219 3.3 The National Surveys of Family Growth (NSFG) 220
Shares of childless women ages 40-44 rose from seven per cent in the 1973-1976 rounds to 18 per cent 221 in the 1995 round of the NSFG. In the rounds conducted during the 2000s, the shares of childless 222 women had settled at 15 per cent (Table 2 and Figure 4). Among childless women ages 40-44 the 223 smallest shares were experienced by the temporarily childless. If the measurements had been taken at 224
the end of women’s reproductive period, as was done in the cohort fertility tables, there would not be 225 any temporarily childless women. As women ages 40-44 is the oldest category that can be analyzed, 226 the temporarily childless women have a significant impact on the overall trends in childlessness. Since 227 women are postponing births to higher ages, a larger amount of births are borne by older women; thus, 228 an increasing proportion of women in the 40-44 age group still expect to bear children. While the share 229 of temporarily childless older women has been increasing steadily, it still represents only three per cent 230 of all women and around one-fifth of all childless women. The share of all women who are 231 involuntarily childless has been relatively stable at an average of five per cent. In the 1973-1976 232 rounds, the share of involuntarily childless women as a proportion of all childless women was 60 per 233 cent because the overall numbers of childless women were relatively small. In the latest rounds, about 234 one-third of childless women would probably want to have children, but for one reason or another—235 primarily related to a health issue—they have been unable to achieve this goal. 236 The NSFG definitions used to distinguish between voluntary and involuntary childlessness appear 237 to be straightforward and clear (see section 2.3 above). However, scholars have pointed out that an 238 unknown segment of the women who at the end of their reproductive period report being voluntarily 239 childless or having become involuntarily childless were postponing childbearing for various reasons 240 until it became too late for them to bear children (Rindfuss et al. 1988: throughout). In other words, 241 some, possibly many, women wind up being unintentionally childless as a result of having postponed 242 childbearing. Regardless of how the childlessness occurred, using NSFG definitions, the percentage of 243 the voluntarily childless increased from one-third in the 1970s rounds to approximately one-half of 244 childless women in subsequent rounds (Table 2). 245 246 Table 2 Women aged 40-44 and their childless status, National Survey of Family Growth, in per cent, 247 United States 248
All women
1973-1976
1982
1988
1995
2002
2006-2010
One or more children
93
88
86
82
85
85
Childless
7
12
14
18
15
15
Total
100
100
100
100
100
100
All women
1973-1976
1982
1988
1995
2002
2006-2010
One or more children
93
88
86
82
85
85
Voluntarily childless
2
5
8
10
6
8
Involuntarily childless
4
4
5
5
6
5
Temporarily childless
1
1
1
3
2
3
Total
100
100
100
100
100
100
Childless women
1973-1976
1982
1988
1995
2002
2006-2010
Voluntarily
31
53
55
59
44
49
Involuntarily
60
38
36
26
40
30
Temporarily
9
9
10
16
16
21
Total
100
100
100
100
100
100
Sources: Abma and Martinez (2006); Martinez et al (2012); Mosher and Bachrach (1982), author’s 249 calculations. 250 Note: Sub-categories of childless do not add up to total due to rounding. 251 252 Fig. 4 Per cent distribution of childless women aged 40-44 by childless status, National Survey of 253 Family Growth, United States 254
255 Sources: Abma and Martinez (2006); Martinez et al (2012); Mosher and Bachrach (1982), author’s 256 calculations 257 258 259 3.4 Personal Characteristics and Attitudes of Childless Women 260
There is ample evidence from several rounds of the NSFG that childless women, and particularly the 261 voluntarily childless, are disproportionately white, are employed full-time, and have a higher 262 education; and are less likely to be currently or formerly married and are less religious (Abma and 263 Martinez 2006). For example, data from the 2002 round show that among women aged 35-44, 69 per 264 cent of the voluntarily childless had some college or higher education, compared to 17 per cent among 265 all women of that age; 76 per cent of the voluntarily childless were working full-time, compared to 51 266 per cent among all women; 79 per cent were non-Hispanic white, compared to 71 per cent among all 267 women; and 35 per cent never attended religious services, compared to 17 per cent among all women 268 (Abma and Martinez 2006). 269 Among the women aged 35-44, the voluntarily childless also differed from the temporarily and 270 involuntarily childless in terms of economic characteristics. They had the highest individual and family 271 incomes, the most extensive past work experience, and were the most likely to be employed in 272 professional and managerial occupations. For example, according to the results of the 1995 round, 57 273
per cent of the voluntarily childless had individual annual earnings of over US$25,000, compared to 41 274 per cent of the temporarily childless and 36 per cent of the involuntarily childless; and 84 per cent had 275 worked more than 15 years, compared with 72 per cent of the temporarily childless and 77 per cent of 276 the involuntarily childless (Abma and Martinez 2006). 277 On the whole, the voluntarily childless tend to differ from women who have children and from the 278 temporarily or the involuntarily childless in terms of their attitudes regarding gender egalitarianism, 279 work, and family. For example, in their responses to questions in the 1995 round, 82 per cent of 280 voluntarily childless versus 72 per cent of women with children disagreed with the statement “a man 281 can make long-range plans, a woman cannot;” and 84 per cent of the voluntarily childless versus 75 per 282 cent of the women with children agreed with the statement “young girls are entitled to as much 283
independence as boys.” The voluntarily childless also stood out in their response to the question of 284
whether “women are happier if they stay at home and take care of their children;” 87 per cent of them 285 disagreed, compared with around 76 per cent of the women who had children or were temporarily or 286 involuntarily childless (Abma and Martinez 2006). 287 288 289 4 Reasons and Motivations for Remaining Childless 290
In a discussion of the biological factors which contribute to childbearing motivations, Foster (2000: 291
227) argued that because of their genetic predisposition to nurture and the effects of hormones, “most 292 women, motivated by a genetically developed desire to nurture, will choose to have at least one child, 293 given reasonably favorable circumstances.” Moreover, McQuillan et al. (2008: 17) established that 294 motherhood is valued by mothers and non-mothers alike, and that “there is no evidence that valuing 295 motherhood is in conflict with valuing work success among non-mothers, and among mothers the 296
association is positive.” Yet for prolonged periods a fifth of U.S. women, i.e. around 20 per cent, 297 remained childless. Why? 298 In the first place about five per cent of women cannot or should not bear children; they are 299 involuntarily childless, mostly due to fecundity impairments or health issues (Figure 4 and Table 2). 300 Then there are the temporarily childless, i.e. those that are still expecting to have a child. However, 301 these women can no longer be considered temporarily childless once they have reached the end of their 302 childbearing period. The remainder of women remains childless for a wide variety of reasons. 303 People grow up and live in differing social, cultural, and economic circumstances which influence 304 their decisions regarding childbearing. They live aided or obstructed by a material world, and are 305 affected by an array of social norms. They may also have their own independent reasons for not having 306 children. Both the material conditions and the norms affecting their decisions may change over time. If 307 we were to accept the notion that every woman has a natural desire to have children, irrespective of her 308 surroundings, there would not be any voluntary childlessness. Indeed, there was a time in U.S. history 309 when only around eight per cent of white women and only about five to six per cent of black women 310 were childless. Among these women, the rates of voluntary childlessness must have been negligible. 311 The 1973-1976 round of the NSFG found that only two per cent of women reported being voluntarily 312 childless, which implies that this share might have been even lower during the 1960s among white 313 women. Moreover, the five to six per cent rate of childlessness among black women leaves very little 314 room for voluntary childlessness. On the other hand, as was pointed out above, at certain points in time 315 around 20 per cent of white women and almost 30 per cent of black women were childless, which 316
implies that the shares of “voluntary” childlessness were large. 317 The basic explanation for these extreme high and low childlessness rates is the fact that the former 318 occurred at a time of economic hardship and psychological stress for large strata of the population 319 affecting family life during the Great Depression which started in 1929 and lasted through the early to 320 mid-1930s. Conversely, the low childlessness rates occurred when a majority of the population 321 experienced favorable economic and social conditions for childbearing after the Second World War. In 322 his recently published book, Labor’s Love lost: The Rise and Fall of the Working-Class Family in 323 America, Cherlin (2014) masterfully describes in great detail changes in American family life over the 324
past two centuries. He characterizes “the Great Depression [as] a cataclysmic event in the United States 325 in its depth and duration” (Cherlin 2014:60). Based on contemporary sociological research of 326 Komarovsky (1940), Cherlin discusses the effect of the Depression, inter alia, on reproductive 327 behavior. 328 329 Their sex lives often deteriorated: in twenty-two out of thirty-eight families for which adequate 330 information was collected, the frequency of sexual relations declined--including four families in which 331 sex stopped altogether. In some cases, however, couples reduced sexual activity not because of 332
emotional strain but in order to lower the chance that the wife would become pregnant. Without 333 modern means of birth control such as the pill or the IUD, financially struggling couples did what they 334
could to avoid having another mouth to feed. One parent said, “It is a crime for children to be born 335
when the parents haven’t got enough money to have them properly” (Cherlin 2014: 79). 336 337
The low shares of childlessness make clear sense in light of Cherlin’s characterization of the living 338 conditions of American families in the post-World War II years. 339 340 Why did young couples have so many children? One reason lay in the unique life histories of the 341 generation who were in their twenties and thirties. They experienced the Great Depression as children 342 or adolescents and then a world war erupted as they reached adulthood. After enduring these two 343
cataclysmic events, the “great generation,” as they are sometimes called, was pleased in peacetime to 344 turn inward toward home and family. ... Family life was the domain in which they found … security. 345 Raising children provided a sense of purpose to adults who had seen how fragile the social world 346
could be. … Moreover, conditions were favorable for family formation and fertility: unemployment 347 rates were low, wages were rising, and the government had enacted the GI Bill, which offered low-348 interest home mortgage loans to veterans so that they could buy single-family homes. ... Employers in 349 the rapidly expanding American economy were forced to offer higher wages in order to attract new 350 workers because they were in short supply (Cherlin 2014:115). 351 352 What remains to be clarified are the social, cultural, and economic circumstances shaping childlessness 353 levels and trends prior to the Great Depression of the 1930s and the levels and trends unfolding during 354 the two to three last decades of the 20th century, as well as the peak and subsequent decline in 355 childlessness in the early 21st century. 356 It could be considered odd that for 40 years (or 40 birth cohorts, i.e. 1867-1907) childlessness was 357 at a similar level as during the Great Depression (Figure 1). Morgan (1991) has argued that the period 358 of high childlessness in late 19th and early 20th centuries was mainly due to a strong motivation to delay 359 marriage and childbearing, which eventually resulted in many women remaining childless, even though 360 that was not their initial intention. Childbearing delays were significantly more pronounced in the 361
economically more advanced states of the northeast. Many young women working in mills “may have 362 been important income earners. Pressure for them to marry may have been replaced by pressure to 363
continue supporting the family” (Morgan 1991: 801). Furthermore, the harsh conditions of the 364 economic depression of the 1890s might have had an impact similar to that of the Great Depression of 365 the 1930s, even though it was not as long or as deep. In addition, the risk of remaining childless would 366 have been greater when childbearing was delayed, as sub-fecundity and sterility increases among 367 women in their thirties. Finally, growing numbers of women were entering professions during this 368 period, and these women tended not to marry; or, if they married, they often remained childless. 369 Turning our attention to the end of the 20th century and the early 21st century, numerous societal 370 developments have been taking place simultaneously, each of which has played a role in shaping 371 contemporary childbearing behavior, and has thus contributed to trends in childlessness. These include: 372 373
The re-emergence of marriage and childbearing postponement (Kohler et al. 2002; 374
Hašková 2007; Goldstein et al. 2009; Frejka 2011) 375
Rising female labor force participation rates, which are now almost as high as those of 376 men (Oppenheimer 1994; Bianchi 2011) 377
The work-family dilemma for employed women (Bianchi 2011) 378
The status of the childcare infrastructure (Laughlin 2013) 379
The increase in women’s earnings, and the growth in their income relative to that of men 380 (Cherlin 2014: 126; Wang et al. 2013) 381
The growing empowerment of women (Anonymous 2009) 382
High rates of incarceration (Tsai and Scommegna 2012) 383
The deployment of men and women in wars (Adams 2013) 384
Technological developments in production and communication, and their impact on the 385 composition of the work force (Karoly and Panis 2004; Economist Intelligence Unit 386 2014) 387
The hollowing out of the work force (Cherlin 2014: 124-125) 388
Changes in the class structure of society, with education playing the decisive role (Cherlin 389 2014) 390
Growing job insecurity, particularly among the less educated (Farber 2010) 391
Changing marriage and cohabitation patterns (Cherlin 2009) 392
Changing income and wealth distribution patterns (Saez and Zucman 2014) 393
Income stagnation among a large share of the population (Krugman 2007; Fry and 394 Kochhar 2014) 395 396 The above developments may influence women and their partners—in various ways, at different stages, 397 and to differing degrees—in their inadvertent or conscious deliberations about whether to remain 398 childless. 399 On the other hand there are those, including professionals such as psychologists and physicians, 400 who have argued that some women and men decide to remain childless for their own subjective 401 reasons. These individuals presumably engage in an independent decision-making process in which 402 they focus on their personal motivations and preferences, rather than allowing themselves to be 403 influenced by their circumstances. Scott (2009: 75-110; 222) reported the results of a survey of 404 childless individuals which found that the six most compelling motivation statements for not having 405 children were: 406 407
I love our life, our relationship, as it is, and having a child won't enhance it. 408
I value freedom and independence. 409
I do not want to take on the responsibility of raising a child. 410
I have no desire to have a child, no maternal/paternal instinct. 411
I want to accomplish/experience things in life that would be difficult to do if I was a 412 parent. 413
I want to focus my time and energy on my own interests, needs, or goals. 414 415 Taking into account the wide range of circumstances and personal subjective reasons which can affect 416
people’s decisions about whether to have children can help us to better understand the increase in the 417 share of women who remained childless which occurred during the final decades of the 20th century 418 and into the 21st century. However, the reasons for the apparent reversal in this trend in the early years 419 of the 21st century have yet to be explored. That is a topic for discussion and research in the near 420 future, especially if this trend continues. 421 422 423 5 Black childlessness: Trends and Explanations 424
For almost 60 birth cohorts (1883-1942) childlessness was higher among black than among white 425 women (Figure 1). At its peak black childlessness was 2.4 times higher than it was among white 426 women – in the 1924 and 1925 birth cohorts. Starting with the cohorts born in the early 1940s, this 427 trend was reversed, and black women became less likely than white women to be childless. Among the 428 youngest cohorts, those born in the late 1950s and the 1960s, the shares of black and of white 429 Americans who are childless are converging at around 12 to 15 per cent (Figures 1 and 2). The 430 relatively low childlessness among black women and the convergence with white childlessness since 431 the end of the 20th century is generally confirmed by data from the Fertility Supplements of the Current 432 Population Survey as well as the National Surveys of Family Growth. 433 The basic reasons for high black childlessness were analogous to those shaping white childlessness, 434 namely difficult economic and social settings, psychological stress and social norms. In addition, living 435 conditions of black Americans were incomparably more difficult than those of whites. Racial 436 segregation, discrimination, and inequalities have been basic features of American society throughout 437 its history (Massey 2011), and are reflected in virtually all aspects of life, such as economic 438 opportunities, remuneration, schooling, housing, and access to health and reproductive services. 439 Farley (1970: 217-226) was the first to analyze deteriorating health conditions of blacks 440 systematically, and their effect on reproductive behavior during the first three decades of the 20th 441 century. An increase in the prevalence of venereal diseases, such as syphilis and gonorrhea may have 442 been an important factor generating the fertility decline and the increase in childlessness among blacks, 443 which culminated in the 1930s. Farley was criticized by McFalls (1973: 18) and others who argued in 444
favor of “a more conservative interpretation of the importance of VD in the natality history of the black 445
population.” Yet McFalls (1973: 18) conceded that “health factors undoubtedly played a more 446
significant role” than other societal factors. 447 But what explains the decline in black childlessness and the crossover from relatively high to 448 relatively low levels of childlessness from the 1941 to the 1942 birth cohorts? The decline in the 449 childlessness rate of black women started with the cohorts most affected by the Great Depression, 450 namely those born around 1915, and lasted until the 1948 cohort, from a share of 30 per cent to six per 451 cent (Figures 1 and 2). The childlessness decline among blacks took more than twice as long as that for 452
white women, 33 compared to 14 cohorts. The childlessness descent for white women also started with 453 the cohorts most affected by the depression of the 1930s, but stopped when living conditions started to 454 improve significantly after the Second World War and essentially settled at that level for over 20 birth 455 cohorts. Among black women childlessness stopped declining temporarily for a few birth cohorts – 456 those born between 1926 and 1931 – but then resumed its decline with new force. Black childlessness 457 declined from 20 per cent in the 1931 cohort to six per cent in the 1948 cohort. 458 The passage of the Social Security Act in 1935 strengthened government support for health 459 activities (Farley 1970: 230-235). Title VI of that act appropriated money “for the purpose of assisting 460
States, counties…. in establishing and maintaining adequate public health service, including the 461
training of personnel for State and local health work...” This was an important element in the 462 development of the health system. The resulting improvements in the health of the black population in 463 turn led to declines in childlessness. 464 Moreover, there may be some justification to assume that improvements in living conditions and 465 educational attainment levels among the black population during the second half of the 20th century 466 were associated with the long-term decline in childlessness. This progress was both absolute as well as 467 relative to that of the white population. While living conditions for blacks remained inferior to those of 468 whites, the disparities were narrowing as blacks were catching up. On average, incomes of blacks were 469 rising faster than those of whites, especially during the 1990s (Figure 5). Rates of poverty among 470 blacks were also improving. Based on the definition of poverty of the U.S. Census Bureau, the ratio of 471 blacks to whites who were living in poverty declined from 3.4 in 1970 to 2.1 in 2010 (DeNavas-Walt et 472 al. 2012: Table B-1). In addition, educational attainment levels of blacks were increasing faster than 473 those of whites. Between 1960 and 2009, the shares of blacks aged 25 and older who had graduated 474 from high school rose from 20.1 to 84.1 percent, whereas the corresponding shares of whites increased 475 from 43.2 to 87.1 per cent (U.S. Census Bureau 2012: Table 225). Over the same period, the shares of 476 blacks aged 25 and older who had graduated from college grew from 3.1 to 19.3 per cent, while the 477 corresponding shares of whites increased from 8.1 to 29.9 per cent (U.S. Census Bureau 2012: Table 478 225). 479 480 481 Fig. 5 Households by total money income (in 1,000 of constant 2008 U.S. dollars) and race of 482 householder, black as percent of white income, 1967 to 2010 483
484 Source: DeNavas-Walt et al. (2012), Table A-2 485 486 What might be the reasons for the most recent turnaround -- the doubling in black childlessness from 487 six per cent in the 1948 birth cohort to 12 per cent in the 1968 cohort? The numerous societal 488 developments shaping childlessness that have been taking place around the turn of the century listed 489 above, together with the subjective motivations of women for not having children, surely played a role 490
in influencing contemporary childbearing behavior and thus contributed to the increase in childlessness 491 of black women. 492 Other important factors which might have influenced the recent rise in black childlessness are 493 changes in union formation and marital trends, and in fertility trends within unions. According to 494 Cherlin (2009: 169), “the larger story for African Americans is a sharp decline in marriage that is far 495
greater than among other groups.” In 2010 the share of black married women over age 18 was a mere 496 31 per cent compared to 61 per cent in 1960. In contrast, among white women this share declined from 497 74 to 55 per cent (Cohn et al. 2011). These developments are in line with the findings of Espenshade 498 (1985: 209), who concluded that “at least since 1960 in the United States, a weakening of marriage has 499 been under way. The fading centrality of marriage in the lives of American men and women is more 500
noticeable for blacks than for whites.” Only 24 per cent of black women aged 15-44 were married 501 compared with 46 per cent of white non-Hispanic women according to the NSFG 2006-2010 round 502 (Copen et al 2012: 12). 503 A comprehensive, albeit complex, set of explanations for declining marriage rates among blacks 504 has been revealed by research conducted by Banks (2011). Most black women want to marry and have 505 children, as getting married is seen as a marker of status and social prestige, and remains an aspiration. 506 Almost all black women would prefer to have a partner of the same race, as they are acculturated to 507 date and marry black men, and rarely marry across racial lines. In the African American community, 508 however, there is a considerable shortage of successful black men who are educated, employed, and 509 have respectful earnings. One reason for this shortfall is the extraordinarily high rate of incarceration of 510 black men (Massey 2011:10; Tsai and Scommegna 2012). Second, black men are up to three times 511 more likely than black women to marry a person of a different race. Third, at all educational levels 512
men’s attendance and attainment rates are far below those of women. In these circumstances, many 513 black women remain single or marry less educated black men. In such unions, women tend to be better 514 educated and earn more money than their spouse, which can result in tensions over gender roles. Such 515 marriages have a high potential to dissolve. Hence a high divorce rate among blacks is another reason 516 why their marriage rates are low. 517 Data on trends in the types of first unions for women aged 15 to 44 confirm the decline in 518 percentages of women who are married. Shares of marriages in first unions declined from 25.2 per cent 519 in 1995 to 12.5 per cent in the 2006-2010 round of the National Survey of Family Growth (Table 3). 520 Over the same period, the share of unions which were cohabitations increased from 35.4 to 49.2 per 521 cent. Consequently, the percentages of black women of reproductive age who were not in any union 522 hardly changed between the 1995 and the 2006-2010 NSFG rounds, i.e. instead of getting married a 523 large share of black women were living in a consensual union. That implies that the recent increase in 524 childlessness of black women does not appear to be associated with a decline in the percentage of 525 women who are in a union. The combined shares of cohabiting and married women were 60.6 and 61.7 526 per cent in 1995 and 2006-2010, respectively. 527 What did change dramatically between 1995 and 2013 was the fertility rate of unmarried black 528 women; it declined by 17 per cent, from 74.5 to 61.7 births per 1,000 unmarried black women (Table 529 4). It was this significant decline in the fertility rate which was associated with the rise in black 530 childlessness between the 1948 and the 1968 birth cohorts (Figure 2). It is worth noting that the fertility 531 rate of unmarried black women was almost twice the rate of unmarried white non-Hispanic women. 532 Nonetheless, the decline in black fertility, overall and especially of unmarried – cohabiting and never 533 married – women, was apparently the decisive factor in the recent rise of black childlessness. 534 535 536 Table 3 Type of first unions, women ages 15-44, United States 537
Year
No union
Cohabitation
Marriage
Total
1995
39.4
35.4
25.2
100
2006-10
38.4
49.2
12.5
100
Source: Copen et al.( 2013); Martin et al .( 2015) 538 539 540 Table 4 Births per 1,000 women ages 15-44, by race, United States 541
Year
All black women
Black unmarried women
White non-Hispanic
unmarried women
1995
71.0
74.5
28.1
2006
71.4
70.7
32.4
2013
64.7
61.7
31.7
Source: Copen et al. (2013); Martin et al. (2015) 542
543 544 545 6 Epilogue 546
More than ever in U.S. history, women and couples can regulate their fertility. They have access to a 547 wide variety of means to prevent childbearing, and there is over 20 years of experience with assisted 548 reproductive technologies (ART) which can alleviate the burden of infertility. A Division of 549 Reproductive Health at the Centers for Disease Control has a long history of surveillance and research 550
in women’s health and fertility, adolescent reproductive health, and safe motherhood. In response to a 551 congressional mandate, CDC has started to strengthen existing data collection efforts initiated by the 552 American Society for Reproductive Medicine (ASRM) and the Society for Assisted Reproductive 553 Technology (SART), and to develop a national system for monitoring ART use and outcomes. 554 The facts, i.e. the childlessness levels and trends since the late 19th century, are reasonably well 555 known. But often the mechanisms that shaped the facts have not been thoroughly deciphered, although 556 some of the basic circumstances affecting levels and trends of childlessness are quite obvious, namely 557 the concurrent economic and social conditions and cultural norms. 558 The U.S. population has experienced periods of very high and very low childlessness. The 559 challenging living conditions in the 1930s appear to have been the main cause of the high levels of 560 childlessness observed in that period. In contrast, the favorable living standards and enlightened public 561 policies of the 1940s, 1950s, and 1960s were instrumental in maintaining low levels of childlessness. 562 Living conditions of African Americans were far more difficult than those of white Americans; 563 hence higher black than white childlessness during much of the 20th century. Subsequently black 564 childlessness declined to levels below those of whites which in part was likely to have been due to 565 improvements in the health and living conditions of blacks, even though these conditions continued to 566 be inferior to those of whites. 567 In the recent past, i.e. since the 1970s through the late 1990s/early 2000s, the three independent 568 sources of data indicate that the overall childlessness rate doubled (Figures 1-4 and Tables 1 and 2). 569 This was the case among white as well as among black women, although not quite for identical birth 570 cohorts (Figure 2). 571 While history provides a general understanding of the principal causes of childlessness, the 572 experience of the past few decades points to the complexity inherent in identifying more specific 573 factors shaping levels and trends of childlessness. In section 4 above, 15 such societal factors discussed 574 in the literature are listed. In addition, people claim to have personal motivations and preferences for 575 not having children. Six of the most compelling ones were also listed above. What appears lacking is 576 an overall picture of the interaction of the elements which shape childlessness, and how these change 577 over time. 578 As of the early 2010s, around 12 to 16 per cent of U.S. white and black women over age 40 remained 579 childless. Among Hispanic women this share was lower, about 11 percent were childless. Whether 580 these percentages will increase or decline is impossible to predict. It depends on whether the material 581 world will be aiding or obstructing family formation, and how cultural norms and personal attitudes 582 will change over time. 583 584
Literature 585 Abma, J.C., Martinez, G.M. (2006). Childlessness among older women in the United States: Trends and profiles. Journal 586 of Marriage and Family, 68, 1045-1056. 587 Adams, C. (2013, March 14). Millions went to war in Iraq, Afghanistan, leaving many with lifelong scars. McClatchy 588 Newspapers, http://www.mcclatchydc.com/news/nation-world/national/article24746680.html. Accessed 27 July 2015 589 Anonymous (2009, December 30). Female power. The Economist, http://www.economist.com/node/15174418. Accessed 590 27 July 2015. 591 Bachu, A. (1995). Fertility of American Women: June 1994. U.S. Bureau of the Census. Current Population Reports, 592 Populations Characteristics (P20-482). Washington, DC: U.S. Government Printing Office. 593 Banks, R.R. (2011). Is Marriage for White People? How the African American Marriage Decline Affects Everyone. New 594 York: Dutton. 595 Bianchi, S.M. (2011). Changing families, changing workplaces. Future Child, 21, 15-36. 596 Cherlin, A.J. (2009). The marriage-go-round: The state of marriage and the family in America today. New York: Vintage 597 Books. 598 Cherlin, A.J. (2014). Labor’s love lost: The rise and fall of the working-class family in America. New York: The Russell 599 Sage Foundation. 600 Cohn, D’V., Passel, J.S., Wang, W., Livingston, G. (2011). Barely half of U.S. adults are married. A record low. Pew 601 Research Center. http://www.pewsocialtrends.org/2011/12/14/barely-half-of-u-s-adults-are-married-a-record-low/. 602 Accessed 27 July 2015 603 Copen, C.E., Daniels, K., Vespa, J., Mosher, W.D. (2012). First marriages in the United States: Data from 2006–2010 604 National Survey of Family Growth. National Health Statistics Reports, 49, 1-21. 605 Copen, C.E., Daniels, K., Mosher, W.D. (2013). First premarital cohabitation in the United States: 2006–2010 National 606 Survey of Family Growth. National Health Statistics Reports, 64, 1-15. 607 DeNavas-Walt, C., Proctor, B.D., Smith, J.C. (2012). Income, poverty, and health insurance coverage in the United 608 States: 2011 (P60-243). Washington, DC: U.S. Government Printing Office. 609 Espenshade, T.J. (1985). Marriage trends in America: Estimates, implications, and underlying causes. Population and 610 Development Review, 11, 193-246. 611 Economist Intelligence Unit (EIU) (2014). What’s next: Future global trends affecting your organization - Evolution of 612 work and the worker. SHRM Foundation. 613 Farber, H.S. (2010). Job loss and the decline in job security in the United States. In Abraham, K.G., Spletzer, J.R., Harper 614 M. (Eds.), Labor in the New Economy (pp. 223–262). Chicago: Chicago Press. 615 Farley, R (1970). Growth of the black population. Chicago: Markham Publishing Co. 616 Foster, C. (2000). The limits to low fertility: A biosocial approach. Population and Development Review, 26, 209-234. 617 Frejka, T. (2011). The role of contemporary childbearing postponement and recuperation in shaping period fertility 618 trends. Comparative Population Studies – Zeitschrift für Bevölkerungswissenschaft, 36, 927-958. 619 Fry, R., Kochhar, R. (2014). America’s wealth gap between middle-income and upper-income families is widest on 620 record. Pew Research Center. http://www.pewresearch.org/fact-tank/2014/12/17/wealth-gap-upper-middle-income/. 621 Accessed 27 July 2015. 622 Goldstein, J., Sobotka, T., Jasilioniene, A. (2009). The end of lowest-low fertility? Population and Development Review, 623 35, 663-700. 624 Hamilton, B.E., Cosgrove, C.M. (2010). Central birth rates, by live-birth order, current age, and race of women in each 625 cohort from 1911 through 1991: United States, 1960–2005. Hyattsville, MD: National Center for Health Statistics. 626 [2012 addition with 2006-2009 data] http://www.cdc.gov/nchs/nvss/cohort_fertility_tables.htm. Accessed 27 July 627 2015. 628
Hašková, H. (2007). Fertility decline, the postponement of childbearing and the increase in childlessness in Central and 629 Eastern Europe: A gender equity approach. In Crompton, R., Lewis, S., Lyonette, C. (Eds.), Women, men, work and 630 family in Europe (pp. 76-85). Basingstoke, Hampshire: Palgrave Macmillan. 631 Heuser, R.L. (1976). Fertility tables for birth cohorts by color: United States, 1917–73. Rockville, MD: U.S. Department 632 of Health, Education, and Welfare, National Center for Health Statistics. 633 http://www.cdc.gov/nchs/data/misc/fertiltbacc.pdf. Accessed 27 July 2015. 634 Karoly, L., Panis, C.W.A. (2004). The 21st century at work forces shaping the future workforce and workplace in the 635 United States. RAND Corporation, Santa Monica. 636 Kohler, H.-P., Billari, F.C., Ortega, J.A. (2002). The emergence of lowest-low fertility in Europe during the 1990s. 637 Population and Development Review, 28, 641-680. 638 Komarovsky. M. (1940). The unemployed man and his family. New York: Octagon Books. 639 Krugman, P. (2007). The conscience of a liberal. New York: W. W. Norton. 640 Laughlin, L. (2013). Who’s minding the kids? Child care arrangements: Spring 2011. Current Population Reports P70-641 135. U.S. Census Bureau, Washington, DC. 642 Martin, J.A., Hamilton, B.E., Osterman, M.J.K., Curtin, S.C., Mathews, T.J. (2015). Births: Final data for 2013. National 643 vital statistics reports. 64. Hyattsville, MD: National Center for Health Statistics. 644 Martinez, G., Daniels, K., Chandra, A. (2012). Fertility of men and women aged 15–44 years in the United States: 645 National survey of family growth, 2006–2010. National Health Statistics Report 51. Hyattsville, MD: National Center 646 for Health Statistics. 647 Massey, D.S. (2011). The past and future of American civil rights. Daedalus, 140, 33-54. 648 McFalls, J.A. (1973). Impact of VD on the fertility of the US black population, 1880–1950. Social Biology, 20, 2-19. 649 McQuillan, J., Greil, A.L., Scheffler, K.M., Tichenor, V. (2008). The importance of motherhood among women in the 650 contemporary United States. Lincoln: University of Nebraska. 651 Morgan, S.P. (1991). Late nineteenth- and early twentieth-century childlessness. American Journal of Sociology, 97, 779-652 807. 653 Mosher, W.D., Bachrach, C.A. (1982). Childlessness in the United States. Journal of Family Issues, 3, 517–543. 654
Oppenheimer, V.K. (1994). Women’s rising employment and the future of the family in industrial societies. Population 655 and Development Review, 20, 293-342. 656 Rindfuss, R.R., Morgan, S.P., Swicegood, G. (1988). First births in America: Changes in the timing of parenthood. 657 Berkeley: University of California Press. 658 Saez, E., Zucman, G. (2014). Wealth inequality in the United States since 1913: Evidence from capitalized income tax 659 data. Working Paper 20625. New York: National Bureau of Economic Research. 660 Scott, S.S. (2009). Two is enough: A couple’s guide to living childless by choice. Berkeley California: Seal Press. 661 Sobotka, T. (2009). Shifting parenthood to advanced reproductive ages: Trends, causes, and consequences. In Tremmel, 662 J. (Ed.), A young generation under pressure? (pp. 129-154). Berlin, Heidelberg: Springer. 663 Tsai, T., Scommenga, P. (2012). U.S. has world's highest incarceration rate. Washington, DC: Population Reference 664 Bureau. 665 US Bureau of the Census (1975). Historical statistics of the United States, colonial times to 1970, Washington, DC: 666 Bicentennial Edition. 667 US Census Bureau (2012). The 2012 statistical abstract. http://www.census.gov/compendia/statab/. Accessed 27 July 668 2015. 669 Wang, W., Parker, K., Taylor, P. (2013). Breadwinner moms: Mothers are the sole or primary provider in four-in-ten 670 households with children; public conflicted about the growing trend. Pew Research Center. 671 http://www.pewsocialtrends.org/2013/05/29/breadwinner-moms/. Accessed 27 July 2015. 672
