Original Research Article

Socioeconomic Status, Education, and Reproduction in Modern Women:An Evolutionary Perspective

SUSANNE HUBER,1* FRED L. BOOKSTEIN,2,3 AND MARTIN FIEDER2

1Research Institute of Wildlife Ecology, University of Veterinary Medicine Vienna, Austria2Department of Anthropology, University of Vienna, Austria3Department of Statistics, University of Washington, Seattle, Washington

ABSTRACT Although associations between status or resources and reproduction are positive in premodern soci-eties and also in men in modern societies, in modern women the associations are typically negative. We investigatedhow the association between socioeconomic status and reproductive output varies with the source of status and resour-ces, the woman’s education, and her age at reproductive onset (proxied by age at marriage). By using a large sample ofUS women, we examined the association between a woman’s reproductive output and her own and her husband’sincome and education. Education, income, and age at marriage are negatively associated with a woman’s number ofchildren and increase her chances of childlessness. Among the most highly educated two-thirds of the sample of women,husband’s income predicts the number of children. The association between a woman’s number of children and her hus-band’s income turns from positive to negative when her education and age at marriage is low (even though her meanoffspring number rises at the same time). The association between a woman’s own income and her number of childrenis negative, regardless of education. Rather than maximizing the offspring number, these modern women seem to adjustinvestment in children based on their family size and resource availability. Striving for resources seems to be part of amodern female reproductive strategy—but, owing to costs of resource acquisition, especially higher education, it maylead to lower birthrates: a possible evolutionary explanation of the demographic transition, and a complement to thehuman capital theory of net reproductive output. Am. J. Hum. Biol. 22:578–587, 2010. ' 2010 Wiley-Liss, Inc.

Life history theory predicts that individuals optimizethe allocation of their available resources to maximize fit-ness, so that resource availability influences resource allo-cation strategies. Because an individual’s energy andresources are finite, the allocation of these resourcesbetween competing life history goals involves trade-offs(Clutton-Brock, 1991; Daan and Tinbergen, 1997; Stearns,1992). A first trade-off is between current and futurereproduction (Gadgil and Bossert, 1970; Williams, 1966)and a second is between offspring number and offspringfitness (Lack, 1947, 1954; Smith and Fretwell, 1974; Wil-liams, 1966). An assumed trade-off between offspringquantity and quality is central to both evolutionary andeconomic explanations of so-called demographic transi-tions (Becker, 1981; Easterlin, 1987; Harpending et al.,1990; Kaplan, 1996; Lancaster, 1997; Mace, 1998; Rogers,1990; Turke, 1989), which are characterized by a declinein offspring number despite plentiful resources (Bor-gerhoff Mulder, 1988; Hill, 1993; Low, 2000). Some studiesprovide evidence for the existence of such a quantity-qual-ity trade-off (Borgerhoff Mulder, 2000; Gillespie et al.,2008; Penn and Smith, 2007; Strassmann and Gillespie,2002), whereas other studies do not (Goodman and Koupil,2009; Kaplan et al., 1995; Mueller, 2001; Stevenson et al.,2004).In most of Europe and North America and in New Zea-

land and Australia, demographic transitions are notedbetween 1880 and 1920 as well as during the second halfof the 20th century (Caldwell, 1982; Van de Kaa, 1997). Indeveloping countries, a change in fertility occurred as wellalbeit in the more recent of these windows (Lee, 2003).Even in ancient Rome, upper classes seemed to limitreproduction (Brunt, 1971; Dixon, 1992; Parkin, 1992). Alaw instituted by the Emperor Augustus rewarded coupleshaving children and penalized those without (Boone and

Kessler, 1999). Inasmuch as the global demographic tran-sition results in population aging and associated economicand political challenges (Lee, 2003), its causes are widelydiscussed among demographers, economists, sociologists,and evolutionary anthropologists. The lowering of fertilitymay be driven by increasing competition (BorgerhoffMulder, 1998; Low et al., 2002; Voland, 1998), economicgains, and costs of marriage and having children (Becker,1981) [which may be phrased as high costs and benefits ofparental investment (Kaplan et al., 2002; Mace, 2007,2008)], women’s education (Becker et al., 1990; Low et al.,2002; Robinson, 1997), changing payoffs of human capitalinvestment (Kaplan et al., 2002), breakdown of extendedkinship networks (Turke, 1989), cultural evolution (Boydand Richerson, 1985; Newson et al., 2005), or social strati-fication (Harpending and Rogers, 1990).Reproductive success entails success in child rearing

through the moment that the children can produce off-spring of their own. Because the variance of fertility infemales is fairly low, survival of children to reproductiveage is likely the more important determinant of a woman’sreproductive success (Strassmann and Guillespie, 2002).One way to increase chances of successful offspring sur-vival is to strive for resources (Turke, 1989; Wiederman,1993). On this theory, women are expected to adjust theirreproductive output according to their resource availabil-ity and to prefer mates who offer access to more resources

*Correspondence to: Susanne Huber, Research Institute of WildlifeEcology, Savoyenstrasse 1, A-1160 Vienna, Austria.E-mail: susanne.huber@vetmeduni.ac.at

Received 29 June 2009; Revision received 18 January 2010; Accepted 31January 2010

DOI 10.1002/ajhb.21048

Published online 14 April 2010 in Wiley Online Library (wileyonlinelibrary.com).

AMERICAN JOURNAL OF HUMAN BIOLOGY 22:578–587 (2010)

VVC 2010 Wiley-Liss, Inc.

(Buss, 1989, 1999). Correspondingly, data from hunter-gatherers, rural and agricultural societies (BorgerhoffMulder, 1988; Chagnon, 1988; Cronk, 1991; Hill andHutardo, 1996; Mealey, 1985; Voland, 1990), and likewisemodern societies (Fieder et al., 2005; Fieder and Huber,2007; Hopcroft, 2006; Nettle and Pollet, 2008; Weedenet al., 2006) show that the association between wealth orsocioeconomic status and reproduction in men is positive.(The principal reason, according to Fieder and Huber(2007) and Nettle and Pollet (2008), is the increased prob-ability of childlessness in men who have low income.)

Yet in modern women, the relationship between socioe-conomic status and number of children is negative (Fiederet al., 2005; Fieder and Huber, 2007; Goodman and Kou-pil, 2009; Hopcroft, 2006; Nettle and Pollet, 2008; Weedenet al., 2006). This likely results from the incompatibility ofeducation, work, and rearing of offspring for women(Morgan, 2003; Marini, 1984). Educational attainment,for instance, delays the age at first birth (Low et al., 2002;Kaplan et al. 2002; Rindfuss et al., 1996). Nettle and Pol-let (2008) argue that women may compensate for thistrade-off between education, income generation, and chil-dren by seeking mates who have high income. In line withthat argument, wealthy men are more likely to marrythan poorer men (Kanazawa, 2003; Nakosteen andZimmer, 1997; Oppenheimer, 2003). Even women withhigh-paying jobs tend to value the mate’s resources highly(Buss, 1989; Wiederman and Allgeier, 1992).

This study examines how the association between socio-economic status and number of children in women variesdepending on the source of status and resources as theyarise in either of two channels, from the woman or fromher husband. In a sample of half a million contemporaryUS women, we examine the association between her life-time fertility (or its obverse, the probability of childless-ness) and both her own and her husband’s resource avail-ability (income) and educational attainment. We specifi-cally analyze how the relationship between resources andnumber of children varies by woman’s education and byage at marriage, adjusting for the confounding factors ofage and ethnicity. We predict that a woman’s income andeducational attainment will be negatively associated withher number of children. Her husband’s income, on theother hand, should be positively associated with her num-ber of children (and thus negatively with childlessness).Husband’s education could have both positive and nega-tive effects on a woman’s reproductive output—it shouldincrease his income but delay reproductive onset—and wehave no prediction about the balance of these two contrarytendencies.

METHODS

We used the 5% US census sample from 1980 providedby IPUMS-USA (Minnesota Population Center. IntegratedPublic Use Microdata Series—International: Version 4.0.Minneapolis, MN: University of Minnesota, 2008). De-scriptive statistics of the census are given in Tables A1and A2. Total fertility rate in the United States is rela-tively high for so developed a country (Hagewen and Mor-gan, 2005), and fertility patterns are more segmented byfemale education than in other Western societies (Les-thaege and Moors, 2000). We limited our analyses towomen in their first marriage, and assigned a husbandonly if he lived within the same household as the woman

proband. (Thus, husbands living separately are omittedfrom our sample.) We expect that most of the husbands inour sample are the fathers of the women’s offspring. Inplace of a measure of age at reproductive onset, which wedo not have, we use the woman’s age at marriage (which,according to Low et al. (2002), is also associated with totalnumber of children albeit to a lesser degree than age atreproductive onset). We limited our analysis to spousalpairs both aged between 45 and 66 years. All but 0.5% ofAmerican women have finished reproduction by the age of45 years (Ventura et al., 2003). The final file, a sample of504,496 married couples, contains the following variables:age, number of biological children, total income per year,highest grade completed, and the age at marriage. Income(in $US) is reported for each partner in the marriage astotal pretax personal income or losses from all sources forthe year 1979. This variable was truncated at the source,with any value greater than $75,000 reported as $75,000.Education is according to the following categories: 1—none or preschool; 2—grade 1, 2, 3, 4; 3—grade 5, 6, 7, 8;4—grade 9; 5—grade 10; 6—grade 11; 7—grade 12; 8—1to 3 years of college; 9—4 years of college or more. Age ofmarriage is reported by exact year except that all agesolder than 24 years are pooled in a ‘‘251’’ category. De-scriptive statistics of the sample and bivariate correla-tions among the variables are given in Tables A3–A6.The distributions of income and number of children are

skewed. We performed a square-root transformation ofboth parameters before the analyses. To avoid leading ze-ros in our tables of regression coefficients, the incomeshad already been expressed in thousands of US dollarsbefore this additional transformation. Because incometypically rises with age, we adjusted for current age. Anal-yses of the woman’s probability of childlessness were logis-tic regressions on her husband’s income, education, andcurrent age and the woman’s age at marriage. Analyses ofnumbers of offspring were linear regressions on the wom-an’s and her husband’s income, education, and currentage and the woman’s age at marriage either (i) for allwomen of our sample or (ii) sorted by female educationalcategory. We additionally fitted the woman’s offspringcount to a linear mixed model of a woman’s and her hus-band’s income, education, and current age, along with thewoman’s age at marriage (i) for all women of our sample,and (ii) sorted by female educational category (see Appen-dix). In these analyses, the woman’s and her husband’sethnicity are entered as random factors. In addition, wecalculated the linear model separately for White andBlack marital pairs (owing to small sample size otherraces were not analyzed separately). Finally, we calcu-lated age-adjusted regression slopes for woman’s incomeand husband’s income against number of children for eachsubset of woman’s educational level by age at marriage (atotal of 90 subsets). We largely abstain from reportingnull-hypothesis statistical significance levels owing to thevery large sample size (Ziliak and McCloskey, 2008) andthe unreasonableness of the null hypothesis itself. In allreports of fitted models, quantities in parentheses afterregression coefficients are their standard errors.

RESULTS

A woman’s probability of remaining childless decreaseswith increasing husband’s income (Fig. 1a). The logistic

579STATUS, EDUCATION, AND REPRODUCTION IN AMERICAN WOMEN

American Journal of Human Biology

regression equation of the woman’s childlessness (encodedas 1 5 childless and 0 5 at least one child) on the woman’sand her husband’s income (as transformed), education,and age, and the woman’s age at marriage has coefficientsfor logit (childlessness) as follows: incomewoman b 5 0.09(0.004), incomehusband b 5 20.079 (0.004), educationwoman

b 5 20.016 (0.004), educationhusband b 5 0.006 (0.003),age at marriagewoman b 5 0.118 (0.001), agewoman b 50.009 (0.002), agehusband b 5 0.019 (0.002), Nagelkerke R2

5 0.093. The association between childlessness andincome holds true even when untransformed income datawere used in the analysis (data not shown). Increasingown income (Fig. 1b), age and age at marriage all raise awoman’s chances of remaining childless, whereas increas-ing woman’s education and increasing husband’s incomedecrease them.A woman’s [transformed] number of children is posi-

tively associated with her husband’s [transformed] incomebut negatively associated with her own [transformed]income, with her education, and with her husband’seducation: offspringwoman � 3.134 (0.012) 2 0.005 (0.0003)agewoman 2 0.027 (0.001) incomewoman 2 0.01 (0.001) educa-tionwoman 2 0.032 (0.0002) age at marriagewoman 2 0.007(0.0003) agehusband 1 0.005 (0.001) incomehusband 2 0.009(0.001) educationhusband, R2 5 0.078. Correspondingly,the mean number of children declines with increasingwoman’s educational category (Table 1). The older thewoman at marriage and the higher her educational level,the lower her number of children (Fig. 2). The prediction ofa woman’s number of children by her income, her education,and her husband’s income and education remains essentiallyunchanged in a linear mixed model adjusting for ethnicity,age, and age at marriage (Table 2). The same holds true if

Fig. 1. Percentage of childless women (a) per husband’s incomedeciles, and (b) per women’s income deciles.

TABLE1.Mea

noffspringcountandlinea

rregressionof

thewom

an’stransformed

number

ofch

ildrenon

age,ageatmarriage,transformed

income,transformed

husband’sincome,

andhusband’sed

uca

tion

,sorted

bylevel

ofwom

an’sed

uca

tion

Education

al

categorywoman

N(%

oftotal)

No.

children,

Mea

n(SD)

R2

Incomewoman

B(SE)

Age w

oman

B(SE)

Incomehusband

B(SE)

Ageatmarriagewoman

B(SE)

Education

husband

B(SE)

Age h

usband

B(SE)

12476(0.5)

4.42(2.90)

0.124

20.067(0.013)

20.007(0.004)

20.039(0.011)

20.035(0.002)

20.046(0.007)

20.004(0.004)

27283(1.4)

4.14(2.78)

0.103

20.061(0.007)

20.009(0.003)

20.030(0.007)

20.033(0.001)

20.028(0.005)

20.006(0.003)

358246(11.5)

3.39(2.33)

0.099

20.036(0.002)

20.007(0.001)

20.020(0.002)

20.035(0.001)

20.027(0.002)

20.008(0.001)

423036(4.6)

3.21(2.11)

0.088

20.025(0.003)

20.007(0.001)

20.016(0.003)

20.034(0.001)

20.016(0.002)

20.009(0.001)

533220(6.6)

3.13(2.00)

0.085

20.021(0.003)

20.009(0.001)

20.015(0.003)

20.033(0.001)

20.011(0.002)

20.007(0.001)

633593(6.7)

2.99(1.90)

0.079

20.032(0.002)

20.008(0.001)

20.011

(0.003)

20.031(0.001)

20.017(0.002)

20.009(0.001)

7226704(44.9)

2.91(1.78)

0.071

20.031(0.001)

20.004(0.000)

10.006(0.001)

20.033(0.000)

20.005(0.001)

20.008(0.0004)

868529(13.6)

2.93(1.74)

0.066

20.022(0.001)

20.004(0.001)

10.012(0.001)

20.032(0.001)

20.002(0.002)

20.006(0.001)

951409(10.2)

2.76(1.64)

0.077

20.028(0.001)

10.002(0.001)

10.019(0.002)

20.036(0.001)

10.011(0.002)

20.007(0.001)

Note:

Thez-statistic

accordingto

whichon

ewou

ldtest

thesignificance

oftheseregressioncoefficien

tsis

theratioof

theBto

itsstandard

error;

theserangeupto

abou

t31forwom

an’sow

nincomeandupto

abou

t9for

husb

and’sincomein

thedata

here.

580 S. HUBER ET AL.

American Journal of Human Biology

only White couples are included in the analysis (n 5 467,041;Table A7), whereas the relationship is the reverse if only Blackcouples are used (n5 26,817; Table A8). The interaction maybe because of lower sample size and higher average offspringcount. Mean 6 SD offspring count in Black couples is 3.66 62.76 and in Whites 2.97 6 1.83. Again, results are essentiallyunchanged using untransformed income data (data not shown).

Within each of the lower six categories of women’s edu-cation, number of children is negatively associated withhusband’s income, but the association is positive withinthe upper three education categories, which total 68.7% ofthe sample. These associations between a woman’s num-ber of children and her husband’s income within educationcategories remain essentially unchanged in a linear mixedmodel adjusting for ethnicity, age, and age at marriage(Table A9). Yet within each of the same nine categories ofwoman’s education, a woman’s number of children isalways negatively associated with her own income, andalso—except for the highest category—with her husband’seducation. The age-adjusted regression coefficientsbetween the number of children and husband’s incomebecome increasingly less negative and ultimately positivewith increasing woman’s educational level and age atmarriage (Fig. 3a). If we sort these categories by meannumber of children, the regression coefficient for numberof children on income turns out negative for the subsetswith the higher average number of children but increas-ingly less negative and then more positive for the subsetswith lower average numbers of children (Fig. 4a). Theassociation between a woman’s number of children andher own income, calculated separately by education-by-age-at-marriage subset, likewise becomes less negativewith increasing educational level, but usually remainsnegative (Fig. 3b), as is the case if we sort the categoriesby mean number of children (Fig. 4b).

Resource availability in terms of the income of eitherspouse is positively associated with woman’s educationallevel. However, the highest regression coefficient is foundbetween the two education levels. Coefficients in theregression of educationwoman are as follows: incomewoman

b, 0.181 (0.001); incomehusband b, 0.133 (0.001); educa-tionhusband b, 0.450 (0.001); age at marriagewoman b, 0.019(0.0005); agewoman b, 20.009 (0.001); agehusband b, 20.001(0.001) (square root of income in 1000$). In addition, the

percentage of husbands within the highest income quarterrises sharply with female education from 5% in the lowesteducation category through 15% in the sixth female edu-cation category to 49% in the highest female educationcategory.

DISCUSSION

In this large sample of contemporary US women,resource availability in terms of husband’s income, but notin terms of the woman’s own income, was usually posi-tively associated with reproductive output. The probabilityof remaining childless decreased and mean number of chil-dren increased with increasing husband’s income. A wom-an’s own income, on the other hand, was negatively associ-ated with her number of children and increased the chan-ces of remaining childless. For women, higher income istypically associated with postponing of reproduction byreason of longer education (see Low et al., 2002) or work-force participation, whereas the rearing of offspring bywomen usually hinders full professionalism with its higherwage levels. (But it is not clear whether women fit workingpatterns to family obligations or vice versa: see, forinstance, Brewster and Rindfuss, 2000.) Not only does edu-cational attainment delay reproduction but also early ageat first birth may interfere with further education (Marini,1984). Postponement of reproduction in turn is a majorcause of low fertility (Morgan, 2003). Because the negativeassociation between number of children and woman’sincome is stronger than its positive association with hus-band’s income, the overall income of the married couple isnegatively associated with number of children (data notshown).Our findings are in keeping with other analyses in con-

temporary America. Higher husband income predicted

TABLE 2. Linear mixed model using a woman’s [transformed] numberof children as dependent variable, her own [transformed] income in1000$, education, age and age at marriage, as well as her husband’s[transformed] income in 1000$, education and age as fixed factors,

and her own and her husband’s ethnicity as random factors

Estimate SE

Intercept 3.184 0.053Educationwoman (reference: 9)

1 0.163 0.0142 0.144 0.0083 0.022 0.0054 20.015 0.0055 20.020 0.0056 20.042 0.0057 20.021 0.0038 0.010 0.004

Educationhusband (reference: 9)1 0.174 0.0142 0.134 0.0073 0.016 0.0044 20.021 0.0055 20.032 0.0046 20.061 0.0047 20.044 0.0038 20.034 0.003

Incomehusband 0.004 0.0006Incomewoman 20.030 0.0006Age at marriagewoman 20.034 0.0002Agewoman 20.004 0.0003Agehusband 20.008 0.0003Ethnicitywoman

a 0.013 0.008Ethnicityhusband

a 0.005 0.003

aEstimates of covariance parameters.

Fig. 2. Mean number of children of women calculated separately bywoman’s educational level and age at marriage. [Color figure can beviewed in the online issue, which is available at wileyonlinelibrary.com.]

581STATUS, EDUCATION, AND REPRODUCTION IN AMERICAN WOMEN

American Journal of Human Biology

higher fertility in wives of Harvard-educated men (Weedenet al., 2006). Analogously, Hungarian women marryinghigher status mates have increased numbers of survivingchildren (Bereczkei and Csanaky, 1996), whereas in urbanSouth India, an area where the demographic transitiondid not start until the second half of the 20th century, theassociation between husband’s income and a woman’s fer-tility is negative (Shenk, 2009). A woman’s own educationand income, on the other hand, are negatively associatedwith her offspring count (Fieder et al., 2005; Fieder andHuber, 2007; Goodman and Koupil, 2009; Hopcroft, 2006;Nettle and Pollet, 2008; Shenk, 2009; Weeden et al., 2006)and increase her risk of remaining childless (Dobritz,2003; Hewlett, 2002; Kaplan et al., 2002; Nettle and Pollet,2008; Neyer and Hoem, 2008).Across categories of decreasing education and age at

marriage, which are also categories of increasing averagefamily size, the association between number of childrenand either the woman’s or her husband’s income becomesmore and more negative, possibly because some of the sur-plus resources are invested in the human capital (Becker,1993) of existing offspring rather than in producing moreoffspring. Only if family size was small was the associationpositive between a woman’s number of children and herspouse’s income. This may be a reason that the relation-ship between husband’s income and woman’s offspring

number was negative in Black couples: on average, Blackcouples had more children than White couples. Thus, ouranalysis suggests that rather than maximizing number ofchildren, modern American women adjust offspringnumber and investment based on resource availabilitytogether with educational level. Such a strategy wouldbe reasonable. In answer to high reproductive invest-ment cost, one should produce as many offspring as willeventually be recruited to the reproducing population,but no more. Penn and Smith (2007) and Clark (2007)suggest that premodern or transitional women, when-ever given the opportunity, prefer smaller families toreduce their net costs of reproduction. Models by Mace(1996, 1998) suggest that maximum fertility very rarelymeans maximum fitness for women, and Turke (1989)argues that for women, the number of children leadingto maximal reproductive success should be lower thanfor men. In addition, intermediate levels of offspringproduction are favored by women in an East Africancommunity (Borgerhoff Mulder, 2000).Our study has certain limitations. We have no data on

actual paternity, although as these are all first marriagesit is likely that most of the children counted are childrenof the husband of the female proband. In addition, we can-not date the actual onset of reproduction (hence the proxywe used, age at marriage). In addition, the reported valueof income is capped at $75,000, a ceiling that is certainlyexceeded in any representative US population in 1980.Again in keeping with the public capitalism of the UnitedStates, income is ‘‘income from all sources.’’ These disad-vantages, it seems to us, are outweighed by the large sizeof the sample, which permits us to examine regressioncoefficients within each of 90 age-education subsets aver-aging 5630 women each.We limited our analyses to married individuals to allow

for the concomitant analysis of the association of a wom-an’s number of children and her own and her husband’sincome. Because marital status is associated with theprobability of childlessness (Fieder and Huber, 2007),childless women are thus underrepresented in our sample(7.4% vs. 14.1% for our age group in the overall US cen-sus) and the average offspring count is correspondinglyhigher. The incompatibility of family and career is likelyto be higher in unmarried women than in married women,and thus the association between a woman’s income andher chances of remaining childless is likely even strongerthan in the logistic regressions reported here. Amongunmarried women, we would also expect a stronger nega-tive relationship between a woman’s own income and edu-cation and her number of children. The associationbetween a woman’s fertility (whether measured as child-lessness or instead as number of children) and her part-ner’s income and education, on the other hand, shoulddepend on the extent to which a woman can dispose of herpartner’s resources.In Whites, the association between woman’s number of

children and husband’s education was zero or negative de-spite the positive association between number of childrenand husband’s income. This underlines the importance ofa man’s resources rather than his education per se for amodern American woman’s reproductive output. On theother hand, our data are also consistent with the claimthat women select their spouses on the basis of educationrather than income—we clearly have a very high degreeof assortative mating for education in these data. Higher

Fig. 3. (a) Association between a woman’s number of children andher husband’s income (adjusted for husband’s age) calculated for eachcombination of woman’s educational level and age at marriage (hence,for each of 90 combinations). (b) Association between a woman’s num-ber of children and her own income (adjusted for age) calculated foreach combination of woman’s educational level and age at marriage.[Color figure can be viewed in the online issue, which is available atwileyonlinelibrary.com.]

582 S. HUBER ET AL.

American Journal of Human Biology

education certainly increases a woman’s opportunity tomarry a man of both high education and high income.Among the highly educated women, husband’s incomewas positively associated with number of children. Yet thestriving for higher education as a woman’s strategy toincrease her own resources (income) has negative effectson reproductive output. Two paths, of opposite sign, leadfrom woman’s education to number of children, and itseems that the one of negative effect dominates.

Assortative mating for education has both advantagesand disadvantages. On the one hand, highly educatedwomen are likely to benefit from their husband’s resour-ces, and husbands benefit economically from having morehighly educated wives (Becker, 1973). On the other hand,the tendency of highly educated women to marry men ofthe same or higher levels of education limits available

marriage partners and possibly lowers marriage rates(Coppola, 2004). Kaplan et al. (2002) further argue thatgreater selectivity in obtaining a suitable partner is amajor reason for the postponement of reproduction inmore educated women, who may thereafter, ‘‘by accident,’’as it were, fall short of their intended fertility.Thus, the increased resource availability at which these

American women were striving through higher educationis associated with lower number of children, for two rea-sons: postponement of reproduction and higher workforceparticipation. In addition, women in modern societies typ-ically lack help from close kin, reducing the available timeand resources for reproduction (Turke, 1989). In addition,higher socioeconomic status leads to increasingly highercosts of rearing offspring—wealth increases the cost ofadditional children (Lawson and Mace, 2009), costs notnecessarily commensurate with the additional resourcesaccompanying that higher status. In line with Becker(1981), Mace (1998, 2008) interprets worldwide fertilitydecline in terms of the relative costs of child rearing.Extraordinary levels of investment may be necessary ifchildren are themselves to be competitive in the marriagemarket (Becker, 1981; Kaplan et al., 1995). This is anotherpossible mechanism for the demographic transition (Bor-gerhoff Mulder, 1998): that the cost of children rises fasterthan the overall standard of living.We conclude that rather than maximizing number of

children, modern American women may be adjusting off-spring number and investment based on resource avail-ability and on the marginal cost of offspring at higherresource levels. Although the striving for resourcesseems to be part of modern female reproductive strategyand although resources reduce the chances of remainingchildless, the same resources ultimately lead to a declinein birth rate as a consequence of higher education, theprincipal mechanism by which those resources areacquired in contemporary America. Higher educationdelays reproduction, increasing the risk that women donot reach their intended family size. It also may dispro-portionately raise the cost of investment per child. Inthis way, evolutionarily acquired traits of human matingstrategies together with the standard human capitaltheory of investment in offspring combine to explain thelow fertility patterns of a range of modern societies, notmerely a set of regressions from one recent US censussubsample.

ACKNOWLEDGMENTS

Data source: US census sample from 1980 provided byIPUMS-USA (Minnesota Population Center. IntegratedPublic Use Microdata Series—International: Version 4.0.Minneapolis, MN: University of Minnesota, 2008).

LITERATURE CITED

Becker GS, Murphy KM, Tamura R. 1990. Human capital, fertility, andeconomic growth. J Polit Econ 98:S12–S37.

Becker GS. 1973. A theory of marriage: Part I. J Polit Econ 1973; 81:813–846.

Becker GS. 1981. A treatise on the family. Cambridge, MA: HarvardUniversity Press, 320 p.

Becker GS. 1993. Human capital. Chicago, IL: University Chicago Press.412 p.

Bereczkei T, Csanaky A. 1996. Mate choice, marital success, and reproduc-tion in a modern society. Ethol Sociobiol 17:17–35.

Fig. 4. (a) Association between subsample mean number of chil-dren and the regression coefficient for the woman’s number of chil-dren on her husband’s income (adjusted for husband’s age) calculatedfor each combination of woman’s educational level and age at mar-riage. The curve is a best-fitting unweighted quadratic regression. (b)Association between subsample mean number of children and theregression coefficient for the woman’s number of children on her ownincome (adjusted for age) calculated for each combination of woman’seducational level and age at marriage.

583STATUS, EDUCATION, AND REPRODUCTION IN AMERICAN WOMEN

American Journal of Human Biology

Boone JL, Kessler KL. 1999. More status or more children? Social status,fertility reduction, and long-term fitness. Evol Hum Behav 20:257–277.

Borgerhoff Mulder M. 1988. Reproductive success in three Kipsigiscohorts. In: Clutton-Brock TH, editor. Reproductive success. Chicago,IL: University Chicago Press. p 419–435.

Borgerhoff Mulder M. 1998. The demographic transition: are we any closerto an evolutionary explanation? TREE 13:266–270.

Borgerhoff Mulder M. 2000. Optimizing offspring: the quantity-qualitytradeoff in agropastoral Kipsigis. Evol Hum Behav 21:391–410.

Boyd R, Richerson PJ. 1985. Culture and evolutionary process. Chicago,IL: University Press. 340 p.

Brewster KL, Rindfuss RR. 2000. Fertility and women’s employment inindustrial nations. Annu Rev Sociol 26:271–296.

Brunt PA. 1971. Italian manpower 225 BC-AD 14. Oxford, England: Clar-endon Press. 784 p.

Buss DM. 1989. Sex differences in human mate preferences: evolutionaryhypotheses tested in 37 cultures. Behav Brain Sci 12:1–49.

Buss DM. 1999. Evolutionary psychology. The new science of the mind.Boston, MA: Allyn and Bacon. 456 p.

Caldwell JC. 1982. Theory of fertility decline. London: Academic Press.386 p.

Chagnon NA. 1988. Life history, blood revenge and warfare in a tribal popula-tion. Science 239:985–992.

Clark G. 2007. A farewell to alms. Princeton, Princeton University Press. 432 p.Clutton-Brock TH. 1991. The evolution of parental care. Princeton, NJ:

Princeton University Press. 372 p.Coppola L. 2004. Education and union formation as simultaneous process

in Italy and Spain. Eur J Pop 20:219–250.Cronk L. 1991. Low socioeconomic status and female biased parental

investment: the Mukogodo example. Am Anthropol 91:414–429.Daan S, Tinbergen JM. 1997. Adaptation of life histories. In: Krebs JR,

Davies NB, editors. Behavioral ecology: an evolutionary approach.Oxford, England: Blackwell Science. p 311–333.

Dixon S. 1992. The Roman family. Baltimore, MD: John Hopkins Univer-sity Press. 279 p.

Dobritz J. 2003. Polarisierung versus Vielfalt—Lebensformen und Kinder-losigkeit in Deutschland—eine Auswertung des Mikrozensus. Z Bevol-kerungswiss 28:403–421.

Easterlin RA. 1987. Birth and fortune: the impact of numbers on personalwelfare. Chicago, IL: University of Chicago Press. 228 p.

Fieder M, Huber S, Bookstein FL, Iber K, Schafer K, Winckler G, WallnerB. 2005. Status and reproduction in humans: new evidence for the valid-ity of evolutionary explanations on basis of a university sample. Ethol-ogy 111:940–950.

Fieder M, Huber S. 2007. The effects of sex and childlessness on the associ-ation between status and reproductive output in modern society. EvolHum Behav 28:392–398.

Gadgil M, Bossert WH. 1970. Life historical consequences of natural selec-tion. Am Nat 104:1–24.

Gillespie DOS, Russell AF, Lummaa V. 2008. When fecundity does notequal fitness: evidence of an offspring quantity versus quality trade-offin pre-industrial humans. Proc R Soc Lond B 275:713–722.

Goodman A, Koupil I. 2009. Social and biological determinants of repro-ductive success in Swedish males and females born 1915–1929. EvolHum Behav 30:329–341.

Hagewen KJ, Morgan SP. 2005. Intended and ideal family size in theUnited States, 1970–2002. Pop Develop Rev 31:507–527.

Harpending H, Rogers A. 1990. Fitness in stratified societies. Ethol Socio-biol 11:497–509.

Harpending H, Draper P, Pennington R. 1990. Cultural evolution, parentalcare, and mortality. In: Swedlund A, Armelagos G, editors. Health and dis-ease in traditional societies. South Hadley: Bergin & Garvey. p 241–255.

Hewlett SA. 2002. Creating a life. Professional women and the quest forchildren. New York, NY: Talk Miramax Books. 352 p.

Hill K, Hurtado AM. 1996. Ache life history: the ecology and demographyof a foraging people. New York, NY: Aldine de Gruyter. 561 p.

Hill K. 1993. Life history theory and evolutionary anthropology. EvolAnthropol 22:78–88.

Hopcroft RL. 2006. Sex, status, and reproductive success in the contempo-rary United States. Evol Hum Behav 27:104–120.

Kanazawa S. 2003. Can evolutionary psychology explain reproductivebehavior in the contemporary United States? Soc Quart 44:291–302.

Kaplan H. 1996. A theory of fertility and parental investment in tradi-tional and modern human societies. Yearb Phys Anthropol 39:91–135.

Kaplan HS, Lancaster JB, Johnson SE, Bock JA. 1995. Does observedfertility maximize fitness among NewMexican men? Hum Nat 6:325–360.

Kaplan H, Lancaster JB, Tucker WT, Anderson KG. 2002. Evolutionaryapproach to below replacement fertility. Am J Hum Biol 14:233–256.

Lack D. 1947. The significance of clutch size. Ibis 89:302–352.Lack D. 1954. The natural regulation of animal numbers. Oxford, UK:

Clarendon Press.

Lancaster JB. 1997. The evolutionary history of human parental invest-ment in relation to population growth and social stratification. In:Gowaty PA, editor. Feminism and evolutionary biology. New York, NY:Chapman & Hall. p 466–488.

Lawson DW, Mace R. 2009. Trade-offs in modern parenting: a longitudinalstudy of sibling competition for parental care. Evol Hum Behav 30:170–183.

Lee R. 2003. The demographic transition: three centuries of fundamentalchange. J Econ Perspect 17:167–190.

Lesthaege R, Moors G. 2000. Recent trends in fertility and household for-mation in the industrialized world. Rev Popul Soc Policy 9:121–170.

Low BS, Simon CP, Anderson KG. 2002. An evolutionary ecological per-spective on demographic transitions: modelling multiple currencies. AmJ Hum Biol 14:149–167.

Low BS. 2000. Sex, wealth, and fertility: old rules, new environments. In:Cronk L, Chagnon N, Irons W, editors. Adaptation and human behavior:an anthropological perspective. Hawthorne, NY: Aldine de Gruyter. p323–344.

Mace R. 1996. When to have another baby: a dynamic model of reproduc-tive decision-making and evidence from Gabbra pastoralitsts. EtholSociobiol 17:263–274.

Mace R. 1998. The coevolution of human fertility and wealth inheritancestrategies. Phil Trans Roy Soc Lond B 353:389–397.

Mace R. 2007. The evolutionary ecology of human family size. In: DunbarRIM, Barrett L, editors. The Oxford handbook of evolutionary psychol-ogy. Oxford, UK: Oxford University Press. p 383–396.

Mace R. 2008. Reproducing in cities. Science 319:764–766.Marini MM. 1984. Women’s educational attainment and the timing of

entry into parenthood. Am Sociol Rev 49:491–511.Mealey L. 1985. The relationship between social status and biological success:

a case study of the Mormon religious hierarchy. Ethol Sociobiol 6:249–257.Morgan SP. 2003. Is low fertility a twenty-first-century demographic cri-

sis? Demography 40:589–603.Mueller U. 2001. Is there a stabilizing selection around average fertility in

modern human populations? Popul Dev Rev 27:469–498.Nakosteen RA, Zimmer MA. 1997. Men, money, and marriage: are high

earners more prone than low earners to marry? Soc Sci Quart 78:66–82.Nettle D, Pollet TV. 2008. Natural selection on male wealth in humans.

Am Nat 172:658–666.Newson L, Postmes T, Lea SEG, Webley P. 2005. Why are modern families

small? Toward an evolutionary and cultural explanation for the demo-graphic transition. Pers Soc Psychol Rev 9:360–375.

Neyer G, Hoem JM. 2008. Education and permanent childlessness: Austriavs. Sweden. A research note. In: Surkyn J, Deboosere P, Van Bavel,editors. Brussels: VUB Press. p 91–114.

Oppenheimer VK. 2003. Cohabiting and marriage during young men’s ca-reer-development process. Demography 40:127–149.

Parkin TG. 1992. Demography and Roman society. Baltimore, MD: JohnHopkins University Press. 248 p.

Penn DJ, Smith KR. 2007. Differential fitness costs of reproductionbetween the sexes. Proc Natl Acad Sci U S A 104:553–558.

Rindfuss RR, Morgan SP, Offutt K. 1996. Education and the changing agepattern of American fertility: 1963–1989. Demography 33:277–290.

Robinson WC. 1997. The economic theory of fertility over three decades.Popul Stud 51:63–74.

Rogers AR. 1990. Evolutionary economics of human reproduction. EtholSociobiol 11:479–495.

Shenk MK. 2009. Testing three evolutionary models of the demographictransition: patterns of fertility and age at marriage in urban SouthIndia. Am J Hum Biol 21:501–511.

Smith CC, Fretwell SD. 1974. The optimal balance between size and num-ber of offspring. Am Nat 108:499–506.

Stearns SC. 1992. The evolution of life histories. Oxford, UK: Oxford Uni-versity Press. 249 p.

Stevenson JC, Everson PM, Grimes M. 2004. Reproductive measures, fit-ness, and migrating Mennonites: an evolutionary analysis. Hum Biol76:667–687.

Strassmann BI, Gillespie B. 2002. Life-history theory, fertility and repro-ductive success in humans. Proc R Soc Lond B 269:553–562.

Turke PW. 1989. Evolution and the demand for children. Popul Dev Rev15:61–90.

Van de Kaa DJ. 1997. Options and sequences: Europe’s demographic pat-terns. J Popul Res 14:1–29.

Ventura SJ, Hamilton SE, Sutton PD. 2003. Revised birth and fertilityrates for the United States, 2000 and 2001. National Vital StatisticsReports 51:1–6.

Voland E. 1990. Differential reproductive success in the Krummhorn popu-lation (Germany 18th and 19th century). Behav Ecol Sociobiol 26:65–72.

Voland E. 1998. Evolutionary ecology of human reproduction. Annu RevAnthropol 27:347–374.

Weeden J, Abrams MJ, Green MC, Sabini J. 2006. Do high-status peoplereally have fewer children? Hum Nat 17:377–392.

584 S. HUBER ET AL.

American Journal of Human Biology

Wiederman MW, Allgeier ER. 1992. Gender differences in mate selection crite-ria: sociobiological or socioeconomic explanation? Ethol Sociobiol 13:115–124.

Wiederman MW. 1993. Evolved gender differences in mate preferences:evidence from personal advertisements. Ethol Sociobiol 14:331–352.

Williams GC. Natural selection, the costs of reproduction, and a refine-ment of Lack’s principle. Am Nat 1966; 100:687–690.

Ziliak ST, McCloskey DN. 2008. The cult of statistical significance: how thestandard error costs us jobs, justice, and lives. Ann Arbor, MI: Univer-sity of Michigan Press. 321 p.

APPENDIX

TABLE A1. Descriptive statistics of overall US census 1980, limited toindividuals aged 46 to 65 years

Mean SD n

WomenIncome in 1000$ 5.90 7.10 1,166,288No. childrena 2.82 2.18 1,113,157Age 54.90 5.35 1,113,157Age at marriageb 22.06 5.82 1,109,180

MenIncome in 1000$ 18.93 14.77 1,044,806Age 54.73 5.34 1,001,928Age at marriageb 24.93 6.12 985,721

aNo data available for men.b4.8 % of the women and 5.7% of the men were never married.

TABLE A2. Frequency of educational categories in overall US census1980, limited to individuals aged 46 to 65 years

Womenn (%)

Menn (%)

1 None or preschool 9791 (0.9) 9254 (0.9)2 Grade 1, 2, 3, or 4 26,155 (2.3) 33,639 (3.4)3 Grade 5, 6, 7, or 8 170,313 (15.3) 176,264 (17.6)4 Grade 9 61,123 (5.5) 53,240 (5.3)5 Grade 10 81,473 (7.3) 65,613 (6.5)6 Grade 11 77,963 (7.0) 58,072 (5.8)7 Grade 12 444,946 (40.0) 312,061 (31.1)8 1–3 years of college 139,611 (12.5) 124,865 (12.5)9 41 years of college 101,782 (9.1) 168,920 (16.9)

Total 1,113,157 (100.0) 1,001,928 (100.0)

TABLE A3. Descriptive statistics of the sample

Mean SD n

Whole sampleIncomewoman in 1000$ 4.76 6.79 503,059Incomehusband in 1000$ 21.16 15.06 502,260No. childrenwoman 3.01 1.91 504,496Agewoman 53.57 4.79 504,496Agehusband 55.83 4.82 504,496Age at marriagewoman 21.82 4.67 504,496

Educationwoman 5 1Incomewoman in 1000$ 2.68 4.43 2476Incomehusband in 1000$ 10.29 9.24 2474Number of childrenwoman 4.42 2.90 2476Agewoman 53.68 4.80 2476Agehusband 56.18 4.82 2476Age at marriagewoman 22.16 6.70 2476

Educationwoman 5 2Incomewoman in 1000$ 2.52 4.27 7277Incomehusband in 1000$ 10.87 9.79 7266Number of childrenwoman 4.14 2.78 7283Agewoman 53.90 4.83 7283Agehusband 56.43 4.85 7283Age at marriagewoman 22.00 6.72 7283

Educationwoman 5 3Incomewoman in 1000$ 2.82 4.70 58136Incomehusband in 1000$ 13.55 10.07 58037Number of childrenwoman 3.39 2.33 58246

TABLE A3. (Continued)

Mean SD n

Agewoman 54.54 4.87 58246Agehusband 57.02 4.73 58246Age at marriagewoman 21.28 5.57 58246

Educationwoman 5 4Incomewoman in 1000$ 3.12 5.04 22990Incomehusband in 1000$ 15.51 10.39 22958Number of childrenwoman 3.21 2.11 23036Agewoman 53.83 4.82 23036Agehusband 56.45 4.72 23036Age at marriagewoman 20.71 5.09 23036

Educationwoman 5 5Incomewoman in 1000$ 3.28 5.08 33147Incomehusband in 1000$ 16.65 11.03 33108Number of childrenwoman 3.13 2.00 33220Agewoman 53.69 4.75 33220Agehusband 56.28 4.71 33220Age at marriagewoman 20.67 4.82 33220

Educationwoman 5 6Incomewoman in 1000$ 3.84 5.68 33508Incomehusband in 1000$ 17.54 11.72 33439Number of childrenwoman 2.99 1.90 33593Agewoman 53.91 4.64 33593Agehusband 56.37 4.62 33593Age at marriagewoman 20.71 4.61 33593

Educationwoman 5 7Incomewoman in 1000$ 4.54 6.19 226036Incomehusband in 1000$ 21.31 13.59 225654Number of childrenwoman 2.91 1.78 226704Agewoman 53.37 4.75 226704Agehusband 55.57 4.81 226704Age at marriagewoman 21.75 4.30 226704

Educationwoman 5 8Incomewoman in 1000$ 5.72 7.55 68259Incomehusband in 1000$ 26.89 17.54 68169Number of childrenwoman 2.93 1.74 68529Agewoman 53.39 4.77 68529Agehusband 55.45 4.81 68529Age at marriagewoman 22.50 4.17 68529

Educationwoman 5 9Incomewoman in 1000$ 9.36 9.97 51230Incomehusband in 1000$ 31.33 19.52 51155Number of childrenwoman 2.76 1.64 51409Agewoman 53.18 4.83 51409Agehusband 55.07 4.87 51409Age at marriagewoman 23.74 4.23 51409

(Continued)

TABLE A4. Frequency of husband’s educational categories

N (%)

1 None or preschool 2773 (0.5)2 Grade 1, 2, 3, or 4 13,058 (2.6)3 Grade 5, 6, 7, or 8 82,931 (16.4)4 Grade 9 25,570 (5.1)5 Grade 10 32,606 (6.5)6 Grade 11 30,319 (6.0)7 Grade 12 163,109 (32.3)8 1–3 years of college 62,671 (12.4)9 41 years of college 91,459 (18.1)

Total 504,496 (100.0)

TABLE A5. Frequency of ethnicities

Women, n (%) Husbands, n (%)

White 468,245 (92.8) 469,023 (93.0)Black/Negro 26,974 (5.3) 27,129 (5.4)American Indian or Alaska Native 1472 (0.3) 1433 (0.3)Chinese 1756 (0.3) 1793 (0.4)Japanese 2952 (0.6) 2250 (0.4)Other Asian or Pacific Islander 2360 (0.5) 2148 (0.4)Other race, nec 737 (0.1) 720 (0.1)Total 504,496 (100.0) 504,496 (100.0)

585STATUS, EDUCATION, AND REPRODUCTION IN AMERICAN WOMEN

American Journal of Human Biology

TABLE A6. Pearson correlation coefficients

Agehusband Incomehusband Incomewoman Educationwoman Educationhusband

Age atmarriagewoman

Agewoman 0.780 20.089 20.055 20.074 20.070 0.214Agehusband 20.131 20.041 20.113 20.126 0.004Incomehusband 0.011 0.324 0.410 0.048Incomewoman 0.213 0.123 0.041Educationwoman 0.603 0.127Educationhusband 0.137

TABLE A7. Linear mixed model including only White couples(n 5 467041) and using a woman’s [transformed] number of children

as dependent variable, her own [transformed] income in 1000$,education, age and age at marriage, as well as her husband’s

[transformed] income in 1000$, education and age as fixed factors

Estimate SE

Intercept 3.015 0.013Educationwoman (reference: 9)1 0.211 0.0162 0.165 0.0093 0.005 0.0054 20.037 0.0065 20.042 0.0056 20.066 0.0057 20.034 0.0048 0.001 0.004

Educationhusband (reference: 9)1 0.206 0.0152 0.130 0.0073 0.013 0.0044 20.019 0.0055 20.030 0.0046 20.061 0.0057 20.042 0.0038 20.033 0.003

Incomehusband 0.005 0.0006Incomewoman 20.029 0.0006Age at marriagewoman 20.034 0.0002Agewoman 20.003 0.0003Agehusband 20.008 0.0003

TABLE A8. Linear mixed model including only Black couples(n 5 26,817) and using a woman’s [transformed] number of children

as dependent variable, her own [transformed] income in 1000$,education, age and age at marriage, as well as her husband’s

[transformed] income in 1000$, education and age as fixed factors

Estimate SE

Intercept 3.697 0.072Educationwoman (reference: 9)1 0.026 0.0622 0.152 0.0353 0.238 0.0254 0.220 0.0285 0.242 0.0276 0.250 0.0277 0.145 0.0238 0.116 0.025

Educationhusband (reference: 9)1 0.167 0.0482 0.176 0.0293 0.101 0.0254 0.029 0.0305 0.007 0.0296 0.034 0.0297 0.008 0.0258 20.003 0.027

Incomehusband 20.015 0.004Incomewoman 20.040 0.004Age at marriagewoman 20.034 0.0008Agewoman 20.020 0.002Agehusband 20.005 0.001

TABLE A9. Linear mixed model sorted for female education withwoman’s [transformed] number of children as dependent variable,her own [transformed] income in 1000$, age, and age at marriage,as well as her husband’s [transformed] income in 1000$, educationand age as fixed factors, and her own and her husband’s ethnicity as

random factors

Estimate SE

Educationwoman 5 1Intercept 3.055 0.256Educationhusband (reference: 9)1 0.318 0.1162 0.388 0.1203 0.312 0.1194 0.067 0.1555 0.062 0.1586 0.131 0.1627 0.048 0.1298 20.015 0.140

Incomehusband 20.035 0.011Incomewoman 20.058 0.013Age at marriagewoman 20.033 0.002Agewoman 20.008 0.004Agehusband 20.003 0.004Ethnicitywoman

a

Ethnicityhusbandb 0.055 0.045

Educationwoman 5 2Intercept 3.420 0.154Educationhusband (reference: 9)1 0.354 0.0672 0.201 0.0593 0.118 0.0594 0.055 0.0755 0.046 0.0786 0.049 0.0857 0.107 0.0688 0.096 0.079

Incomehusband 20.030 0.007Incomewoman 20.061 0.007Age at marriagewoman 20.033 0.001Agewoman 20.007 0.003Agehusband 20.006 0.003Ethnicitywoman

b 0.016 0.022Ethnicityhusband

b 0.011 0.021Educationwoman 5 3Intercept 3.418 0.082Educationhusband (reference: 9)1 0.275 0.0362 0.195 0.0253 0.068 0.0244 0.012 0.0255 0.0007 0.0266 20.028 0.0277 20.005 0.0248 0.004 0.029

Incomehusband 20.012 0.002Incomewoman 20.037 0.002Age at marriagewoman 20.036 0.0006Agewoman 20.007 0.0009Agehusband 20.007 0.0009Ethnicitywoman

b 0.021 0.014Ethnicityhusband

b 0.010 0.007Educationwoman 5 4Intercept 3.518 0.103

(Continued)

586 S. HUBER ET AL.

American Journal of Human Biology

TABLE A9. (Continued)

Estimate SE

Educationhusband (reference: 9)1 0.245 0.0822 0.131 0.0393 0.034 0.0334 20.003 0.0345 20.007 0.0346 20.021 0.0357 20.017 0.0338 20.025 0.038

Incomehusband 20.008 0.003Incomewoman 20.026 0.003Age at marriagewoman 20.036 0.0009Agewoman 20.005 0.001Agehusband 20.010 0.001Ethnicitywoman

b 0.030 0.021Ethnicityhusband

b 0.010 0.010Educationwoman 5 5

Intercept 3.477 0.078Educationhusband (reference: 9)1 20.021 0.0722 0.046 0.0313 20.011 0.0224 20.045 0.0245 20.057 0.0236 20.064 0.0247 20.054 0.0228 20.036 0.025

Incomehusband 20.009 0.003Incomewoman 20.022 0.003Age at marriagewoman 20.035 0.0008Agewoman 20.007 0.001Agehusband 20.007 0.001Ethnicitywoman

b 0.009 0.008Ethnicityhusband

b 0.010 0.008Educationwoman 5 6

Intercept 3.392 0.091Educationhusband (reference: 9)1 0.202 0.0832 0.134 0.0293 0.038 0.0184 0.030 0.0205 20.030 0.0196 20.064 0.0187 20.006 0.0188 20.025 0.020

Incomehusband 20.005 0.003Incomewoman 20.032 0.002Age at marriagewoman 20.033 0.0008Agewoman 20.006 0.001Agehusband 20.009 0.001Ethnicitywoman

b 0.007 0.007Ethnicityhusband

b 0.028 0.021Educationwoman 5 7

Intercept 3.142 0.060Educationhusband (reference: 9)1 20.114 0.0452 0.055 0.0173 0.012 0.0064 20.012 0.0075 20.012 0.0066 20.037 0.0077 20.041 0.0048 20.021 0.005

Incomehusband 0.006 0.0009Incomewoman 20.031 0.0008Age at marriagewoman 20.033 0.0003Agewoman 20.003 0.0005Agehusband 20.009 0.0004Ethnicitywoman

b 0.015 0.009Ethnicityhusband

b 0.007 0.005Educationwoman 5 8

Intercept 2.967 0.052Educationhusband (reference: 9)1 0.118 0.0762 0.068 0.038

TABLE A9. (Continued)

Estimate SE

3 0.022 0.0124 0.006 0.0185 20.011 0.0146 20.065 0.0137 20.022 0.0068 20.041 0.006

Incomehusband 0.011 0.001Incomewoman 20.021 0.001Age at marriagewoman 20.033 0.0006Agewoman 20.004 0.0008Agehusband 20.007 0.0008Ethnicitywoman

b 0.011 0.007Ethnicityhusband

a

Educationwoman 5 9Intercept 2.649 0.058Educationhusband (reference: 9)1 20.006 0.1022 0.086 0.0473 20.020 0.0184 20.096 0.0325 20.039 0.0266 20.046 0.0227 20.040 0.0088 20.023 0.008

Incomehusband 0.017 0.002Incomewoman 20.026 0.001Age at marriagewoman 20.036 0.0006Agewoman 0.003 0.0009Agehusband 20.007 0.0009Ethnicitywoman

b 0.002 0.004Ethnicityhusband

b 0.012 0.009

aCovariance parameter is redundant.bEstimates of covariance parameters.

(Continued)

587STATUS, EDUCATION, AND REPRODUCTION IN AMERICAN WOMEN

American Journal of Human Biology