Marital Name Changing Attitudes and Plans of College Students: Comparing Change Over Time and Across Regions

Abstract

This study examines time period and regional effects on U.S. college students’ attitudes and plans regarding marital naming. Data were gathered at a Midwestern college in 1990 and 2006 and at an Eastern university in 2006 (N=867). No time period effect was identified for marital name plans in the Midwest samples. Women were neither more nor less likely to plan to retain their birth name in 1990 as compared to 2006. A time period effect was found for attitudes: Midwest women in 2006 were more likely to say a woman was more committed to the marriage if she took her husband's last name as compared to Midwest women in 1990. This indicates that women in the Midwest may have become more conservative over time. We also found regional differences: women in the East were significantly more likely than women in the Midwest to plan to keep their birth surname upon marriage. Findings suggest a trend toward more conservative attitudes over time and location although plans, perhaps due to the rareness of maintaining a birth surname upon marriage, have not changed.

Introduction

In the U.S., women have made considerable strides toward equity in the workforce, housework, childcare, and other attitudes regarding expectations of men and women's behaviors (England 2010). Yet, there still remains one practice that is resistant to change: women's last name selection at the time of marriage (Goldin and Shim 2004). The practice of married women taking their husband's last name originates from the patriarchal family system under which women were considered their husband's property (Foss and Edson 1989; Suarez 1996). Although women were not always legally compelled to take their husband's surname, several states had laws in place until the mid to late 1970s that denied women access to rights such as registering to vote if they did not use their husband's surname (Goldin and Shim 2004). Such legal requirements have now been overturned, in part due to much feminist activism on the issue (Lillian 2009), but the social norm that married women change to their husband's surname still strongly persists. Indeed, studies consistently show that most married women in the U.S.—90% or more—change to their husband's last name (Brightman 1994; Gooding and Kreider 2010; Johnson and Scheuble 1995).

Marital name selection continues to be an important feminist issue because it is part of a broader constellation of ideologies about gender and family life and is especially useful as a lens for analyzing relationship commitment and tension between family identity and women's autonomy (Nugent 2010). Recent research suggests that marital naming may actually be a better indicator of gender attitudes than other commonly used survey items that tend to focus on men and women having ‘separate spheres’ of responsibility for work versus home life (Hamilton et al. 2011). Some research suggests that non-conventional marital naming has become more common over time (Goldin and Shim 2004), but recently journalists in the popular press have speculated that there has been a movement in the last decade back to more conventional naming choices (Dubecki 2007; Hampson 2007; Nolan 1998). There is however, little empirical evidence of such a change.

Of the empirical research that does exist, there have been mixed results (Goldin and Shim 2004; Kerns 2011; Scheuble et al. 2000). Some of this is likely due to data availability as data on trends in women's marital surnames are generally unavailable, based on non-representative samples, or difficult to obtain. Studies have also differed in whether they ask about future plans for marital naming or analyzed actual practices based on surveys or content analyses of records such as newspaper wedding announcements. Additionally, very few empirical studies have focused on the influence of region on marital name choice. Much of the research has been done in the Eastern United States, which may differ in terms of the social climate for gender and family attitudes (Goldin and Shim 2004; Hoffnung 2006). Studies focusing on the percentage of women retaining their birth surname have vast differences in findings and one explanation may be the region of the country where the data were gathered.

In this study, we use three different samples to compare marital naming plans and attitudes toward marital name choice among students at a Midwestern college in 1990 and 2006 and among students at an Eastern university in 2006. Although generally not representative, previous studies show that college-educated women are more likely to make nontraditional marital naming choices than their counterparts (Gooding and Kreider 2010; Johnson and Scheuble 1995; Scheuble et al. 2000); hence, if there is a trend moving back toward more traditional choices, it is likely to be more visible in the plans of college students.

Although we only have data from the East at one time point and, therefore, cannot test regional differences over time, we can test hypotheses about recent regional differences and evaluate if students in these two regions have similar plans and attitudes. We use regression analysis to examine four main questions: 1) Is there change between 1990 and 2006 in male and female students’ attitudes about marital naming? 2) Is there change between 1990 and 2006 in female students’ plans for marital naming? 3) Is there a regional difference (Midwestern versus Eastern) in male and female students’ attitudes about marital naming in 2006? 4) Is there a regional difference in female students’ plans about marital naming in 2006?

Practices and Attitudes Regarding Marital Naming

How common is keeping one's birth name at the time of marriage? In U.S. national studies, 90 to 95% of married women change to their husband's last name (Dubecki 2007; Johnson and Scheuble 1995; Lillian 2009). Studies in the U.S. and Canada using less representative samples (e.g., New York Times wedding announcements, Harvard graduates, listserv surveys) that tend to over-represent women in higher socioeconomic statuses (SES) have found higher rates of women retaining their birth surname upon marriage: in these studies, 13 to 35% of women make this nonconventional marital name choice (Goldin and Shim 2004; Hoffnung 2006; Lillian 2009; Scheuble et al. 2000).

Trends Over Time

Trends over time show inconsistent results, likely because most studies have relied on non-representative samples of higher SES women. In a content analysis of 2,383 wedding announcements in the New York Times from 1966 to 1996, Scheuble et al. (2000) found that women were significantly more likely to have chosen nonconventional last name choices (i.e., kept their birth surname or hyphenated) in the late 1980s and 1990s than in the 1960s and 1970s. In 1966, about 3% of women made a nonconventional last name choice, and by 1996, this had increased to 29% of women. Goldin and Shim (2004) reported similar findings in their content analysis of wedding announcements in the New York Times from 1980 to 2001. The percentage of women who kept their surname rose from 20% in 1980 to 35% in 2001

Hoffnung (2006), who also examined New York Times wedding announcements, conducted a content analysis of 480 wedding announcements in the years 1982, 1987, 1992, 1997 and 2002. While the descriptive findings showed that in more recent years, women were slightly more likely to keep their birth surnames at the time of marriage (30.8% in 1987 compared to 34.2% in 2002), these fluctuations by year were not statistically significant, leading to the conclusion that “for educated women, nontraditional naming has neither become less or more common since the 1980s” (p. 824).Hoffnung (2006) also examined last name choice in a sample of 50 women (25 white women and 25 women of color) from five New England colleges and universities. These women were first studied in 1992–1993 and the naming questions were included in the 2003 survey. Of these women, 28.6% kept their birth name.

A longitudinal study by Goldin and Shim (2004) using reunion book data compared marital naming choices of female Harvard graduates of the classes of 1980 and 1990 ten years after finishing the undergraduate degree. Forty-four percent of women graduating in 1980 retained their birth surname upon marriage as compared to 32% of the women graduating in 1990. In the same study, the authors also examined Massachusetts birth records in 1990, 1995 and 2000, focusing on married couples who had both completed at least a 4 year degree and had their first child. In 1990, 21% of women had kept their birth surname, but only 13% did so in 2000. These different data sources—Harvard graduates and Massachusetts birth records—indicate a recent decline in women keeping their birth surnames at marriage, but the data is also limited to the Eastern region of the U.S.

Lillian (2009) examined marital naming plans and practices in data from an on-line survey sent to multiple listservs. Respondents (n=2641) were primarily from the U.S. (75%), but 10% were from Canada and 15% were residents of other countries. Twenty-four percent of her respondents did not change their birth surnames when they married. Canadian women were more likely than women from the United States to keep their birth surname when they married. Lillian also found that younger women were less likely to keep their birth name; she concluded that there was a slowdown in women's birth name retention and possibly a reversal in this trend.

Overall, these studies present inconsistent findings about trends in women's last name choice at the time of marriage. The New York Times, Harvard, and Massachusetts and listserv samples contain predominantly higher SES women, making it difficult to generalize the findings to the population. Additionally, the definition used to identify non-conventional marital names also differs among the studies (Goldin and Shim 2004; Scheuble et al. 2000), as some include women hyphenating both surnames, and others include only women keeping the birth surname alone.

Based on the Northeastern (Harvard and Massachusetts) findings (Goldin and Shim 2004), we would expect that the percentage of unmarried college women selecting nonconventional marital names should have decreased from 1990 to 2006. This is not consistent, however, with The New York Times data, which found a continued increase in nonconventional naming choices through 2000. The studies that have examined trends in marital naming have focused, to a large extent, on well educated women and women in the East. Because our data come from college students at all points in their undergraduate career—including some who will drop out of college—these women are not as educated as women in other trend studies. Less than one-fifth of college women in 1990 (Scheuble and Johnson 1993) planned to keep their birth surname, which indicates that this behavior is clearly not normative. There is also strong social pressure for women to follow conventional practices and change their last names to that of the husband. One goal of this research is to determine if women are more or less likely to plan to keep their birth name at marriage over a 16 year period. This represents a test of whether marital name choice has followed suit with other changes in gender attitudes over the last few decades or if marital name choice remains overwhelmingly conservative—as one of the last normative gendered expectation within U. S. society.

Practices Versus Attitudes

Practices may be indicative of social norms, but do they adequately reflect attitudes? There is evidence that social expectations for women in relation to work and family life have changed substantially over the last several decades and become much more progressive (Bolzendahl and Myers 2004; Brewster and Padavic 2000). Brooks and Bolzendahl (2004) found that, over time, there have been higher levels of support for women's rights and less agreement with survey items that have a more restrictive view of gender roles. Why have these changes in norms not been reflected in an increase in the percentage of women who retain their birth surname upon marriage?

One potential explanation of this issue arose in a study by Nugent (2010): she conducted a content analysis of 600 internet posts on the topic of children's surnames. Although women sometimes had preferences to keep their birth surnames (and give these names to their children) partners, relatives, and others “enforce[d] the cultural mandates of single shared surnames” (Nugent 2010, p. 158). As such “cultural reality” thwarted women's “egalitarian agendas” because women are often held accountable to different standards than men when making sacrifices for home and family life. This justifies viewing attitudes as separate from plans or practices for marital naming and suggests that women's attitudes about marital naming may be more progressive than their actual plans. This is consistent with the theory of cultural lag that some elements of culture change far more slowly than others and that norms may change more rapidly than actual behaviors (Ogburn 1964). While women are accepting of other woman retaining their birth surname when they marry (Scheuble and Johnson 1993) they do not plan to do so themselves.

Regional Differences

One major factor that has been under-examined in marital naming research is regional differences. There are no published studies that have systematically examined regional differences in naming attitudes and plans. One study of actual naming practices found that women from the South were more likely than women in other regions to maintain their birth name as a middle name when they marry (Johnson and Scheuble 1995), but no other regional differences or regional effects seem to influence marital surname choice.

Given the lack of research directly on regional differences in marital naming, we suggest that it is useful to turn to research on gender and family attitudes, because marital naming should fit within these broader ideologies (Hamilton et al. 2011). Research on gender attitudes finds that people living in the South tend to support more traditional gender roles than men and women in other parts of the United States (Carter and Borch 2005; Powers et al. 2003; Rice and Coates 1995), with little difference noted among other regions. Rice and Coates (1995), in an examination of gender roles by region, found that Southerners, Midwesterners and those in the West are all more conservative than those in the Eastern U.S. about women working if her husband can support her. Also, those in the South are slightly more conservative than those in the Northeast when asked whether a woman should help support a husband in his career or have a career of her own.

Studies of regional differences in traditional family values, such as views on same-sex marriage, divorce and single parenting, find some basis for differences in the Midwest and East, with Midwestern states more likely to support more conservative and traditional family practices than in the East (Cahn and Carbone 2010). Moore and Ovadia (2006) find that living in an area with a larger proportion of college graduates increases levels of tolerance—measured as treating people with different viewpoints fairly. In contrast, living in areas with a large proportion of Evangelical Protestants decreases tolerance.

Overall, the Eastern portion of the United States has a higher percent of people completing a Bachelor's Degree (Chrissey 2009) than does the Midwest. The East also has lower levels of Evangelicalism and more religious pluralism than does the Midwest (Silk 2005; Smith et al. 2002). Although there is no direct evidence of differences in marital naming plans and practices between these two regions, broader cultural differences in the two regions suggest that we should find more support and acceptance of nontraditional naming plans in the East than in the Midwest.

Additional Influences on Marital Naming Attitudes and Practices

Studies indicate that there are various individual and contextual factors that may influence marital naming attitudes and practices. Research has particularly shown differences between men and women's attitudes, as well as the influence of other family and community characteristics.

In a study of Midwestern college students (N=258), Scheuble and Johnson (1993) found that female students were more accepting than male students of married women keeping their birth names (92% versus 57%, respectively). Male students were also much more likely to view marital name-changing more important for relationship commitment than female students. Forbes et al. (2002) examined the attitudes of 197 college students about hyphenating surnames upon marriage. They found that both male and female students had positive perceptions of men and women who hyphenate. Other studies have shown that women who kept their birth name or hyphenated their surnames were perceived as less communal or less committed to the marital relationship than those changing their last name to that of their husband's (Etaugh et al. 1999; Stafford and Kline 1996). Nugent (2010) and others have argued that men's greater preferences for married women to change their names stems from invisible male privilege: women are expected to be more self-sacrificing for family, spouse and children. Men have been found to have more conservative attitudes about gender roles than women, possibly because men view gains by women as a threat to male privilege (Brewster and Padavic 2000; Goode 1992). This might be particularly the case for marital naming, which is an issue that strikes ‘closer to home’ than more public issues such as equal pay. Because of these gender differences in prior research and rationale for why men might be more opposed to non-conventional naming practices, we split our analyses by men and women and focus on within gender differences across time and region.

Number of years in college may be important for exposing students to more diverse practices and attitudes. Both men and women in higher education are more progressive than the general population (Brewster and Padavic 2000) and increased exposure to higher education may create more liberal attitudes relative to those who have just begun their college education. We expect that students who are further along in their education may have more progressive views on marital naming than other students.

Research suggests that there is both individual and intergenerational influence on gender-related attitudes (Moen et al. 1997). Family characteristics such as parental education level should be one mechanism to influence attitudes and plans regarding marital naming because higher education is linked to more progressive gender attitudes (Brewster and Padavic 2000). Students with more educated parents should be more exposed to such attitudes, which may carry over in to their personal attitudes and plans about marital naming.

Based on prior research (Moore and Ovadia 2006), we also expect that smaller community size may be linked to more traditional attitudes and practices about marital naming. Smaller communities are more likely to be rural areas as well as more homogenous in terms of population characteristics (Abrahamson and Carter 1986; Moore and Ovadia 2006; Nunn et al. 1978). As a result, we expect that students coming from smaller communities would be less exposed to nonconventional marital naming practices and should be less likely to espouse them.

Research Questions and Main Hypotheses

This research examines changes in plans and attitudes toward marital name choice. We build on prior research by focusing on changes over time and regional differences in plans and attitudes via the following four research questions:

1. Has there been a change between 1990 and 2006 among Midwestern college students’ attitudes in the belief that women who take their husband's surname at marriage are more committed to the marriage than those who do not change their surnames? This research question tests the competing explanations for whether attitudes about non-conventional marital naming have become more progressive over time, or if there has been a slowdown and possible reversal of accepting attitudes.

2. Have the marital name plans of women college students in the Midwest become more or less conventional between 1990 and 2006? Studies based on well-educated and non-representative samples would predict that women in 2006 will be more likely to plan to keep their birth surname upon marriage than in 1990 (Goldin and Shim 2004; Scheuble et al. 2000). An increase in the percentage of college students who plan to maintain their birth names when they marry would support these studies. In contrast, England (2010) includes women changing their last name to that of their husband as an indicator of a slow-down in gender changes. This research question is also a test of competing evidence: research that shows an increase in women retaining their birth names upon marriage versus the position of a slowdown in changes in gender role expectations for women.

3. Are there differences among college students in the Midwest and East in their views about commitment and marital naming? We expect that college students in the East will be less likely to tie commitment to the marriage to the wife's changing her name because of research showing more tolerance for progressive views in the East.

4. Is there a difference in 2006 in the marital naming plans of women college students in the Midwest and the East? Although there has been no previous research comparing region (other than the South) we expect that women in the Midwest will be more likely to choose conventional marital surnames than women in the East because of differences in the social climates of these regions.

Method

Sample I

The first data set consists of 258 personal interviews from students attending a small residential Midwestern college in the fall of 1990. Approximately one-third of full-time students were randomly chosen to be interviewed. Only four students refused. Sixty-one percent of the respondents were women and 39% were men; 32% were in their first year of college, 28% were in their second year, 22% were in their third year and 18% were in their fourth or fifth year. Race was not asked on this survey since the students were predominantly white.

Sample II

Sample II is also based on personal interviews with 246 students attending the same small residential Midwestern College as Sample I. These data were gathered in the fall of 2006. Fifty-nine percent of the respondents were women and 41% were men. The class rank of the respondents consisted of 28% first-year students, 28% second-year students, and 22% third-year students. Fourth- and fifth-year students comprised another 22% of the group. Ninety-two percent of the respondents in this survey were white and 8% (N=25) were non-white. Because the race variable was not asked in other two surveys, we did not analyze differences due to race across cohort and time period. A preliminary analysis for this sample only showed that race did not have a significant effect upon marital name plans or attitudes.

Sample III

The third data set is a convenience sample of students at a large Eastern university in a small city. The participants were enrolled in an introductory sociology or communications course during the fall of 2006. These were general education courses, so they encompassed a full range of student majors. Any student willing to participate was allowed to take part in the study, and an assistant distributed the surveys to the participants during class. A total of 369 students (54% women and 46% men) participated in the survey. Thirty-seven percent of these students were in their first year of college, 18% were in their second year, 20% were in their third year, and 26% were in their fourth or fifth year. This survey did not include a question about race of the respondent because students were predominantly white.

Dependent Variables

Two dependent variables were analyzed. The first item asked: “Do you think that a woman who changes her last name to that of her spouse is more committed to the marriage than a woman who does not?” This was coded “0” if no and “1” if yes. The second item asked the respondents whether they plan to change their last name to that of their spouse when they marry (0=no, 1=yes).

Independent and Control Variables

In order to examine the effects of cohort and region, we created two variables. “Cohort” compares respondents from the Midwestern college in 1990 (coded as 0) with respondents at the same Midwestern college in 2006 (coded as 1). “Region” compares the Midwestern college (coded 1) in 2006 with the large Eastern university in 2006 (coded 0).

We also examine the effects of six additional variables that have been found to affect marital naming plans and attitudes in other studies. Class rank was coded into four categories: first-year students (1) to 4 years or more (4). The size of the community where the respondent lived when s-/he was 16 years of age was coded into five categories ranging from less than 500 people (1) to 500,000 or more people (5). Mother and father's education was coded into five categories: grade school or less (1); high school (2); some college (3); college degree (4); and post-college education (5). The final variable included in this analysis is the respondents’ dating status which was measured in four categories, ranging from less to more serious relationships: single (1); casually dating (2); seriously dating (3); and engaged or married (4).

Missing Values

The majority of variables had less than 1% of item-level missing data. The highest was father's education with 5.4% of the data missing. In order to have the maximum number of cases available for the logistic regression analysis, missing data were imputed with the ICE procedure in Stata. Twenty data sets were imputed and the estimates were pooled using the MIM procedure in Stata to yield accurate standard errors and significance tests. We also ran the analysis using listwise deletion as a missing data strategy. The results were very similar. We present the regression analysis using the imputed data.

Results

Descriptive Statistics

Table 1 presents the percentage responses to the marital naming items by sample, split by gender of respondent. We conducted chi-squared tests, first comparing the two Midwest samples for differences between years and then comparing the Midwest and East 2006 samples for regional differences. The 2006 Midwestern sample has the lowest percentage of women planning to keep their birth surname at the time of marriage (4.3%); whereas women in the East 2006 sample had the highest at 11.6%. There was no significant difference between the Midwest samples, but the difference between the Midwest and East 2006 samples was significant (p.<05). For the question about whether or not women were more committed if they changed to their husband's last name, Midwestern women in 2006 were significantly more likely to agree (10.1%) than Midwestern women in 1990 (2.7%); the difference between Midwestern and Eastern women's attitudes, however, was not statistically significant.

Table 1.

Plans and attitudes of students in the three samples

	Planning to keep birth surname		More committed if change name
	%	N	%	N
Women
Midwest 1990	7.3	138	2.7	148
Midwest 2006a	4.3	141	10.1 *	139
East 2006b	11.6 *	189	5.2	191
N		468		478
Men
Midwest 1990	97.9	95	30.5	82
Midwest 2006a	96.9	98	23.5	98
East 2006b	100.0	151	31.2	157
N		344		337

Chi-square test of difference in the three samples within gender

^*p<.05, ^**p<.01, ^***p<.001

^aWithin gender comparison made with Midwest 1990 sample

^bWithin gender comparison made with Midwest 2006 sample

If we look at the percentages by gender, men across all the samples are overwhelmingly planning to keep their birth names and are much more likely than women to agree with the notion that a woman who changes to her husband's name is more committed to the marriage. There were no statistically significant differences for men across any of the comparisons made here, suggesting that men's attitudes are fairly stable across time and region.

Table 2 displays descriptive data for the control variables that are included in the regression models. Difference of means tests were conducted comparing the Midwest 1990 and 2006 samples and then the Midwest and East 2006 samples within gender categories. For women in the Midwest, both mother's and father's education were lower on average in 2006 compared to 1990. Midwestern women in 2006 were also much less likely to be in a more committed relationship than women in 1990. The only significant difference between the Midwest and East 2006 samples was that father's education was higher on average for Eastern women in 2006. For Midwestern men, those surveyed in 2006 came from smaller communities on average than those surveyed in 1990. Midwestern men in 2006 also came from smaller communities relative to Eastern men in 2006. Father's education was lower on average for Midwestern men in 2006 compared to those surveyed in 1990. Finally, dating status differed for both sets of sample comparisons: Midwestern men in 2006 were less likely to be in a committed relationship than Midwestern men in 1990; Eastern men in 2006 were less likely to be in a committed relationship than Midwestern men in 2006. These various sample differences further suggest the importance of controlling for these variables in the final regression models.

Table 2.

Descriptive information for control variables used in the logistic regression models

	Midwest 1990			Midwest 2006a			East 2006b			Min	Max
	N	Mean	SD	N	Mean	SD	N	Mean	SD
Women
Class rank	157	2.18	1.09	143	2.29	1.11	194	2.26	1.19	1 1st year	5 5th year
Community size at age 16	151	2.95	1.18	143	2.78	1.38	191	2.94	.96	1 <500 people	5 500,000+people
Mother's education	137	3.74	.77	143	3.38	1.00***	193	3.58	1.05	1 grade school or less	5 post-college
Father's education	140	3.66	.92	142	3.39	1.13*	189	3.70	1.11*	1 grade school or less	5 post-college
Current dating status Men	157	2.24	.88	143	1.86	.93***	194	1.94	.90	1 single	4 engaged/married
Class rank	98	2.44	1.20	101	2.49	1.11	163	2.45	1.28	1 1st year	5 5th year
Community size at age 16	90	3.20	1.26	101	2.60	1.35**	161	3.04	1.03**	1 <500 people	5 500,000+people
Mother's Education	90	3.69	.79	100	3.48	1.02	162	3.40	1.08	1 grade school or less	5 post-college
Father's education	85	3.75	.77	99	3.42	1.03*	161	3.68	1.18	1 grade school or less	5 post-college
Current dating status	98	2.30	.81	101	2.00	.99*	163	1.74	.89*	1 single	4 engaged/married

N is number of cases before imputation. Difference of means tests performed in the three samples within gender Community size is where a student lived before college. Dating status is ordinal, reflecting relationship commitment: single, casual dating, serious dating, engaged/married

^*p<.05

^**p<.01

^***p<.001

^aWithin gender comparison made with Midwest 1990 sample. ^b Within gender comparison made with Midwest 2006 sample

^bWithin gender comparison made with Midwest 2006 sample

Regression Analysis

The main focus of this research is the effect of time period and region on women's marital name plans and on both men's and women's attitudes toward marital naming issues. Table 3 presents the logistic regression models for the dependent variables, “More committed if change name” and “Planning to keep birth surname.” This table presents the effect of time before and after including the control variables. The analyses were conducted separately for women and men. To reiterate, we are testing competing hypothesis that attitudes about marital naming have become more versus less conservative overtime. Model 1 includes only the period effect. Compared to Midwestern women in 1990, Midwestern women in 2006 were more than three times as likely to believe that women are more committed to the marriage if they change their last name to that of their husband (OR=3.40). This period effect remained significant when the control variables were added in Model 2. Only one of the control variables—father's education—was statistically significant. For each unit increase in the category of father's education, female students were 1.7 times more likely to agree that married women who change their surname are more committed to the marriage. For Midwestern men, the year of the survey did not have a significant effect on whether they felt a married woman was more committed if she changed her name. Additionally, none of the control variables had significant effects in Model 2. To address our first research question, the findings in Model 1 and 2 suggest that while men's attitudes have not changed over time in the Midwest, women's attitudes appear to have become more conservative about marital naming and commitment.

Table 3.

Logistic regression models comparing the Midwest samples (1990 v. 2006)

	More committed if change name				Planning to keep birth surname
	Model 1		Model 2		Model 3		Model 4
	b	exp(b)	b	exp(b)	b	exp(b)	b	exp(b)
Women (n=300)
Year (1990 = 1; 2006=2)	1.22	3.40*	1.35	3.84*	–0.36	.61	–0.58	.61
Class rank			–0.13	.88			.22	1.32
Community size at 16			.03	1.03			.25	1.17
Mother's education			–0.14	.87			.35	1.55
Father's education			.55	1.73*			–0.56	.70*
Current dating status			.04	1.05			–0.26	.80
Constant	–4.66		–6.32		–2.24		–2.09
Nagelkerke R2	.04		.08		.00		.07
Men (n = 199)
Year (1990 = 1; 2006=2)	–0.31	.73	–0.30	.74
Class rank			.01	1.01
Community size at 16			–0.02	.98
Mother's education			.13	1.14
Father's education			–0.15	.86
Current dating status			.22	1.24
Constant	–0.61		–0.99
Nagelkerke R2	.00		.01

Missing values were imputed in 10 datasets and pooled estimates are reported

^*p<.05

**p<.01

***p<.001

Models 3 and 4 present the findings for women only regarding plans to keep their birth surname. These data are presented only for women, as almost each male in these samples said he planned to keep his birth name when he married. Similar to the analysis for attitudes, we were testing competing hypothesis that women's plans for marital naming have become more versus less conservative over time. Time period did not have a significant effect on plans to keep birth surname at marriage, thus indicating that there was not a significant change in plans between 1990 and 2006. The findings in Models 3 and 4 do not suggest support for either hypothesis (i.e. that non-conventional naming practices are increasing or decreasing over time) at least for Midwestern women. Again, father's educational attainment was the only significant control variable. For every unit increase in the measure of father's education, women were 30% less likely to plan to keep their birth surnames at marriage. This finding is not consistent with our expectations, given the association between higher education and more progressive gender attitudes (Bolzendahl and Myers 2004). However, it is possible that men, regardless of education level, remain relatively more opposed to women keeping their birth names and this may be reflected here in their daughters’ attitudes.

Table 4 presents the effects of region of the country and the control and independent variables on the two outcomes. Again, the findings for “More committed if change name” are presented in Models 1 and 2. We hypothesized that both Eastern women and men would have more progressive attitude regarding non-conventional marital naming and relationship commitment compared to those in the Midwest. For both women and men, there was no significant regional effect on attitudes either before or after including the controls. Therefore, the hypothesis for our third research question was not supported here. In Model 2 for women only, father's education was again significant: as father's education increased, women were more likely to agree that women were more committed if they changed to their husband's last name.

Table 4.

Logistic regression models comparing the Midwest and East regions (2006)

	More committed if change name				Planning to keep birth surname
	Model 1		Model 2		Model 3		Model 4
	b	exp(b)	b	exp(b)	b	exp(b)	b	exp(b)
Women (n=337)
Region (Midwest=0; East= 1)	–0.68	.51	–0.84	.43	.97	2.63*	1.04	2.83*
Class rank			–0.20	.82			.21	1.23
Community size at 16			–0.11	.89			.26	1.30
Mother's education			–0.05	.95			.56	1.75*
Father's education			.54	1.71*			–0.27	.76
Current dating status			.08	1.09			–0.35	.70
Constant	–0.85		–1.72		–4.90		–6.85
Nagelkerke R2	.02		.06		.03		.09
Men (n=264)
Region (Midwest=0; East= 1)	.47	1.47	.52	1.67
Class rank			.07	1.07
Community size at 16			–0.10	.90
Mother's education			–0.12	.89
Father's education			–0.05	.95
Current dating status			–0.05	.95
Constant	–2.17	–1.49
Nagelkerke R2	.01	.02

Missing values were imputed in 10 datasets and pooled estimates are reported

^*p<.05

**p<.01

***p<.001

Marital name plans for women only were examined in Models 3 and 4. We expected that women in the East would be more likely to plan a non-conventional marital name than women in the Midwest. For this outcome, there was a significant regional difference. Women in the East were almost three times more likely than women in the Midwest to say they planned to keep their maiden names when they married and this finding was consistent both before and after controls were included (OR=2.63, 2.83 in Models 3 and 4 respectively). Our hypothesis for this outcome was supported. Mother's educational attainment was the only control variable to affect marital name plans. The women students’ plans to keep their birth name increased as mother's education rose. Mother's education may be an important factor in affecting women's marital naming plans for several reasons: first, it may expose younger women to more progressive gender attitudes; second, depending on the mother's educational and career choices, she may have opted to keep her birth name for personal or professional reasons, thereby creating an example that daughters might follow.

Discussion and Conclusion

The objective of this study was to add to the marital naming literature by examining time period and regional effects upon marital naming attitudes and plans. Prior research has produced competing evidence that marital naming has become both more and less conventional over time (Goldin and Shim 2004; Scheuble et al. 2000). We used this tension in the literature to see which hypothesis was supported in our data. In terms of time period effects, we found that attitudes about marital naming appear to have become more conservative over time, but there was not a significant change in marital name plans in the 16-year period we examined. Women were neither more nor less likely in 2006 to say they planned to change their last name to that of their spouse when they married than they were in 1990. This suggests overall that there is a disjuncture between attitudes and plans, and that neither of the hypotheses (increasing or decreasing non-conventional choices) were supported in our study regarding actual plans for women's marital naming.

Although some researchers have reported an increase in the percentage of women keeping their birth surname (Scheuble et al. 2000) and others a decrease (Goldin and Shim 2004), our findings show time to have had little influence on marital naming plans, similar to the findings by Hoffnung (2006). The difference between our findings and those of previous studies may be due to sampling differences and the fact that we asked about plans rather than about actual behavior. Our respondents were not as highly educated as those in other studies and this certainly can affect the findings as well. Yet because our respondents are Midwestern and Eastern students at all points in their college careers, rather than college graduates from an elite school or people who announce their weddings in a prestigious newspaper, we suggest that our findings may be more amenable to generalization to a wider population.

Further analysis (not shown here) found in both time periods that, among the women planning to keep their birth surname, none thought a woman is more committed to the marriage if she changes her last name to that of her husband. Although there appeared to be no change in the attitudes of women who were less conventional, there appeared to be a movement toward more conservative attitudes in women who plan to follow the traditional practice of changing their marital surname. This may signify an increasing polarization between women with traditional and non-traditional values.

Another goal of this research was to examine the effect of region upon marital naming attitudes and plans among women in 2006. We hypothesized that those in the East would be more progressive in terms of both attitudes and plans than those in the Midwest because of different social climates (Cahn and Carbone 2010; Rice and Coates 1995). We did not find a regional difference in terms of attitudes about marital naming and commitment; however, we did find that women at the Eastern university were significantly more likely to say they planned to keep their birth surname than the women at the Midwestern college. Although differences in marital naming practices and gender attitudes have focused on comparing the South to the rest of the country (Johnson and Scheuble 1995), there are a number of reasons why women in the East are more likely to make a non-conventional last name choice than women in the Midwest. In particular, those students living in the East may have more tolerant views toward social issues (Carter and Borch 2005; Powers et al. 2003; Rice and Coates 1995) and be more likely to think keeping one's birth name at the time of marriage is an acceptable marital name choice.

Despite the speculation that women are becoming more traditional about marital last name choice in the most recent decade (Dubecki 2007; Hampson 2007), we found little systematic evidence in the current study that this is happening—rather there appear to be uneven changes dependent on time period and region that appear in plans and attitudes. Changes in gender expectations have occurred more rapidly in the labor force and more slowly in the area of home and family (Blau et al. 2006; England 2010, 2011). Our findings are consistent with those findings, as there has been little change in the percentage of women planning to keep their name over the 16-year period between 1990 and 2006. Retaining one's birth surname at the time of marriage is still an uncommon event (Brightman 1994; Johnson and Scheuble 1995), and our findings confirm this; only 5–12% of the college women in our samples plan to keep their birth name when they marry.

Changing one's birth surname at the time of marriage is clearly ingrained in the norms and values of marriage in the United States (at least for white women; see Hoffnung (2006) and Kerns (2011) for research addressing racial/ ethnic variation). Although there is no evidence that women are less likely to plan to change their last name upon marriage now as compared to 2006, there is some evidence in our study that women who planned to take their husband's last name were more likely to say that doing so is an indicator of commitment to the marriage. Larger, more representative surveys of marital naming attitudes, plans, and behaviors are needed to resolve questions about national trends and regional differences more definitively.

Limitations

This study sought to contribute to prior literature on marital naming by addressing gaps in the existing research; however, there are still some limitations to note. First, this study only used two time points and two regions for comparison. Future research should focus on addressing multiple points over time and examining various regions for differences in plans and attitudes. Second, this study relied on three convenience samples drawn from two schools—nationally representative data from the general public would of course be beneficial for better generalizing findings to the broader population. Third, although racial and ethnic variations in marital naming practices have been documented in prior research (Hoffnung 2006; Kerns 2011), we were unable to control for race/ethnic differences in this study. A study oversampling for racial and ethnic minorities would be beneficial in order to ascertain whether regional and time period differences hold across these subgroups.