Higher Education 32: 1-22, 1996. (E) 1996 KIuwer Academic Publishers. Printed in the Netherlands.
1
Divorce, gender role, and higher education expansion YANFENG TIAN Telesec, LMB 34 Madison Building, The Library of Congress, Washington, D.C. 20540, USA
Abstract.In Western societies the past two decades have witnessed a great increase of women's participation in higher education and a multiplication of responsibilities as a result of the expanded role of women. This article examines higher education enrollment between 1970 and 1987 in the United Kingdom and the United States. It is guided by a theoretical model contending that educational expansion occurs where people choose education as an adaptation to increased deprivation and uncertainty. By linking the expanded role of women to the expansion of women's participation in higher education, and by using the rising rate of divorce as an indicator of the increased disenfranchisement among women in patriarchal societies, this study attempts to uncover the hidden forces behind the recent expansion of women's participation in higher education. Multiple regression is used for the statistical analysis. The results show that the divorce rate is positively related to women's enrollment in higher education in the United States. The results also show divergent enrollment patterns between the two sexes regarding the effect of unemployment in both the United States and the United Kingdom. It is argued that gender role expectation is crucial to understanding the different effects of divorce and unemployment on the divergent enrollment patterns of men and women. The article also explores important differences in the areas of educational opportunities, the magnitude of the impact of divorce, government policies regarding women's welfare, and the differing role of credentials in social mobility in the two countries under study.
Introduction T h e degree o f e x p a n s i o n o f w o m e n ' s enrollment in higher education during the last two decades has been remarkable. In 1970, only a few of the industrialized nations had w o m e n ' s enrollment a b o v e forty percent. B y 1986, however, w o m e n ' s enrollment in m o s t countries represented in O E C D a p p r o a c h e d fifty percent. Several countries, including the United States, have pushed w o m e n ' s e n r o l l m e n t a b o v e fifty percent. This study seeks to explore the i m p a c t o f social changes experienced by w o m e n on the expansion of higher education in the United K i n g d o m and the United States. 1
Theories of educational expansion Several theories try to explain educational expansion. Research literature on tertiary education and c o m p a r a t i v e studies has, however, concentrated
2 on human capital theory, conflict theory and world system theory. Human capital theory, which emphasizes the changing structure of the job market and the changing patterns of the needs for specific jobs, was developed in the context of modernization theories. This theory perceives a linear, marketrelated relationship between the need for trained human resources and the expansion of education. Education is considered as an investment that will pay off later in the form of better paying and/or more prestigious jobs. As returns on educational investment increase, so will the number of people seeking more education (Becker 1964; Blaug 1970; Freeman 1976; Mincer 1958; Schultz 1961). Human capital theory has lost much ground since the late 1970s. It was challenged by conflict theorists with evidence that the supply of highly educated persons in the job market has in some cases grown more rapidly than the demand, and that the surplus has resulted in an inflation of credentials and declining returns to specific levels of educational attainment. Conflict theory argues that this credential crisis, and educational expansion in general, are an outcome of a process of status competition among groups. Because education is closely linked to power and status, different groups will seek to improve their position by seeking more education. Therefore, in a final analysis, it is competition among groups for status and prestige rather than the needs for more trained personnel that accounts for educational expansion (Collins 1971, 1979; Boudon 1974). Research guided by world system theory, however, rejects both human capital theory and conflict theory (Meyer et al. 1977; Ramirez & Rubinson 1979; Ramirez & Boli-Bennett 1982). World system theory holds that in modern societies, the state is the major institution responsible for nation building and social development. The actual implementation of those goals, however, is dependent on individual efforts. It is suggested that citizenship is the key to link the individual with the state. Citizenship entails that individuals are contributors to national development as well as legitimate consumers enjoying all benefits provided by the state. School education is essential to teaching skills necessary for the citizens to engage in nation building and individual development. The state as an institution is the major actor in the development of the educational system. Educational expansion is therefore a product of the joint effort from the state and individuals to produce the citizen. Although world system theory hints at a possible role of individuals in educational expansion, its emphasis is on the state and the role of the state in educational expansion. All three theories above can help us understand certain aspects of educational expansion. However, comparative studies under those theoretical models are often characterized by macro analysis of a large number of cases, and
the distinctive structure of an individual educational system that may encourage or constrain a particular case of educational growth is lost. Moreover, theories and comparative studies that deal with educational expansion have long been gender blind. Some of the recent studies (Benavot 1989; Gamier & Hage 1991; Karen 1991; Roger 1991; Waiters 1986) have focused on how women's participation in education either has affected the national economic growth or has been affected by the labor market, group competition, or other conventional demand or supply factors. None of them, however, has related the increase of women's participation in education to the possible impact of recent social changes being experienced by women. From this perspective, Craig and Spear's theoretical model, which permits a closer examination of particular social aggregates in a given educational expansion, has something more to offer. John Craig and Norman Spear's theoretical model has been mostly articulated in the essay "Explaining educational expansion: an agenda for historical and comparative research" (Craig & Spear 1982). They believe that any social change is likely to have different consequences for different social actors. Educational expansion is most likely for those groups, primary or corporate actors, who have found themselves lost in the changing social structures and in the experience of growing deprivation and uncertainty, or perceive themselves as such. For them, education is likely to become a vehicle to change, and educational expansion is the result of using education as an adaptation. For this reason, the decline of the old social structure is often a precedent for a rapid educational expansion. The key concept that links educational expansion and a particular social aggregate in this theory is that of deprivation and uncertainty. Craig and Spear distinguished two types of uncertainty; unbounded uncertainty and bounded uncertainty. The less bounded the uncertainty, the more likely the deprivation is real, the range of options is extensive but vague, and the harder it is to foresee the outcome of the various options. For bounded uncertainty, the opposite is true. The more bounded the uncertainty, the more likely that the deprivation is relative, and the options and their outcomes appear more or less approachable. The authors believe that in modem society the general movement is toward pushing unbounded uncertainty to bounded uncertainty where relative deprivation can only be alleviated through achievement, and education is often regarded as a necessity for such an achievement. However, the direction of change is not only affected by the actual or perceived deprivation and uncertainty, but also by options available to people who are seeking a change.
4
Role change and women's experience of uncertainty To explore the forces behind the recent expansion of women's participation in higher education, one must understand how women's lives have been affected by recent social changes, and how women have increasingly found themselves in a situation of deprivation and uncertainty. During the past two decades, women's roles have been greatly expanded by their presence in the paid employment and other places that were traditionally reserved for men. Women, however, continue to do the majority of housework; the expanded role of women has not resulted in a fair redistribution of household labor. Nor has the structural change of the job market followed a redistribution of economic compensation; women are still largely confined to jobs offering low pay and little prestige. Evidence worldwide demonstrates that the change is "asymmetric", and the assumption of more work outside the home by women has resulted in "role overload" of women (England & Swoboda 1988). Nowhere is the "role overload" of women so drastic and demanding as what follows after divorce, when many women are thrown into economic hardship and overburdened with childcare responsibilities. In patriarchal societies, though varying in degrees, women typically earn less than men. The rising rate of divorce during the past two decades across industrialized countries has resulted in a large number of households headed by separated or divorced women. Financial support for children from their ex-husbands is limited. Depending on the structure of government assistance, the economic situation for female-headed families ranges from "economically weak" in Sweden to a poverty rate that is "disproportionately high" in the United States (Dahn & Kamerman 1988; Goode 1993; Weitzman & Maclean 1992). Divorce dramatizes the dilemma of the expanded role of women; women are increasingly expected to be both wage earner and care taker without adequate support and fair compensation for their contribution to the family and society. Hence, the higher the divorce rate, the greater the number of women are exposed to increasing deprivation and uncertainty. Consequently, they are motivated to seek alternatives to reduce their uncertainties. Going to college exemplifies one way women respond to their disenfranchised status. A college degree is seen as a means of achieving economic and social mobility. While little research has been done in this area, there is evidence that in the United States divorced women are over-represented in colleges (Aslanian & Brickell 1988), and that financial need and role change were motivating factors for divorced women to return to college (Clayton & Smith 1987). Moreover, studies in the United States also indicate that female college students with divorced parents had more work experience, were more independent and responsible (Vess, Schwebel & Moreland 1983), and were
more career oriented (Measel11992). Apparently, the effect of divorce and the seeking of alternatives are not only gender related, but also inter-generational in character.
Purpose of the study The purpose of this research is to apply Craig and Spear's theory of educational expansion to the increased participation of women in higher education during the past two decades in the United Kingdom and the United States. By relating the concept of deprivation and uncertainty to the expanded role of women and to the increased rate of divorce, this study seeks to explore a new way of understanding the causal forces of the recent expansion of women's participation in higher education. Not only is this study an examination of the social forces that shape educational expansion, it is also a comparative investigation of how the expansion patterns differ between the two sexes and how the structure/nature of higher educational institutions in different countries influence those patterns.
Research methods
Research hypotheses As described above, Craig and Spear distinguished two types of uncertainty: unbounded uncertainty and bounded uncertainty. While there is no further deliberation about the dimension of deprivation and uncertainty, the example given by the authors shows that in their own study of university expansion in Europe, occupations in decline were treated as an equivalence of uncertainty. Accordingly, they found that male children with fathers in declining occupations were responsible for most of the enrollment increase in the late eighteenth and the early nineteenth centuries (Craig & Spear 1982). It appears that an immediate consequence of losing a trade to make a living would be economic deprivation and uncertainty. In this light, the concept of uncertainty and deprivation proposed by Craig and Spear suggests an emphasis on economic consequence. This study seeks to explore the relationship between the increase of women's participation in higher education and the impact of recent social changes being experienced by women. As pointed out above, in industrialized countries divorce rates have been rising rapidly during the last two decades, and divorce tends to intensify .women's economic hardship and multiplication of responsibilities. Financial hardship is only part of the problem. For
many divorced women with children, the physical and psychological stresses caused by the "overload" of responsibilities are equally overwhelming. To miss the non-economic dimension of deprivation and uncertainty is to miss the socially-structured gender differences. Therefore, this study takes divorce rate as an indicator of deprivation and uncertainty that is characterized by financial hardship and multi-burden of responsibilities. This study assumes that a higher rate of divorce would put more women in deprivation and uncertainty, and likely lead more women to seek higher education as an alternative. The US and the UK have some of the highest divorce rates in the world (Goode 1993; Kurian 1991). This study predicts that divorce rate in a highdivorcing country is likely to be associated with the increase of higher education enrollment. Craig and Spear's theory maintains that only those social aggregates who perceive the threat of deprivation and uncertainty are likely to seek the remedy of further education. Since the consequences of divorce are asymmetrically damaging to women, one can predict that in a high-divorcing country, an increase in the divorce rate will mostly affect women's, but not men's, enrollment in higher education. Accordingly, this study is designed to analyze and compare enrollment patterns of both men and women in higher education in the UK and the US during the past two decades.
Variables In the analysis, divorce rate is used as the independent variable and female/male enrollment as the dependent variable. Divorce rate is, technically, the number of divorces per 1000 population. Divorce rate is used here as an indicator of deprivation and uncertainty to which women as a group are ever increasingly exposed. Alternative measures of divorce are also available- the number of divorces per 1000 marriages for the UK, and the number of divorces per 1000 married women 15 years of age and over for the US. The analysis result will also be tested with the alternative measures of divorce for each country. Besides the increase of the divorce rate, there are other factors that may affect higher education enrollment. These factors must be held constant. One such factor is the population of the available pool for higher learning candidates. A related factor is the number of high school graduates as college candidates. Another possible factor is the effect of the larger economic environment. This study uses the unemployment rate as an indicator of the overall economic situation. Findings from previous research on the effect of the economy on education were mixed. A recent cross-national study (Windolf 1992) shows that a bad economic situation is positively related to university expansion.
Since women as a group are relatively new in the job market and in the higher education scene, it would be interesting to see how women's enrollment responds to a bad economic situation during which women are more likely to lose their jobs before their male counterparts. The general atmosphere of higher education in each country is another source of influence. The change of atmosphere, whether in the national economy, politics, or population in general, will certainly affect the overall growth of higher education among both men and women. Since any measure of enrollment that contains women would produce a confounding statistical effect, male student enrollment as an indicator of the general trends of the system growth is used as a control variable. It is here that the analysis concept differs between the two sexes. In the analysis of female enrollment, male enrollment, as a control variable, is an indicator of the general trends of the system growth. In the analysis of male enrollment, it would be inappropriate for female enrollment to serve the same purpose as an indicator since women's participation in education has universally lagged behind men's, and since provision for equal access to education in general and higher education in particular is only a recent accomplishment. Therefore, female enrollment, as a control variable in the analysis of male enrollment, is used as an indicator of the trend for educational equality. Altogether, we have the following variables: divorce rate is used as the independent variable and female/male enrollment as the dependent variable. Enrollment of the opposite sex, female/male population for ages of 15 to 64, the female/male unemployment rate, and the number of female/male high school graduates 2 (undergraduate enrollment when the dependent variable is graduate enrollment), are used as control variables. Student enrollment of each sex has subvariables. There are equal numbers of variables and subvariables for each sex. Data for women and men has been collected and analyzed separately. All data covers the period from 1970 to 1987 and is from published sources. 3
Statistical analysis Multiple regression is used for the analysis. Two-tail test is adopted and the confidence level is at 0.05. The statistical analysis is comprised of two steps. First, all original data is adjusted into the mode of the annual rate of increase. That is, original numbers are converted into the mode of percentage changes in a given year over the previous year.
8
Table 1. Variables* Variables
United Kingdom & United States
independent variable dependent variable (five subvariables)
divorce rate 1. total tertiary enrollment 2. university enrollment 3. non-university enrollment4 4. undergraduate enrollment5 5. graduate enrollment
control variable
1. enrollment of the opposite sex 2. number of high school graduates (undergraduate enrollment when the dependent variable is graduate enrollment) 3. population for ages 15 to 64 4. unemployment rates
*All variables have separate male and female variables and subvariables except for the divorce rate.
Equation: Y = (Xl - X)(lOO/Xl)
where Y = the annual rate of increase, Xl = variable in a given year, and X = variable in the previous year. Second, stepwise multi-regression is introduced. Stepwise multi-regression is a procedure in multiple regression often used to study theoretical constructs about relationships among variables. Although all variables adopted in this study except for divorce rate have been tested in various studies one way or another for their possible influence on educational expansion, little is known about how they are related to divorce rate or how they differ between the two sexes at the college level. Stepwise procedure works in such a way that it serves to screen out any variable that does not significantly contribute to the predictability. Logically, if divorce rate is not in any way contributing to the enrollment growth, it will be removed from the analysis process as a potential contributor. Equation: Y = b l X l + b2X2 + ...bnXn + bo, where Y is the predicted dependent variable, bib2 ... bn are the regression coefficients for the respective independent variables, X1X2 ... Xn are the independent variables, and bo is the Y-intercept.
Results
United States In Table 2, all five female enrollment variables responded to the divorce rate in the positive direction, and the increment R2s range from the lowest of 0.1616 for female graduate enrollment to the highest of 0.5464 for nonuniversity enrollment. All are significant at 0.05 or less. The result shows that divorce rate has an independent effect on female enrollment. Divorce rate has a greater effect on female non-university enrollment while the least affected is female graduate enrollment. Female unemployment also shows an independent effect on female graduate enrollment, though in the negative direction. Both the increment R 2 0.0844, and the SRC - 0.3083, are significant at 0.05. That means that a bad economic situation is negatively related to female enrollment of certain section. A striking difference is revealed when we look at the effect of divorce on male enrollment from Table 3. Out of five male enrollment variables, divorce rate is significantly related to only one variable, male graduate enrollment. The increment R 2 is 0.0798, and the SRC is in the negative direction. No positive relation between divorce rate and male enrollment is found. What is equally striking is the relationship between male unemployment and male enrollment. Out of five male enrollment variables, four are positively related to the unemployment rate and significant at 0.05 or less. The increment R2s range from the lowest 0.2051 to the highest 0.5061. The unemployment rate has a greater effect on male four-year college/university enrollment (R 2 0.4162) than on male two-year college enrollment (R 2 0.2051). Male undergraduate enrollment (in both two-year and four-year colleges) yields R 2 0.4986.
United Kingdom In Table 4, only one enrollment variable, female university enrollment, has the increment R 2 0.0066 with a negative SRC. But it fails to reach the significance level at 0.05. However, the unemployment rate is negatively related to female university enrollment with the increment R 2 0.0089 significant at 0.05. In Table 5, male unemployment yielded increment R2s 0.3516 and 0.3528 for male tertiary enrollment and male non-university enrollment, respectively, with both significant at 0.01. Both SRCs are in the positive direction. As in the US, the unemployment rate in the UK has different effects on student enrollment of different sexes. In both the US and the UK, the unemployment
Variable
Graduate
5
0.6324*** 0.1154
0.7545 . 0.2396
SRC SE
0.3798** 0.1294
0.5639***
Increment R 2 0.1616"*
0.4759** 0.1082
SRC SE
0.5370*** 0.1137 0.4857***
0.5549*** 0.1824
Increment R 2 0.2223**
SRC SE
0.3977* 0.1567
SRC SE 0.2544***
0.1308"
Increment R 2 0.2033* 0.6166"* 0.0790
Control
.
-0.3109 0.3324
0.0661
.
.
0.4298* 0.0539
0.3672**
0.3083* 0.0369
0.0844*
Female unemployment rate
CN
CN
CN
CN
Stepwise R 2
* p < = .05. **p < = .01. ***p < = .001. SRC = Standardized regression coefficient. CN = Regression intercept.
Undergraduate
Non-university Increment R 2 0.5464***
3
4
University
0.5915"** 0.1235
SRC SE
2
0.4612"*
0.5187"* 0.1025
Control
Female Female H.S. Male population graduate/ enrollment 15 to 64 undergraduate
Control
Increment R 2 0.2613"*
Divorce rate
Control
Total tertiary
Independent
1
US Dependent
F
Table 2. Stepwise analysis of female enrollment in the United States
2.946356 0.67428957
0.8099
2.602137 0.44935640
0.7080
3.362250 0.72486751
0.8008
2.697127 0.49583270
0.7673
2.644194 0.42345246
0.7226
Dependent
Total tertiary
University
Non-university
Undergraduate
Graduate
US
1
2
3
4
5 0.9403*** 0.0888
SRC SE
-0.3489* 0.0923
0.5639***
0.4874"* 0.2051
SRC SE
Increment R 2 0.0798*
0.1963"*
0.5629*** 0.1637
SRC SE
Increment R 2
0.5291"**
0.4458* 0.1648
SRC SE
Increment R 2
0.1925"
Increment R 2
0.4757** 0.1838
population 15 to 6 4
SRC SE
Male
enrollment
Control
Female
Control
0.1927"*
Divorce rate
Control
Increment R 2
Independent
0.4056** 0.0806
a0.1322"
graduate/ undergraduate
Male H.S.
Control Male
0.5031 ** 0.0304
0.4986***
0.4818"* 0.0469
0.2051"*
0.5661"* 0.0163
0.4162"*
0.5282** 0.0263
0.5061"**
unemployment rate
CN
CN
CN
CN
Stepwise R z
- 1.229680 0.38981712
0.7559
- 1.361531 0.91404326
0.6949
-1.354003 1.26923801
0.7342
-0.990916 0.48947921
0.6087
- 1.206919 0.82268863
0.6989
*p < = .05. **p < = .01. ***p < = .001. SRC = Standardized regression coefficient. CN = Regression intercept, a The control variable is undergraduate enrollment.
Variable
M
Table 3. Stepwise analysis of male enrollment in the United States
Dependent
Total tertiary
University
Non-university
Undergraduate
Graduate
UK
1
2
3
4
5
0.9054 0.0569 0.1814 0.4259 1.2436 0.6074*** 0.7794*** 0.2006 0.2949* 0.5430* 0.2495
Increment R 2
SRC SE
Increment R 2
SRC SE
Increment R 2
SRC SE
-0.1058" 0.0556
SRC SE
Male enrollment
Control
0.9600***
Divorce rate
Control
Increment R 2 0.0066
Independent
-0.1100" 0.0202
0.2124"* 0.1245
Female unemployment rate
0.0089*
Female H.S. graduate/ undergraduate
Female population at 15 to 64
0.0081
Control
Control
CN
CN
CN
CN
Stepwise R 2
*p < = .05. **p < = .01. ***p < = .001. SRC = Standardized regression coefficient. CN = Regression intercept.
Variable
F
Table 4. Stepwise analysis of female enrollment in the United Kingdom
4.244281 0.68317315
0.2949
2.982859 0.43628889
0.6074
-0.484115 3.96459613
0.1814
3.059296 0.41334672
0.9835
t~
Dependent
Total tertiary
University
Non-university
Undergraduate
Graduate
UK
1
2
3
4
5
0.2949* 0.5430* 0.2355
SRC SE
0.7794*** 0.1631
SRC SE
Increment R 2
0.6074***
-0.5229* 0.0720
SRC SE
Increment R 2
0.2007*
0.9798*** 0.0474
SRC SE
Increment R 2
0.9600***
Increment R 2
0.8637** 0.0299
0.3528**
0.5930** 0.0368
unemployment rate
Male
Control
SRC SE
graduate/ undergraduate
Male H.S
Control
0.3516"*
Male population at 15 to 64
enrollment
Control
Female
Control
Increment R 2
Divorce rate
Independent
CN
CN
CN
CN
CN
*p < = .05. **p < = .01. *** p < = .001. SRC = Standardized regression coefficient. CN = Regression intercept,
Variable
M
Table 5. Stepwise analysis of female enrollment in the United Kingdom
-1.545388 1.28783618
0.2949
-1.770304 0.702611
0.6074
1.589211 0.46327074
0.5535
-2.234350 0.40411560
0.9600
1.422222 0.66175776
0.3516
Stepwise R 2
t.~
14 rate has a greater effect on male than on female enrollment in the absolute magnitude. Test
The results of the analysis for both the US and the UK have been tested by replacing the independent variable, the number of divorces per 1000 population, with the number of divorces per 1000 marriages for the UK and the number of divorces per 1000 married women 15 years of age and over for the US. The results of the tests for both countries are consistent with those that have been presented above.
Discussion In summary, the multiple regression analysis of this study shows that in the US the increased divorce rate is positively associated with the increase of female enrollment and negatively associated with male enrollment of certain section. The findings also reveal that the unemployment rate is positively related to male enrollment but negatively to female enrollment. No relationship is found between divorce rate and female enrollment in the UK. There are several noticeable findings in the US. First, the analysis shows that the divorce rate has a greater effect on female enrollment in two-year colleges than in four-year colleges. This finding is in line with the report by the American Council on Education (1987) that except for recent female high school graduates, women college students were largely part-time and concentrated in community colleges. This fact largely reflects both the economic situation of women and the lower cost of two-year colleges. There is evidence that those from lower income families are more likely to attend twoyear colleges (Astin et al. 1983; Clowes, Hinkle & Smart 1986). Second, in contrast to female enrollment, the growth of male enrollment generally is not affected by the rising rate of divorce. But divorce rate does have a significant but negative effect on male enrollment at the graduate level. An explanation may partly be found from the report prepared for the Carnegie Commission. In a comparison of men and women at graduate schools (Feldman 1974), it was reported that marital status had a different effect on the performance of graduate students of different sexes. Marriage had a detrimental effect on female graduate students but a positive effect on male graduate students. Interestingly, the effect of divorce went diverged too; both married men and divorced women were much more committed and had much better academic performances than divorced men, who were marked with lack of commitment and a less satisfactory academic performance.
15 Divorce effect in the cultural contexts
While findings in the US support the research hypothesis, findings in the UK do not. However, further investigation of both "pull and push" factors of educational expansion reveals some important differences between the US and the UK. Here, we consider demand as the push side and supply as the pull side; the rise in divorce as a force leading to the demand, and educational opportunity feeds the supply side. In 1987, the divorce rates in the UK were 2.9 per 1000 population/12.6 per 1000 married couples. In the same year, divorce rates in the US were 4.8 per 1000 population/20.8 per 1000 married women 15 and over. It is estimated that about one third of marriages in the UK and one half of marriages in the US end in divorce. In the US, divorce is widely distributed across all social segments. According to one calculation nearly fifty percent of American fathers who at present have a daughter are likely to see their daughters heading their own household (Garfinkel & Wong 1990). In contrast, divorce and separation in the UK are most likely to be found among the lower social classes and the unskilled manual occupational groups (Haskey 1984; Fletcher 1988), the segments for whom college education has been elusive. The skewed distribution of divorce in the UK to some extent reflects a lower level of a legal recognition of divorce and the reluctantlychanging attitude toward divorce (Agell 1992). While the UK has a lower divorce rate, its social welfare system takes better care of women in general, and female-headed families in particular (Joshi 1990; Kamerman & Kahn 1991; Weitzman & Maclean 1992). British women enjoy some benefits that American women do not - such as universal medical care, family allowance, maternal leave, etc. Moreover, in Britain single parenthood is regarded as a sufficient reason for not being available for employment, and the parent has a choice either to remain at home or to work. This is in great contrast to American women's experience revealed by Weitzman's Divorce Revolution where, under the name of sexual equality, divorced women are expected to work regardless of their responsibility for minor children, their physical capability or other complex circumstances (Weitzman 1985). In 1988, the labor force participation was 18 percent for British single mothers with a child aged 0-4 and 37 percent for all British mothers (Snyder 1992). In the same year, it was 45.6 percent for American single mothers with children under 3 and 52.5 percent for all American mothers (Kamerman & Kahn 1991). Apparently, proportionally fewer British women have ever been divorced. Through assistance from the welfare state, British women fare better in face of a marital crisis and are under less pressure to take a job when they have children to care for. Lopata (1993) commented
16 that American women are carrying such a heavy burden that another round of women's revolt may ensue if remedies are not provided. While the perceived deprivation and uncertainty may be translated into a demand for higher education, any demand must be subject to constraint of supplies. To the extent that education can be an option for change, therefore, educational opportunities must be made available in the first place. In the UK and the US, school systems are organized differently due to different historical and cultural precedents. The most basic difference is that the British system has historically been more highly stratified than the American system, which inspired Turner to develop his well-known thesis of "sponsored" versus "contest" mobility (Turner 1960). Since then, the British educational system has gone through many changes aimed at curtailing the effect of social stratification upon educational attainment. The achievement of greater sexual equality in higher education has also been given increasing attention since the mid-1970s. However, as of 1987, women made up only 41 percent of the overall university student population and about one third at the postgraduate level; college students accounted for less than a quarter of the 20-24 age group. The comparable figures for the US was 51, 52 and 60 percent respectively. Ironically, after many changes, it is mostly British women who are continuously confined within the traditional "sponsored system" while British men are more likely to take advantage of non-traditional modes of study such as further education and part-time programs (Bourner et al. 1991; Kerckhoff 1993). There is also evidence that most gains made by working class people in access to higher education in recent years were for men; little progress has been made for working class women in either traditional or non-traditional institutions (Woodley et al. 1987). Gender role and the unemployment effect
Another important finding from this study is the effect of the unemployment rate on student enrollment patterns of different sexes. The finding of this study that the unemployment rate is positively related to male enrollment in both the US and the UK parallels the finding by Craig and Spear's earlier study, though with different methodological approaches. In the US, the effect of unemployment on male enrollment is unevenly distributed; it is greater for four-year than for two-year colleges, which mirrors the reward structure of American society since the risk of unemployment decreases as the level of education increases. In the UK, the effect of unemploymenton male enrollment is slightly higher for non-university (0.3528) than for higher education in general (0.3516). That may be a reflection that non-university institutions in the UK are more
17 vocationally oriented, and more responsive to the national economy and the job market. In contrast to the positive relation with male enrollment, the unemployment rates are negatively related to female enrollment of certain sections in both the US (at graduate level) and the UK (at the university level). That is to say, economic situations affect men and women differently - economic slumps tend to stimulate the growth of male enrollment but dampen that of female enrollment. Both human capital theory and conflict theory can be used to explain the positive effect of unemployment on the growth of educational enrollment, but not the negative effect, especially not the different effects on the divergent enrollment patterns between the two sexes. It is more revealing to compare the divergent enrollment patterns of men and women as the effects of different social forces. That is, female enrollment is significantly affected by the rising rate of divorce while male enrollment is significantly affected by unemployment. In appearance, the two patterns are seemingly unrelated. However, if we look at the phenomenon from the perspective of the traditional gender roles of men and women, we see clearly that in fact the enrollment of both men and women is responsive to the forces that undermine the traditional gender roles: men as breadwinners and women as homemakers. Male enrollment tends to increase when men's role or status as a breadwinner is in question (the increase of the unemployment rate), while female enrollment tends to increase when women's role as a homemaker is in question (the increase of the divorce rate). Female enrollment of certain sections in both the UK and the US also responded to the rise of the unemployment rate. In both cases, the increase of female enrollment is negatively related to the unemployment rate. On the one hand, it may simply reflect women's inferior financial situation. On the other hand, given the fact that women have for a long time been confined at home to be wives and mothers and their contribution to family income was usually marginal, if any, one may argue that the traditional gender role did not expect women to take the risk of investing in education especially at the time of an economic crisis. In contrast men as the principal family providers had to respond according to their traditional gender role expectation and find a way to support the family for either a short or long run. Therefore, a bad economic situation is likely to drive men into, but women away from, college due to different gender role expectation. The same explanation may extend to the positive relationship between the divorce rate and women's enrollment. The weakening of the traditional role of women as wives and mothers by the rapidly increasing marital breakdown is placing more and more women as head of the family and principle family provider, the role traditionally
18 assigned to men only. Women, facing the new reality or aware of such threat, are subsequently likely to take action appropriate to their new role. One would expect that since a higher unemployment rate means a larger number of people having lost a job and experiencing deprivation and uncertainty, it should generate a greater effect on student enrollment. However, this is not the case of the UK in comparison with the US. Although the UK has a higher unemployment rate for both men and women than that of the US, the unemployment effect in the UK on male enrollment is not as extensive as that in the US. Coupled with the non-response of the British female enrollment to the rise in divorce, some speculations come to mind. Halls (1985) once noticed that in the UK the participation rate of age groups in higher education in some regions was consistently higher than those of others. He wondered whether it was caused by cultural tendencies that value education differently. The same could be said about the different participation tendencies between the US and the UK. While there is no systematic study to prove whether Britishers or Americans are more inclined to go to college, there was evidence that Americans' fondness of college education went back to the "frontier" expansion period. The mentality had been such that many more colleges were founded than lower level schools could supply candidates, though a great number of these colleges were just college in name and closed down shortly thereafter (Rudolph 1962). Besides the cultural value of education, there may be an even more influential factor. That is, how education has played a role as a sorting/placing institution in a given society. Craig and Spear pointed out that the direction of change is not only affected by the actual or perceived deprivation and uncertainty, but also by options available to people who are seeking changes. If we can assume that the desire for higher education is closely related to the role education plays in social mobility, the desire for a higher degree would apparently depend as much on educational opportunity as on people's perception of the chances for mobility without a degree. While much attention has been paid to issues such as how different educational systems categorize people into different educational tracks, little attention has been paid to how credentials are related to placing people in different social status. One study of these issues concluded that Great Britain, among a few other nations in Europe, had a lower degree of credentialism and jobs less tied to degrees and qualifications (Muller et al. 1989). It is possible that the higher the value attached to a credential in society, the more people are likely to be motivated to make full use of educational opportunity available to them. In a comparative study of Japan and Western countries, Lynn lamented that young people in both the US and the UK were less motivated to go to college than their counterparts in Japan (Lynn 1988). According to Lynn, the situation in
19 the UK was even worse than that in the US. In spite of generous financial support from the government for college students, there were unfilled seats in British universities and polytechnics partly because not enough young people were inspired to do the academic work necessary for college admission. Do British young people have more alternatives other than attending a higher education institution? At least one study suggested that academically unsuccessful young people in the UK have more options than those in the US (Kerckhoff 1993). An earlier study of engineers in the US and Great Britain also indicated that formal education in Great Britain was less important in determining one's job, and experience was more important than education (Whalley 1984). With these findings in mind, it is reasonable to believe that the relative lack of enthusiasm of British young men and women for a higher qualification is partly because they have other options that can more or less meet their needs beyond going to college to get a degree. This article examines the gender-based patterns of higher education enrollment in the United Kingdom and the United States. The findings show that the enrollment of both men and women is responsive to the forces that undermine the traditional gender roles: men as breadwinners and women as homemakers. Male enrollment tends to increase when men's role or status as a breadwinner is in question (the increase of unemployment), while female enrollment tends to increase when women's role or status as a homemaker is in question (the increase of divorce). In comparing the UK with the US, it appears that while the US divorce rate is higher, it has taken the lead in providing educational opportunity. The UK has a more generous social welfare system and is more resistant to credentialism. The differences in these areas are considered as possible causes for the different enrollment patterns of the two countries.
Acknowledgements The writing of this report was supported by a grant from the Spencer Foundation Mentor Network under the supervision of Professor Gerald Grant of Syracuse University. It was Professor Grant's constant support and encouragement that has made this report possible. My debt to him can only be expressed, not repaid. This report is partly based on a larger research project, which received valuable advice and critiques from Professor Vincent Tinto and Sari Biklen of Syracuse University. I would like to express my gratitude to them. The author also wishes to thank the anonymous reviewer for his comments and suggestions on the report and its improvement.
20
No~s 1Throughout this report, terms like higher education, tertiary education, and college education, will be used interchangeably to indicate all postsecondary institutions. 2 There is no secondary-school diploma in the United Kingdom as in the United States. Compulsory education in Britain is from 5 to 16 years old. Those who are leaving school at age of 16 are called "school leavers", and can take either the GCE (General Certificate of Education) examination in individual subjects or the CSE (Certificate of Secondary Education) examination. School statistics has a category of "school leavers with qualifications", which excludes those whose exam scores are lower than C or no examination qualifications at all. In this study, "school leavers with qualifications" in the United Kingdom are taken as equivalent to high school graduates in the United States. 3 Data for this study is from the following publications: Central Statistical Office, Annual Abstract of Statistics. A Publication of Government Statistical Service, London: HMSO (Series); EC. (1990). Eurostat Demographic Statistics (Microfiche); National Center for Educational Statistics, (1990). Digest of Education Statistics. U.S. Department of Eudcation, Office of Educational Research and Improvement; National Center for Health Statistics, (1991). Vital Statistics of the United States, 1987. Vol. III Marriage and Divorce. DHHS Pub. No. (PHS) 91-1103. Washington DC: U.S. Government Printing Office; OECD Economic Outlook, Historical Statistics 1960-1980/1960-1989. Paris: OECD; UNESCO. Statistical Yearbook. Paris: UNESCO (Series); U.S. Bureau of the Census, StatisticalAbstracts of the United States. Washington, DC: Government Printing Office (Series). 4 The data is from UNESCO's Statistical Yearbook. Non-university institutions do not represent a unified or homogeneous group of higher educational institutions, but an umbrella category that includes diverged types of institutions outside university system at the tertiary level. In the United Kingdom, the non-university institutions mainly comprise institutions in ~ublic sectors. In the United States, it includes primarily two-year colleges. The data is from the statistical yearbooks of each individual country. Under the category of undergraduate students, the United Kingdom includes only those of university undergraduate students. But the United States includes those in both two-year colleges and four-year colleges/universities. Therefore, in this study, the components of undergraduates in the two countries are different.
References Agell, A. (1992). 'Grounds and procedures reviewed', in Weitzman, L.J. and Maclean, M. (eds.), Economic Consequences of Divorce The International Perspective. Oxford: Clarendon Press, pp. 53-66. America Council on Education (1987). Fact Book on Higher Education 1986-1987. New York: Macmillan. Aslanian, C.B. and Brickell, H.M. (1980). Americans in Transition. New York: College Entrance Examination Board. Astin, A.W., Green, K.C., Korn, W.S. and Maier, M.J. (1983). The American Freshman: National Norms for Fall 1983. Los Angeles: UCLA Higher Education Research Institute. Becker, G.S. (1964). Human Capital. New York: Columbia University Press for NBER. Benavot, A. (1989). 'Education, gender, and economic development: a cross-national study', Sociology of Education 62(1), 14-32. Blaug, M. (1970). Economics of Education. Harmondsworth: Penguin. Boudon, R. (1974). Education, Opportunity, and Social Inequality: Changing Prospects in Western Society. New York: Wiley. Bourner, T. with Reynolds, A., Hamed, M. and Barnett, R. (1991). Part-time Students and Their Experience of Higher Education. Buckingham (England): The Society for Research into Higher Education & Open University Press.
21 Clayton, D.E. and Smith, M.M. (1987). 'Motivational topology of reentry women', Adult Education Quarterly 37(2), 90-104. Clowes, D.A., Hinkle, D.E. and Smart, J.C. (1986). 'Enrollment patterns in postsecondary education, 1961-1982', Journal of Higher Education 57, 121-33. Collins, R. (1979). The Credential Society: A Historical Sociology of Education and Stratification. New York: Academic. Collins, R. (1971). 'Functional and conflict theories of educational stratification', American Sociological Review 36, 1002-1018. Cowan, R.S. (1983). More Workfor Mother: The Worries of Household Technologyfrom Open Hearth to the Microwave. New York: Basic Books. Craig, J.E. and Spear, N. (1982). 'Explaining educational expansion: an agenda for historical and comparative research', in Archer, M. (ed.), The Sociology of Educational Expansion, Take-off, Growth, and Inflation in Educational Systems. California: SAGE Studies, International Sociology, pp. 133-160. England, P. and Swoboda, D. (1988). 'The asymmetry of contemporary gender role change', Free Inquiry in Creative Sociology 16(2), 157-161. Feldman, S.D. (1974). Escape from the Doll's House. A Report Prepared for the Carnegie Commission on Higher Education. New York: Mcgraw-Hill Book Company. Fletcher, R. (1988). The Shaking of the Foundations. London: Routledge. Freeman, R.B. (1976). The OvereducatedAmerican. New York: Academic Press. Garfinkl, I. and Wong, P. (1990). 'Child support and public policy', in OECD, Lone-Parent Families. Paris: OECD. pp. 101-126. Gamier M.A. and Hage, J. (1991). 'Class, gender, and school expansion in France: a foursystems comparison', Sociology of Education 64(4), 229-250. Goode, W. (1993). World Changes in Divorce Patterns. New Haven and London: Yale University Press. Halls, W.D. (1985). 'A comparative political and sociological analysis of educational opportunity in Western Europe, 1960-1980', in Watson, K. and Wilson, R. (eds), Contemporary Issues in Comparative Education. London: Croom Helm. pp. 117-133. Haskey, J. (1984). 'Social class and socio-economic differentials in divorce in England and Wales', Population Studies 38, 419-38. Joshi, H. (1990). 'Obstacles and opportunities for lone parents as breadwinners', in OECD, Lone-Parent Families. Paris: OECD. pp. 127-150. Kamerman, S.B. and Kahn, A.J. (1991) (eds.). Child Care, Parental Leave, and the Under 3s. New York: Auburn House. Kerckhoff, A.C. (1993). Diverging Pathways. Cambridge: University Press. Kurian, G.T. (1991). The New Book of World Rankings. New York: Facts on File Publications. Lynn, R. (1988). EducationaIAchievement in Japan. Armonk: Sharpe, Inc. Lopata, H.Z. (1993). 'The interweave of public and private: women's challenge to American society', Journal of Marriage and the Family 55, 176-190. Measell, R.E (1992). 'The impact of coming from a single parent household on the labor force plans of college females: some economic and counseling implications', Journal of Divorce & Remarriage 18(1/2), 210-230. Meyer, J.W., Rubinson, R., Ramirez, EO. and Boli-Bennett, J. (1977). 'The world educational evolution 1950-1970', Sociology of Education 50(4), 242-258. Mincer, J. (1958). 'Investment in human capital and personal income distribution', Journal of Political Economy 66, 281-302. Muller, W., Luttinger, P., Konig, W. and Karle, W. (1989). 'Class and education in industrial nation', International Journal of Sociology 19(3), 3-39. Ramirez, F. and Boli-Bennett, J. (1982). 'Global patterns of educational institutionalization', in Altbach, P.G., Arnove R.F. and Kelly, G.P. (eds.), Comparative Education. New York: Macmillan Publishing Co., Inc. Ramirez, EO. and Rubinson, R. (1979). 'Creating members: the political incorporation and expansion of public education', in Meyer, J.W. and Hannan, M.T., National Develop-
22 ment and The World System: Educational, Economic, and Political Change, 1950-1970. Chicago: University of Chicago Press. Roger, C. (1991). 'Multinational corporate investment and women's participation in higher education in noncore nations', Sociology of Education 65(I), 37--47. Rudolph, E (1962). The American College and University. New York: Knopf. Schultz, T.W. (1961). 'Investment in human capital', America Economic Review 51, 1-17. Snyder, E (1992). The European Women's Almanac. New York: Columbia University Press. Turner, R.H. (1960). 'Sponsored and contest mobility and the school system', American Sociological Review 25(6), 855-867. Vess, J.D., Schwebel, A.I. and Moreland, J. (1983). 'The effects of early parental divorce on the sex role development of college students', Journal of Divorce 7, 83-95. Walters, EB. (1986). 'Sex and institutional differences in labor market effects on the expansion of higher education 1952 to 1980', Sociology of Education 59(4), 199-211. Weitzman, L.J. and Maclean, M. (1992). Economic Consequences of Divorce: The International Perspective. Oxford: Oxford University Press. Weitzman, L.J. (1985). The Divorce Revolution. New York: The Free Press. Whalley, P. (1984). Deskilling engineers? the labor process, labor markets, and labor segmentation, Social Problems 32, 117-132. Windolf, E (1992). 'Cycles of education in higher education 1870-1985: an international comparison', Higher Education 23, 3-19. Woodley, A., Wagner, L., Slowey, M., Hamilton, M. and Fulton, O. (1987). Choosing to Learn: Adults in Education. Philadelphia: The Society for Research into Higher Education & Open University Press.