Employment scarcity and racial differences in the decline of male marriage rate in the United States from 1980 to 2000
- (0) Download
https://www.eduzhai.net American Journal of Sociological Research 2016, 6(3): 74-90 DOI: 10.5923/j.sociology.20160603.03 Scarce Jobs and Racial Differences in the Decline of Marriage among Men in the United States, 1980-2000 Cedric Herring Language, Literacy, and Culture PhD Program, University of Maryland, Baltimore County, Maryland, USA Abstract Recent studies of male marriageability have placed the topic within the context of the dramatic structural changes that have occurred within the U.S. industrial economy. This study seeks to determine the relative impact of demographic and social factors on racial and ethnic differences in marital patterns for men between 1980 and 2000. It employs a ―population and structural change‖ perspective and data from the Current Population Surveys to assess the impact of the interaction of demographic and social variables on marital patterns for black, white, Latino and other males in the United States. The results show that during the 1980s, the proportions of unmarriageable males grew more rapidly among racial and ethnic groups with the lowest percentages of unmarriageable bachelors. Also, the proportion of college-educated black and Latino males who were unwilling to marry surpassed the proportion of college-educated white males who chose to remain single. The implications of the findings for patterns of marriage and family structure in the U.S. are discussed. Keywords Race, Employment Discrimination, Diversity, Marriage 1. Introduction One of the most significant and recurring themes in the area of racial and ethnic inequality is that of differential marital patterns by race. Several studies have searched for the link between male joblessness and marital status (e.g., Wilson and Neckerman, 1986; Sampson, 1987; Wilson, 1987; Henderson and Herring, 2013). Generally, these studies have focused on the differences in the pools of eligible males from which females choose prospective mates (Spanier and Glick 1980; Lichter, LeClere, and McLaughlin 1991; Henderson and Herring, 2013). They have noted that over the past several decades, the ratio of employed black men to the population of young women has declined significantly. The decline in the pool of ―marriageable‖ (i.e., employed) black males is the major reason for changes in the black family (Wilson 1987, 1996; and Henderson, 2009). More recently, studies have linked the issue of black male marriageability to changes in the economic options of black females, the growth of black female-headed households, and changes in the industrial economy (Wilson 1987; South 1991; Lichter, McLaughlin, Kephart and Landry 1992; Henderson, 2013). Still, social scientists have not adequately articulated how demographic and social variables interact over time to produce differences in marital patterns for various racial and ethnic * Corresponding author: firstname.lastname@example.org (Cedric Herring) Published online at https://www.eduzhai.net Copyright © 2016 Scientific & Academic Publishing. All Rights Reserved males. Moreover, they have not sufficiently explored the possibility that declines in marriage reflect declining preferences for marriage among eligible bachelors as well as decreases in the proportion of marriageable males because of a deteriorating labor market. This study seeks to determine the relative impact of demographic and social factors on racial and ethnic differences in marital patterns for men in the United States at the turn of the 21st Century. Specifically, the objectives of this study are to: (1) articulate a theoretical perspective that explains how demographic and social forces interact to produce racial and ethnic differentials in male marriageability; (2) compare patterns in marriage between blacks, whites, Latinos and other racial/ethnic groups; (3) examine the impact of preferences for bachelorhood on male marital patterns for all racial/ethnic groups; and (4) assess the change in these patterns for all men between 1980 and 2000. 2. Disadvantage, Demography, and Marriageability Several studies have examined the connection between male joblessness and marriage (e.g., Becker 1973, 1974, 1981; Wilson and Neckerman, 1986; Ellwood and Crane, 1991; Herring, 2013; Rockquemore and Henderson, 2015). Some of the leading works along these lines have suggested that the decline in the pool of marriageable black males is the main reason for changes in the black family structure. Becker (1981), for example, argues that marriage is attractive only if one party (traditionally the male) has a comparative American Journal of Sociological Research 2016, 6(3): 74-90 75 advantage in generating income from the labor market. When the relative availability of ―marriageable‖ (i.e., employed) men is in short supply, a larger percentage of marriageable men than women must marry for there to be balance in the marriage market. This has serious implications for marital dynamics and patterns. Any worsening of the labor market position of men would produce more jobless men. This in turn would tend to make men less attractive as potential spouses since their possible wives would stand to gain less from such couplings. At the same time, such scarcity of marriageable males would make those men with employment even more attractive, better positioned to take advantage of their comparative advantage in the labor market, and ironically less willing to marry without there being a greater share of the gains from the marital trade. In other words, worsening labor market conditions for men also make it less likely that marriageable bachelors will marry. The concept of marriageability was reintroduced to the debate on racial inequality by Wilson (1987). Wilson (1987) defined a marriageable male as one who is both single and employed. Linking black male unemployment to the transformation in the American industrial economy, he went on to note that the dearth of marriageable males was associated with the rise in female-headed households-particularly in the central city areas. Wilson and Neckerman (1986) argued that it is the decline in the pool of marriageable black males that is the main cause of changes in the black family. They showed that declines in the sex ratio for blacks mirrored declines in their marriage rates in the 1970s and the 1980s. Others using this basic framework have found support for this male marriageability hypothesis (e.g., Sampson, 1987; Testa, Astone, Krogh, and Neckerman, 1989; Bennett, Bloom, and Craig, 1989; Thomas, Herring, and Horton, 1994; Henderson, 2015a; Henderson, 2015b). Despite the significant advancement that the concept of marriageability represents, there are limitations in the manner in which it has been operationalized and implemented in research (Horton and Burgess 1992). Specifically, the concept does not adequately address the impact of underemployment (especially poverty wage employment) on racial differentials in marriageability (Lichter 1988). Moreover, implicit in the current usage of the term is the assumption that declines in marriage is specifically attributable to a rise in the number of blacks of lower socioeconomic status and in the central city areas (Wilson 1987). Research has not fully appreciated the possibility that there have been dramatic declines in marriage among those who are not poor and/or unemployed. Some of the decline in the proportion of married men may be attributable to increases in the number of men who prefer to stay single. Relatedly, the research and debates have focused almost exclusively on blacks and whites. Few researchers have paid attention to the implications of dramatic demographic changes in the Latino and other racial and ethnic minority communities for male marital patterns. Most importantly, the concept of marriageability ignores the broader demographic forces that have a bearing on male marital patterns. Racial differences in marriage patterns have been noted by demographers for some time (Bianchi and Farley 1979; Spanier and Glick 1980; Bianchi 1981; Farley and Allen 1987; Bennett, Bloom, and Craig 1989; Landale and Tolnay 1991; Qian and Preston 1993; Morgan, McDaniel, Miller, and Preston 1993; Ruggles 1994; Herring and Henderson, 2015; Herring and Henderson, 2016). Placed in the context of the general trends on the American family in the post-World War II era, there have been parallels and some convergent trends of marriage, divorce, and cohabitation (Glick 1988; Cherlin 1990; Herring, Horton, and Thomas, 1993; Herring, and Wilson-Sadberry, 1993). Bianchi (1981) noted the relationship between female-headed households and racial inequality. Cherlin (1990) cited the extended family network and cultural adaptation as partial explanations for the differences in black and white patterns of marriage and family formation. Landale and Tolnay (1991) argued that the system of stratification must be considered when examining differences in marital timing by race. One disadvantage of prior studies on black male marital patterns has been the absence of an explicit theoretical framework. The general trend has been to analyze Black male marital patterns in isolation from racial discrimination (Lichter et al. 1992) and apart from tastes for marital statuses. In addition, the linkage between Black male marriageability and the increasing marginalization of workers from other racial groups tends to be ignored (Lieberson 1980). Thus, detailed analyses of the demographic and social characteristics of the population of marriageable males of various racial groups are necessary. Such studies would add significantly to the understanding of the American family and the prospects for change in the future (Landale and Tolnay 1991; Lichter 1992; Herring, Henderson, and Horton, 2014). 3. The Population and Structural Change Thesis The population and structural change thesis maintains that changes in the relative size of minority populations interact with changes in the social structure to exacerbate racial/ethnic inequality. This perspective acknowledges structural arrangements that have relegated larger segments of the black and Latino communities to joblessness. However, it also argues that changes in the social and demographic characteristics of the population interact with changes in the social structure. For example, in the past five decades, African Americans and other racial and ethnic minorities have steadily increased their levels of educational attainment. The educational attainment gap between whites and blacks dwindled to less than a year by 1990. In addition, there has been some movement of racial minorities into nontraditional jobs. But these changes have occurred in the context of a deteriorating 76 Cedric Herring: Scarce Jobs and Racial Differences in the Decline of Marriage among Men in the United States, 1980-2000 labor market. Simply put, as the overall opportunity structure declines, dominant groups attempt to maintain their relatively privileged positions in society by eliminating subordinate groups from competition (Blumer 1958; Bonacich 1972; Smith 1981; Herring, and Henderson, 2012). Moreover, the competitive threat to dominant groups is not simply a function of a change in the size of the subordinate population. Conflict occurs when that segment of the subordinate population increases which is most likely to compete with the dominant population (Bonacich 1976; Lieberson 1980). In the United States, dominant and subordinate group relations have historically taken the form of racial and ethnic conflict (van den Berghe 1967; Wilson 1973; Herring, 2006). In the case of black-white relations, changes in the social structure, in conjunction with changes in the population of labor force participants, led to the wholesale disenfranchisement of blacks from the post-Reconstruction Era to the mid-1960s (Lieberson 1980; Franklin and Moss 1988). Certainly, changes in population size have been linked to inequality, labor force participation and intergroup conflict in past studies (Blumer 1958; Blau 1977; Olzak 1996; Tomaskovic-Devey and Roscigno, 1996). Moreover, the association between social dislocation, black male marriageability and female-headed households has been documented (Wilson 1987). However, none of these studies has focused upon how change in demographic and structural factors interact to produce differences in black-white male marriageability. 3.1. Race and Population Change in the United States The story of population change in the United States over the last two decades has been the phenomenal increase in the various racial and ethnic minority populations (U.S. Bureau of the Census 1993). The most dramatic increases have occurred in the Asian American and Hispanic populations (O'Hare 1992). Over the 1980-1992 period, the former has increased by 123.5% and the latter by 65.3%. In both instances, the increases were primarily due to relatively high rates of immigration (O'Hare 1992). These subpopulations are doing more than contributing to the increase or changing racial and ethnic composition in the United States. They are also bringing into question the social definition of race in this society (Massey 1994; Waters 1994; Horton, Allen, Herring, and Thomas, 2000; Herring, 2005). The black population has also contributed to this increase in the size and diversity of the overall minority population (U.S. Bureau of the Census 1993). Having a growth rate of 16.4%, their increase has been three times that of whites between 1980-1992 (Butler and Herring, 1991; O'Hare 1992). However, the history and sociology of the black experience adds a different dimension to the overall issue of race and population change. Despite gains since the Civil Rights Movement of the 1960s, there is still a disproportionate number of blacks who are disadvantaged (Wilson 1987; Herring, House, and Mero, 1991). Moreover, blacks still experience levels of discrimination that exceed those of other groups— irrespective of class (Lieberson 1980; Feagin 1991; Massey and Denton 1993; Herring, 2009). Hence, an overall increase in the minority population is likely to have a disproportionately negative effect on the black population as various groups compete for existing opportunities. 3.2. Race and Structural Change Equally as important as the increase in racial and ethnic diversity in the United States is the structural context within which diverse groups compete (Olzak 1992). The transformation of the American industrial economy has resulted in the displacement of many American workers. Hit hardest have been those labor-intensive, manufacturing industries and the blue collar workers that they employed (Bluestone and Harrison 1982; Herring and Johnson, 1990; Horton, Thomas, and Herring, 1995; Semyonov and Herring, 2007). Placed in the context of increasing levels of education and job readiness among racial minorities, racial and ethnic conflict is likely to be exacerbated. As jobs become increasingly more difficult to obtain, irrespective of race and education, gains experienced by blacks and other racial minorities are perceived as coming at the expense of white workers (Gans 1988; Bobo and Kluegal 1993). Wilson (1987) focuses upon the effects of the restructuring of the American industrial economy on blue-collar African American workers. But the population and structural change thesis suggests the likelihood that increased labor force discrimination will also be experienced by middle-class blacks and other middle class people of color. There are several considerations that lend support to this hypothesis. First, the increase in the black middle class population means that for the first time in the history of the United States, white middle class workers will have to compete with blacks for professional job opportunities. Second, the restructuring of the American economy is also impacting upon white-collar workers, as companies downsize to meet the new economic realities. In short, the black middle-class is growing at the time when the opportunity structure for middle class people is in a state of decline. The population and structural change thesis also provides a cogent explanation of differential marriageability for different racial and ethnic groups: Labor market dynamics are not guided by pure rationality and perfect competition; rather, a number of concrete processes operate systematically to generate dissimilar employment for whites, blacks, Latinos and others. For example, segmented labor markets offer qualitatively different jobs (Cain, 1976; Beck et al., 1978; Dickens, and Lang, 1988; and Lichter, 1988). Positions in the primary labor market offer relatively low unemployment rates, high pay, good benefits, job security, and the possibility for on-the-job advancement; jobs in the secondary labor market offer relatively high unemployment rates, low pay, poor benefits, seasonal or contingent employment, and little upward mobility on the job. Blacks, American Journal of Sociological Research 2016, 6(3): 74-90 77 Latinos, and other racial and ethnic minority groups are under-represented in the primary labor market and over-represented in the secondary labor market. In addition, normative beliefs about ―minority jobs‖ and ―white jobs‖ (Neckerman and Kirschenman, 1991), employer reliance on ―soft skills‖ rather than tangible ones (Kirschenman, 1992), informal recruitment networks (Braddock and McPartland, 1987), employers' ―tastes for discrimination,‖ (Portes and Sassen-Koob, 1987; and Neckerman and Kirschenman, 1991) and exclusionary practices by labor unions and professional associations (Johnson and Oliver, 1992) all act to steer racial minorities into less stable, racially typed jobs in the secondary labor market. But as racial minorities have acquired credentials that make them more competitive with white males for positions in the primary labor market, the competition for more desirable jobs in a declining labor market has become stiffer. All of these factors contribute to lower overall levels of marriageability for men. Nevertheless, there is a tendency for the same credentials for minorities to result in less access to professional jobs than for whites (Shelton, 1985; and Landry, 1987). Thus, smaller proportions of college-educated minorities occupy positions of authority. So, fewer minorities are able to influence decisions concerning hiring, retention, and promotion of subordinates. Consequently, minorities at different career stages lack the objective sponsorship of their white counterparts. In addition, racial minorities are more likely to be dismissed from their jobs, especially early in their careers (Johnson and Herring, 1989). These structural barriers lead to even lower levels of marriageability for men of color. 3.3. Theoretical Implications for Male Marriage Patterns Population and structural change provides the context within which to assess differences in marriage patterns among various racial groups over time. These differences are expected to be manifested over various social and demographic categories: age, region, urbanicity, education, and occupation. Age is expected to be a major determinant of marital status. However, the pattern here is expected to be differentiated by race. Contrary to arguments that suggest that discrimination is primarily a historical legacy (Wilson 1980), the population and structural change thesis argues that younger minority males will have lower levels of marriageability relative to comparable whites. This relationship is not expected to vary significantly over time, and if anything, will be more pronounced in 1990 than in 1980. Regional variations in male marriageability by race are expected because of racial and ethnic differences in population distributions (O'Hare 1992) and employment opportunities (Lichter 1989). Previous studies have documented more directly the relationship between race, male marriageability, and region (Wilson 1987; Horton and Burgess 1992). In the current study, change in the gap between blacks and whites in marriageability is expected to be greater in the northeast and midwest because these two regions have experienced the brunt of the economic restructuring during the last decades. (Herring and Fasenfest, 1996). The disproportionate percentage of racial minorities found in the central cities makes urbanicity a key variable in this analysis. Past studies have documented the effects of the outmigration of middle class Blacks from the inner city areas (Wilson 1987; Jaynes and Williams 1989). Thus, it would follow that Black male marriageability would be lower in the central city areas than outside them. Moreover, it would be expected that Blacks outside of the central cities would have levels of marriageability that would be more comparable to white males. Yet, the population and structural change thesis leads one to hypothesize a different relationship between race, urban residence and male marriageability: It is precisely outside of the central cities where blacks are most likely to be in competition with whites for existing employment opportunities. Thus, the differential in marriageability between the two groups is argued to be greater outside of the inner city. Education could be expected to be the most important variable relative to racial differentials in marriageability. A male is more likely to be marriageable (i.e., employable) if he is better educated. To the degree that race, per se, has become less important in determining employment prospects minority males with higher educational attainment would be expected to have levels of marriageability that are equal to comparably educated whites (Wilson 1980). On the other hand, the population and structural change thesis would predict a different result. It predicts that the racial differentials in male marriageability will be greater among those with higher educational attainment than among minorities and whites with similarly low levels of educational attainment. To reiterate, it is in the more highly educated and skilled minority males who are most likely to be competing with whites for the most prized jobs in the labor force. In the midst of economic restructuring, it is in this highest educational category where blacks and other racial minorities are expected to have lost the most ground relative to whites in marriageability. Similar findings are expected when examining the racial differentials in marriageability by occupation. Minorities in the higher occupational categories are expected to have lower levels of marriageability than their white counterparts. In addition, these categories are expected to have experienced greater increases in the marriageability differential over time. 3.4. Statement of Hypotheses H1. Blacks and Latinos are significantly less likely than whites to be married and more likely to be unmarriageable, net the effects of demographic and social variables. H2. (a) The proportion of married men decreased over the 78 Cedric Herring: Scarce Jobs and Racial Differences in the Decline of Marriage among Men in the United States, 1980-2000 1980-1990 period. (b) Among racial minorities, these declines in marriage were more likely than those of whites to be due to increases in the number of unmarriageable bachelors. (c) Among whites, declines in marriage were more likely than those of minorities to be due to increasing preferences for bachelorhood. H3. Blacks and Latinos have higher levels of unmarriageability than whites in every region of the country. H4. Black and Latino marriageability will be less than that for whites within every age category. H5. Blacks and Latinos do not have parity with whites in marriageability within categories of education. H6. College-educated black and Latino males are relatively more likely to unwilling bachelors than are their college-educated white counterparts. 4. Data and Methods laborers; and others. (g) Decade was dummy coded to indicate whether the respondent was interviewed in 1980 or 1990. (h) Percentage Black in a community was coded to reflect the proportion of a Standard Metropolitan Statistical Area (SMSA) that was African American, non-Hispanic. (i) Percentage Hispanic in a community was coded to reflect the proportion of an SMSA that was Hispanic, non-Black and non-white. (j) Ratio of marriageable Black bachelors to single Black females in a community was coded to reflect the number of unmarried Black men (between the ages of 18 and 54) with jobs paying wages over the poverty threshold divided by the number of unmarried Black women (between the ages 18 and 54). (k) Ratio of marriageable Hispanic bachelors to single Hispanic females in a community was coded to reflect the number of unmarried Hispanic men (between the ages of 18 and 54) with jobs paying wages over the poverty threshold divided by the number of unmarried Hispanic women (between the ages 18 and 54). The data for this study come from the 1980 and 1990 Current Population Surveys. The study includes only male respondents between the ages of 18 and 54. The sample size is 64,169. The dependent variable, marital situation, is operationalized as a trichotomy: husband, marriageable (i.e., marriageable but ―unwilling to marry‖) bachelor, and unmarriageable (i.e., ―unable to marry‖) bachelor. Currently married men were dummy coded to indicate that they were husbands. Men who were not currently married but who were employed with income above the poverty level for the given year were dummy coded as marriageable bachelors. All others (unmarried and unemployed or with incomes below the poverty level) were dummy coded as unmarriageable bachelors. It is useful to introduce at this point language that will facilitate a clear understanding of the analysis of the dependent variable, marital situation. From this point on, marriageable bachelors will be referred to as ―the unwilling‖ and unmarriageable bachelors will be referred to as the ―unable.‖ The third category, ―husband‖ will remain as is. It will function as the reference category of the dependent variable. The independent variables of the study are as follows: (a) Race was dummy variable coded as white, black, Latino, and others. (b) Age was coded in years, but was collapsed into categories of 18-24, 25-34, 35-44 and 45-54 for tabular presentations. (c) Region was dummy variable coded into four categories: Northeast, Midwest, South, and West. (d) Urbanicity of residence was dichotomized into central (inner) city and non-central city. (e) Education coded as years of attainment (or its equivalent), but was collapsed into a trichotomy (of ―less than high school,‖ ―high school only,‖ and ―at least some college‖) for tabular presentation. (f) Occupation was dummy variable coded into professional, technical and managerial; administrative support and sales; craftspersons; service workers; operators, fabricators and 5. Results Are blacks and Latinos less likely than whites to be married and more likely to be unmarriageable bachelors? Did the proportion of married men decrease between the 1980s and the 1990s? Are college-educated blacks and Latinos more likely than their white counterparts to be unwilling bachelors? Below we provide some preliminary answers to these questions. Figure 1 presents the distribution of male marriageability by race and ethnicity for 1980 and 1990. It shows that for 1980 and 1990 combined, 18% of white men were unwilling (i.e., marriageable) bachelors. This compares with 17% of black males, 14% of Latino males, and 16% of other males. This figure also shows that 21% of white males were unable (i.e., unmarriageable) bachelors, compared with 38% of black males, 29% of Latino males, and 28% of other males. We also see that 61% of white males were husbands, compared with 45% of black males, 57% of Latino males, and 56% of other males. These differences provide a chi square statistic of 1094.5 with six degrees of freedom, and they are statistically significant at p<.01. Figure 1 also shows that the proportion of husbands decreased for all racial and ethnic groups between 1980 and 1990. Among whites, the decline was from 62% to 60%. For Blacks, the decrease was from 48% to 43%. For Latinos, the decline was from 60% to 54%. And for others, the drop was from 56.1% to 55.7%. Among whites, the decrease in the proportion of males who were husbands can be accounted for completely by a corresponding increase in the proportion of the unwilling bachelors. Among blacks, Latinos, and others, however, the decline in the proportion of men who were husbands is attributable, at least in part, to increases in the proportion of unable bachelors. American Journal of Sociological Research 2016, 6(3): 74-90 79 Figure 1 Table 1 presents the relationships among marriageability and race and ethnicity by various sociodemographic variables and decade. These results show that racial and ethnic patterns in marriageability vary by sociodemographic subgroups. This table shows some general patterns: with increasing age, the proportion of husbands increases for each racial and ethnic group, and there were general increases in the proportion of unable bachelors between 1980 and 1990. For example, among those younger than age 25, racial and ethnic differences in marriageability became smaller between 1980 and 1990--not because of a decrease in the proportion of unable bachelors for those groups with the highest proportions of unmarriageable males. Rather, the proportions of unable bachelors grew more rapidly among those ethnic groups with the lowest percentages of them in 1980. In 1980, 59% of young whites, 78% of young Blacks, 61% of young Latinos, and 72% of young others were unable bachelors. By 1990, these percentages had climbed to 68%, 83%, 73%, and 80% respectively. Table 1 also shows that among men between the ages of 25 and 35, the proportion of husbands decreased for every racial and ethnic group between 1980 and 1990. This decline in husbands showed up as increases in the proportion of unable bachelors. While the same basic patterns held true for men between the ages of 35 and 44, the gap between black men who were husbands and all other racial and ethnic groups became greater between 1980 and 1990. During the 1980s, the proportion of men between the ages of 45 and 55 who were unable bachelors grew for whites, blacks, and others. For blacks between ages 45 and 55, however, the increase in the percentage of unable bachelors was greater than for all other racial and ethnic groups. Consistent with the predictions of the population and structural change thesis, blacks and Latinos at every age category were more likely than similar whites to be unable bachelors. This table also shows that there was a general pattern of decreases in the proportion of married men of all races in three of the four regions. For whites, these declines in marriage were accompanied by parallel increases in the percentage of marriageable bachelors. For blacks, this pattern held true only in the northeast. In the midwest and the south, declines in marriage among black males were mostly attributable to increases in the percentage of unable bachelors. Among Latinos, declines in marriage were mostly due to increases in the proportion of unablebachelors in each of the regions. Again, these patterns are in keeping with the expectations of the population and structural change thesis. The table indicates that the percentage of married men in urban settings increased between 1980 and 1990. Meanwhile, there was little change in the proportion of urban men who were unable bachelors, but there was an increase in the rate of unwilling bachelors in urban settings. In contrast, there was a decrease in the percentage of married men outside of urban settings and a corresponding increase in the proportion of men who were unable. While these patterns cut across racial lines, they were consistent with the population and structural change idea that black-white differences would be greater outside of inner city areas. This held true in both 1980 80 Cedric Herring: Scarce Jobs and Racial Differences in the Decline of Marriage among Men in the United States, 1980-2000 and 1990. In 1980, 31% of those without at least a high school degree were unable bachelors. By 1990, this had increased to over 45% of those without high school diplomas. While there were racial differences in the proportion of those without diplomas who were unable, the magnitude of increase among them was similar for all racial groups. Among those with high school diplomas, however, there were racial differences in the patterns. In particular, there was a small decrease in the percentage of high school-educated whites who were unable bachelors, but there was an increase in the unable among those from other racial groups with comparable levels of education. Among those with at least some college education, the changes in marriageability were similar for the various racial groups, as there were small declines in the percentage of the unable for all racial groups and slight increases in the percentage of unwilling for whites, blacks, and Latinos. While the racial difference in the percentage of unwilling bachelors did not increase with level of education, the gap between the percentage of blacks and whites who were unable bachelors did increase with levels of education. Once more, this corresponds to what the population and structural change thesis argues about the racially differentiated effects of education on access to adequate employment and subsequent marital situation. The results presented in Table 1 do not take into account how these factors and others simultaneously affect the relationship between male marriageability and race. Tables 2-4 present more rigorous evidence from multinomial logit analysis that examines this relationship net of other variables such as age, region, urbanicity, education, occupational type, the percentage of the population that is black and Latino, and the ratio of black men to women and the ratio of Latino men to women. The logit (logistic probability unit) model--a special case of the general log-linear model--is appropriate when the dependent variable can take on only limited values, and thus, violates the Ordinary Least Squares (OLS) regression model assumptions that the variables will be continuous and measured on an interval scale. It assumes that the underlying probabilities are logistic, i.e., in the form: F(p) = 1/(1 + e-P) = ln(P/1 -P) where p is the probability of the occurrence of an event and e, an irrational number, is the base of natural logarithms such that ln(ex) = X and the antilog of X is ex . The logit is the logarithm of the odds of success; i.e., the ratio of the probability of the occurrence of an event to the probability of nonoccurrence of that event. The function confines the value of (p) between 0 and 1. When the odds of success are even (.5) the logit (coefficient) is zero; when they are greater than even, the logit has a positive value; and when they are less than even, its value is negative. Column one of each table presents the log odds of being a marriageable (unwilling) bachelor versus being a husband. Column two of each table presents the log odds of being an unmarriageable (unable) bachelor versus being a husband. Unfortunately, the signs of logit coefficients are not sufficient for determining the direction and magnitude of change of corresponding probabilities in polytomous logit models. For this reason, column three of Tables 2-4 presents the odds ratios of being unable versus unwilling. These odds ratios are calculated as follows: antilog(2*CoefficientUnmarriageable Bachelor) Odds Ratio = _____________________________________________ antilog(2*CoefficientMarriageable Bachelor) Ratios of 1.0 indicate that being unwilling (i.e., a marriageable bachelor) and unable (i.e., an unmarriageable bachelor) are equally probable; ratios greater than 1.0 mean that being unable is more likely; and ratios less than 1.0 suggest that the odds of being unwilling are greater. While both the probabilities of being unwilling (Column 1) and an unable (Column 2) may increase (or decrease) in comparison to being a husband, it is their relative odds (Column 3) that is the focus of much of this analysis. Therefore, when both coefficients are in the same direction, comparisons will be made between being unwilling and being unable. Table 2 shows that, net of the correlates of marital status, there were still significant differences in marriageability. Column 1 shows that, net of other factors, blacks were more likely than whites to be unwilling bachelors than husbands, (p < .01), but Column 2 indicates that they were also more likely than whites to be unable bachelors than husbands (p < .01). Column 3 discloses that, net of all other variables, black men were substantially more likely than whites to be unable rather than unwilling bachelors. Net of other factors, Latinos were less likely than were whites to be bachelors than to be married. They were less likely than comparable whites to be unable bachelors and unwilling ones. But they were slightly more likely to be unwilling than unable, net of the other factors. Members of the other racial groups were more likely than whites to be unable bachelors than husbands. Members of the other racial groups were also more likely than were whites to be unable bachelors rather than unwilling ones. The race by decade interaction terms indicate that a higher proportion of blacks than whites became unable bachelors (rather than husbands) during the decade. The interactions also indicate that a lower proportion of Latinos than whites became unwilling bachelors (rather than husbands) during the decade. Other results of note include the finding that unwilling bachelors became more common in the 1990s (versus the 1980s). In addition, during the 1990s, college graduates became less likely to be unwilling bachelors, but more likely to be unable bachelors. During the decade, higher proportions of men with professional positions became unwilling bachelors and fewer of them became unable ones; lower proportions of those with administrative positions, however, became unwilling bachelors, and more of them became unable bachelors. American Journal of Sociological Research 2016, 6(3): 74-90 81 Table 1. Changes in the Distribution of Male Marital Situations by Race and Various Sociodemographic Variables, 1980, 1990, and 2000 82 Cedric Herring: Scarce Jobs and Racial Differences in the Decline of Marriage among Men in the United States, 1980-2000 Table 2. Multinomial Logit Model Predicting the Tendency to be a Marriageable Bachelor or Unmarriageable Bachelor Versus a Husband, Net of Race/Ethnicity and Other Attributesa Net of these structural and compositional changes, there were other factors associated with marital condition. As expected, the tendency to be a husband increased with age, and the tendency to be an unwilling bachelor rather than an unable one also increased with age. Those living in the Northeast, the Midwest, and the South were less likely than those living in the West to be unwilling bachelors, and higher proportions of them were unable rather than unwilling bachelors. Men living in urban settings were more likely to be bachelors than to be husbands. With higher levels of education, men were more likely to be unwilling bachelors than husbands, less likely to be unable bachelors than husbands, and more likely to be unwilling than unable bachelors. Men with professional and administrative positions, craft employment, and those who were employed as laborers were less likely than others to be unable bachelors than husbands. Men who lived in communities with higher concentrations of black and Latino residents were more American Journal of Sociological Research 2016, 6(3): 74-90 83 likely to be bachelors. And finally, men who lived in areas that had higher ratios of marriageable Latino bachelors to Latino females were less likely to be unwilling bachelors than husbands. Generally, these results support the claim of the population and structural change thesis that declines in marriage reflect declining preferences for marriage among the unwilling as well as decreases in the proportion thereof because of a deteriorating labor market. There are changes in marital patterns among men that are associated with changes in education, occupational distribution, racial composition, and sex ratios. These results are not inconsistent with the idea that declines in marriage are in part due to the increasing preference for bachelorhood among men of all racial and ethnic groups who are marriageable bachelors. Table 3 presents similar results for African Americans, and Table 4 presents results for Latinos. As was the case in the overall analysis, bachelorhood became more common among black men in the 1990s. Generally, higher proportions of black males became unwilling bachelors rather than unable ones. During the 1990s, Black men with professional or administrative positions, while already more likely than others to be unwilling bachelors and less likely to be husbands, became even more likely to be unwilling bachelors during the decade. The results also show that African Americans who lived in communities with higher concentrations of black residents were more likely to be bachelors. Similarly, those who lived in areas where the ratio of the black unwilling bachelors to black females was high were still more likely to be unwilling bachelors. These findings underscore the tendency for some men, especially those with options, to choose to remain unmarried. Table 3. Multinomial Logit Model Predicting the Tendency to be a Marriageable Bachelor or Unmarriageable Bachelor Versus a Husband, Among African Americans, Net of Other Attributesa
... pages left unread,continue reading
Free reading is over, click to pay to read the rest ... pages
0 dollars，0 people have bought.
Reading is over. You can download the document and read it offline
0people have downloaded it