Social Mobility from a Gender Inequality
Perspective
A Thesis
Submitted as Partial Fulfilment of the Requirements for the Master of Science Degree in Economics
By Ntorina Kale Course code: EBM877A20 Supervisor: G. (Giampaolo) Lecce, Ph. d
Faculty of Economics and Business University of Groningen
ABSTRACT
Few studies have examined gender inequality in social mobility content. The purpose of this quantitative research was to explore if females are experiencing more or less social mobility pattern than males in order to gain a deeper understanding of the gender inequality in social mobility. Moreover, we looked for the factors that correlate with the mobility difference between the two genders. Using data from Chetty et al. (2014) on relative and absolute mobility of each gender and commuting zone characteristics, the findings revealed an important mean difference on the mobility between females and males. The strongest correlated factors on the mobility gap of the two genders are race, segregation, tax, education and social capital.
Table of Contents
Abstract ... 2
List of Tables and Figures ... 4
Chapter 1 ... 5 1.1 Introduction ... 5 1.2 Research Objectives ... 7 1.3 Research Questions ... 7 1.4 Hypotheses ... 7 Chapter 2 ... 8 2.1 Literature Review ... 8 2.1.1 Gender Inequality ... 8 2.1.2 Social Mobility ... 11
2.1.3 Relationship between Gender Inequality and Social Mobility ... 13
2.1.4 Social Mobility and Economic Results ... 16
Chapter 3 ... 18
3.1 Data and Methodology ... 18
3.2 Results ... 24
Discussion ... 32
Conclusion ... 38
References ... 39
Appendix 2 ... 45 Appendix 3 ... 47
List of Tables and Figures
CHAPTER 1
1.1 Introduction
Intergenerational income persistence in economics impacts the perpetuation and aggravation of the resource gap between the wealthy and poor. Research on social mobility in previous years has mainly concentrated on men. However, there is less or no research on the patterns of social mobility of women counterparts and the factors that associate with the mobility gap of the two genders.
The level of intergenerational mobility in society is perceived as a measure of the degree of equality of economic opportunity. It captures the degree whereby individual situations during childhood lead to success in older years along with the extent at which individuals can use of their own talents and luck. Measuring intergenerational mobility involves understanding how the social and economic status of the married men and women determined the status of their children and comprehending its drivers. This is of supreme importance to get rid of barriers to equal opportunities and ensure a level playing platform in access to education and jobs.
From an economic point of view, social mobility is important because it is a chance for individuals from underprivileged backgrounds to break the borders of the social class. Social mobility is the movement of people and households or families between or within strata in a society. As Chetty et al. (2014) contend, social mobility can vary across geographical areas in the United States. The drivers of these differences in areas are
The fact that women were not a part of the workforce and social classes were represented by the class positions of the husbands explain why women were excluded from mobility studies (Li & Singelmann, 1998). Women are neglected in the labour force because they are perceived to have a much weaker attachment to it, and in the long run, they are employed in lower-class positions. The degree at which the mobility patterns of both genders vary biases the conclusions regarding the fluidity of a society. This thesis offers a research on the social mobility patterns of males-females and the factors that can correlate with the mobility gap of the two genders. The goal is to explain which of the two genders is more mobile and what are the factors that relate with the differences in the mobility patterns of the two sexes.
By using data from Chetty et al. (2014), we came up with a significant mean difference of the mobility patterns of the two genders to the advantage of females. Women have developed their knowledge, skills, and creativity to an excessive degree. Hence, they achieved to get a high relative pay and seize the equal opportunities (Mulligan & Rubinstein, 2008). Moreover, we found that the factors that correlate with the mobility gap of males-females are race, segregation, tax, education, and social capital.
1.2 Research Objectives
The long-term goal of the research is to understand the impact of gender differences on social mobility patterns and the factors that associate with the mobility gap of the two genders. Social mobility is the change in the position of individuals from one status to another. The objective of the present study is to offer a comprehensive literature review for the relationship between gender differences and social mobility, the study constitutes the subsequent sub-objectives:
To understand if the social mobility pattern of females is greater than that of males. To understand the factors that associate with the mobility gap between males-females.
The outcome of this study will be vital to industry practitioners to understand the best tools to be used to avoid gender inequalities in a world that is characterized by social stratification and mobility.
1.3 Research Questions
The following research questions need to be addressed:
Are females experiencing more or less relative/absolute mobility than males overall? Why do some commuting zones of the U.S. show higher differences of relative mobility (absolute mobility respectively) between the two genders than others?
1.4 Hypotheses
𝐻1 There is a statistically significant mean difference between the relative/absolute mobility of the two genders.
CHAPTER 2
2.1 Literature Review
2.1.1 Gender Inequality
Gender is one of the main dimensions of inequality and stratification in American society. It intersects in many and complex ways with such dimensions of inequality and stratification as social position and social mobility (Blau & Kahn, 2017). It is problematic to imagine any element of society that is not in some perspective ‘gendered’.
Gender comprises of the psychologically and socially norms for male and female sanctioned by and expected in a society. The expectations of the roles of gender differ across the cultures and at distinguishable moments in society and within microcultures in a similar society. There exist biases due to culture which results in disparities in the workstation, particularly to the disadvantage of women. The women are affected by irrespective of their level of education, age, and social background. Even though women workers have justified being as diligent in their duties as men, they are ordinarily judged more strictly thereby appear not to be competent. The inequality is obvious in pay as women receive less income compared to men in nearly all the occupations (Abrams, 2012).
The United States was the first developed nation that started supporting equal opportunities at the workplace regardless gender by applying anti-harassment policies and laws (Blau & Kahn, 1996). Lots of economists wondered what the effect of these
(2006), where women started entering into the labour market and seeking for equal
opportunities with their gender counterpart, neglecting their role at home (Aguiar & Hurst, 2007). The more knowledge, skills and creativity women have the more likely it is for them to confront less market access barriers. These limited percentage of high-skilled women manage to get a high relative pay and seize the equal opportunities and higher social stratifications (Mulligan & Rubinstein, 2008).
England (2010) contended that the persistence of cultural devaluation of women’s jobs had weakened the trend towards gender equality. In the American context, the author demonstrates that women abandoned the traditional women jobs and shifted to men’s jobs only when that was the only trajectory of upward mobility present to them. Men had no reimbursements to leave treasured men’s work for devalued women’s work under any circumstances. The elite women have shifted to traditional male jobs that only a few years ago were denied to them in significant quantities.
Regardless of the notably increasing level of education of women, we persist in having gender differences in the labour market results (Goldin, 2006). These differences may be interpreted mainly due to discriminatory behaviour at the workplace and sexes’
inequivalent roles and responsibilities (Goldin & Rouse, 2000). Gramlich (2017) discussed the gender issues that impacted the United States in 2017. He argued that sexual harassment was an issue of gender that affected the United States for a long time. More than 22% of employed women have been sexually assaulted in their workstations, women who encounter discrimination due to gender at their workplace stood at 42%. According to Becker (1971), discrimination tends to take place mostly in non-competitive areas of the economy. He reported that companies that have low discriminatory behaviour and employ low cost female workers face low production costs compared to ones that have high prejudicial behaviour and result in getting out of the business.
Economists continue to question why labour market disparities last and argue if gender differences in competition can give more information to that. If men are more willing to compete than women, then it seems that women are less doubtlessly to ask for promotions or shift to men’s job. In a competitive workplace, the level of performance between the two genders shows to get even larger than it is in a non-competitive one. More specifically, a male indicates to perform greater to an augmentation in rivalry than a high-skilled female who inclines to stay out of competitions (Niederle & Vesterlund, 2011). Psychologists came out with an explanation and stated that the abstention of women from a competitive
environment can be due to the lack of confidence on their skills relatively to men (Soll & Klayman, 2004).
transformations and concerted movements to challenge the subordination of women. The persistence of gender disparity in the economic, legal and political processes that work against it recommends that there are continuous social processes that reconstruct gender inequality.
2.1.2 Social Mobility
Economists differ from sociologists when giving preference to explore mobility between the levels in the distribution of income instead of social strata conceptualized. Social mobility is the most significant notion in a class system of stratification. In class system, social stratification grounded on both birth and personal attainment; an individual merit becomes more significant (Lee & Solon, 2009). As societies become more meritocratic and competitive, elements including social skills, ambition, energy and physical attractiveness and luck were instrumental in social mobility and altering social position (Behrman, 2000). Arguably, social mobility highlights the alterations in social position which happens during the lifetime of an individual.
The move can be upward or downward. Upward social mobility is an alteration in the social status of an individual leading to promoting the individual to a higher position in the system. On the other hand, downward mobility leads to the individual attaining a lower position in the status system (Andrew & Leigh, 2009). Intragenerational mobility and intergenerational mobility highlight the two ways to explore social mobility.
Economists have tried to define social mobility from an economic point of view. Before starting analysing intergenerational mobility, Chetty et al. (2014) separates social mobility into relative and absolute. It is taken into account as relative mobility when the present generation occupies a greater socioeconomic status comparative to the earlier generation. While, absolute mobility measures the downward or upward movement of the children’s earnings comparative to their parents in real terms. The coexisting studies on social mobility might be classified based on the unit of analysis including family, group, and the domain.
The latest significant trend in the literature on intergenerational mobility explores the correlation between the indicators of social mobility and a diversity of significant aggregate outcomes. Agreement in this facet is far from being attained. Corak (2013) and Chetty et al. (2014) discover that social mobility varies across the geographical regions and that it moves positively with economic activity, social and human capital and negatively with equality. In all societies, there is a low mobility and constant across all societies and time, and therefore not related to aggregate variables.
Many attempts have taken place in recent years to compute the variations of intergenerational mobility. The United States is characterized as a high-income nation providing equal opportunities, giving the freedom to achieve what you desire and having the chance to experience an upward social mobility. Hoping that if not the currently
impoverished families to achieve it then at least the upcoming generations. However, the findings seem to contradict the American dream of a promising land for an upward movement (Corak, 2013). The OECD (2011a, p.40) has concluded that countries like the United States and the United Kingdom with high pay gap may prevent upward social
invest in a higher-quality human capital for their children and the economy will still benefit the top 1 percent. On the other hand, Lee et al. (2009) suggests that, by using a better structured data approach for groups born from 1952 to 1975, intergenerational mobility has not altered substantially, despite the fact that income gap has grown during this time.
Other studies have been conducted with regards to the relationship between social mobility and a country’s degree of inequality. There is a common belief that the low social mobility and strong association of income among generations mean the disruption of equal chance standards. Accordingly, the people of countries with high social mobility can tolerate inequality easier. Burtless and Jencks (2003) reveals that social mobility will drop when the difference between low-income parents and income parents expands because high-income parents can afford high quality education for their children. In another research turns out that sons that are raised in more discriminated countries in the 1970s were hardly
probable to undergo social mobility by 1999.Thus, in the most discriminated nations, it is quite challenging and almost impossible to rise from rags to riches (Andrews & Leigh, 2009).
2.1.3 Relationship between Gender Inequality and Social Mobility
Limited progress has been made to understand how gender differences affect the social mobility. A growing body of literature on gender disparity in mobility has started to emerge in the latest years. Hayes and Miller (1993) argued that the convectional exemption of women from mobility research has met strong criticism from feminists and has caused worries among the social stratification researchers regarding the restrained understanding of the process of social mobility because of the omission of women.
have continued to be assigned low-level positions. The Economic Mobility Project (2014), conducted by the Pew Charitable Trusts, examines the mobility of the Americans in
comparison with their parents. It finds that even though daughters’ salary is greater than that of their mothers, the hourly pay gap between daughters and fathers is still important due to gender, payment, work hours and labour sector. Additional to the findings, non married daughters in full employment provide 81 percent on their family earnings and are related with upward mobility in the family income ladder.
In contrast, men have been awarded high-status positions in the occupational hierarchy (Hayes & Miller, 1993). Notably, McGinn & Oh (2017) argue that in the
workplace, cultural and occupational conditions differ by social class. Gender is instrumental in the employment of women when the latter are in the majority in a work setting. Chetty et al. (2014), examined the estimates of intergenerational mobility and argued that there is substantial variation in absolute and relative intergenerational mobility in the social context across countries. Due to differences in social class in different regions, the relative mobility was found to be lowest in children that grow up in the southwest side of the United States and highest in the rural Midwest and Mountain West.
When working in professions where men are the majority, women will face gender-oriented bias (Turco, 2010), and the latest evidence recommends this bias might be higher for upper class comparative to the middle-class women (Rivera & Tilcsik, 2016). The
amplification of gender bias might increase the recognition of upper-class women as female, whereas potentially distorting with class-oriented recognition. The female managers and professionals benefit because of the increase in the presence of women in leadership positions (Lee at al., 2017), though dependence on the minority of leaders who are female might
increase recognition with gender and reduce identification with class.
McGinn & Oh (2017) argues that in earlier research has demonstrated relatively consistent proof when absolute rates of social mobility attained via occupational status are assessed; there is a weaker inclination for daughters than sons to inherit the occupational position or social class of their fathers. In addition, women were discovered to be much more probable downwardly mobile than men. According to Corak (2013), higher rate of inequality promotes lower mobility and ruins opportunities. Mobility changes incentives, opportunities and institutions that develop and transmit skills and characteristics that are most valued in the labour market. Mobility shifts the balance of power to position some groups to structure policies. Individuals who are more concerned with equal opportunities should consider taking care of outcomes of inequality in gender. Parents can transmit opportunities to their children in the form of economic advantage via social connections that facilitate access to capital sources and jobs. In high-income countries such as United States and United Kingdom, their values and demographic diversity implies that it is not desirable to change the intensity of mobility (Marginson, 2016).
peasantry was deteriorating though some of its members were shifting into industry or low-skilled services. Children who came from social class backgrounds in which the parents had professional jobs enjoyed the most conducive conditions to ultimately achieve a high occupational status themselves due to the socialization. Simultaneously, the progressive expansion of the professional class allowed the entrance of the new members particularly the children of service workers garnered good grades at the academic institution. Thus, education is improving social mobility and generating privilege and solidarity. These patterns were observed from 1970 to 1990.
2.1.5 Social Mobility and Economic Results
The vast literature assessing intergenerational mobility perceives that more mobility is a desirable characteristic of the economy and society in general. Andrews et al. (2009), report that any kind of inequality can have consequences on economic growth, health, and political behaviour. However, as a result of the renowned challenges in producing the reliable measures of intergenerational mobility, we still have less knowledge regarding how it relates with significant social and economic results.
correlations. Clark et al. (2014) review the relation among generations of the mean outcomes of persons having the similar surname contending that mobility does not differ in most of these societies and thus is not linked with the economic performance.
A recent research was conducted to shed light on the relation of social mobility and economic outcomes. Güell et al. (2018), compute intergenerational mobility across different geographical regions of Italy and analyse its interaction with economic and social variables. Italy is a centralized country, where institutional set up is similar for all regions, policies and institutions are impossible to be the primary factors of geographical differences in social mobility. They discovered that intergenerational mobility is greater in areas where there is high economic activity, educational achievement, social capital and low inequality.
Furthermore, intergenerational mobility has a positive effect on desirable economic results and a negative on undesirable economic results. However, there are not any
CHAPTER 3
3.1 Data and Methodology
The Opportunity Insights is a non-profit research institution that is headed by
economists from the University of Harvard and the University of Brown. It provides big data on economic mobility for the purposes of mending economic opportunity and inducing policy makers to enable underprivileged young people and families to climb up the economic ladder as well as reach better life standards. Various projects have been conducted using advanced scientific research methods to investigate the issues that are of importance. Due to the existence of ‘big data’, the main factors of upward economic mobility and poverty in the United States have been discovered.
Specifically, the paper “Where is the Land of Opportunity? The Geography of Intergenerational Mobility in the United States” by Chetty et al. (2014) gives us information on the absolute and relative intergenerational mobility of both genders for core sample (1980-82 birth cohorts) by commuting zones and the commuting zone characteristics. These
characteristics are considered crucial for this research as they will be used to investigate the correlation with the difference in relative/absolute mobility between the two genders. Consequently, the data of the paper of Chetty et al. (2014) were selected as the most appropriate to assess the relationship between gender inequality and social mobility.
geographical areas in the United States. As a final point, they explore the features associated with upward mobility such as segregation, income gap, elementary education, social capital, family balance, et cetera. The difference between the two papers is that Chetty et al. (2014) are investigating what factors correlate with upward mobility, whereas this research focuses on what factors associate with the mobility gap of the two genders.
The data that the authors use to calculate absolute and relative mobility are from federal income tax records across 1996-2012. The data comprise income tax return and third-party statistics returns that provide details on the income of those who do not record tax returns. The dataset of children involves all individuals who have legit Social Security Number or Individual Taxpayer Identification Number, were born during the period 1980-1991, and are U.S. civilians of 2013. They preclude individuals who have immigrated to the United States at an adult age since their parents’ income cannot be captured. Each child is matched with a parent (or parents), despite any possible changes in parents’ family status.
The sample, which this research will be based on and is termed as the core sample in the paper of Chetty et al. (2014), consists of all children in the base dataset who are born in the period 1980-1982, whose identification of the parent is able, and whose mean parent earnings for the period 1996-2000 is positive. The measure of children’s income as mean family income takes place at the age of 30 (in 2011-2012).
computation is the elasticity of child earnings with regards to parent earnings ( 𝑑𝐸[𝑙𝑜𝑔𝑌𝑖|𝑋𝑖=𝑥] 𝑑𝑙𝑜𝑔𝑥 ), known as the intergenerational income elasticity (IGE). The usual approach of measuring the IGE is by regressing log child earning (𝑙𝑜𝑔𝑌𝑖) on log parent earnings (𝑙𝑜𝑔𝑋𝑖), with coefficient
𝐼𝐺𝐸 = 𝜌𝑥𝑦𝑆𝐷(𝑙𝑜𝑔𝑌𝑖) 𝑆𝐷(𝑙𝑜𝑔𝑋𝑖)
with 𝜌𝑥𝑦 = 𝐶𝑜𝑟𝑟(𝑙𝑜𝑔𝑋𝑖, 𝑙𝑜𝑔𝑌𝑖) the association between log child earnings and parent
earnings and SD() is the standard deviation. The IGE constitutes the relative mobility because it captures the disparity in (log) results between children of lower and higher earnings
parents.
The second approach that Chetty et al. (2014) is using to estimate relative mobility is by correlating the ranks of child and parents (Dalh et al., 2008). Consider as 𝑅𝑖 the child 𝑖′𝑠 percent rank in the earnings distribution of children and 𝑃𝑖 the parent 𝑖′𝑠 percent rank in the earnings distribution of parents. By regressing child rank 𝑅𝑖 and parents rank 𝑃𝑖 it generates the coefficient 𝜌𝑃𝑅= 𝐶𝑜𝑟𝑟(𝑃𝑖, 𝑅𝑖), which is referred as the rank-rank slope. The rank-rank slopes estimate the correlation between child’s status in the earnings distribution and his parents’ status in the distribution.
Absolute Mobility (AM) refers to children’s propensity to have higher earnings than their parents in actual terms. One of the three measures of the absolute mobility that the authors are using and will be used in this paper is the absolute upward mobility (AUM). According to Chetty et al. (2014), the AUM is “the mean rank (in the national child income distribution) of children whose parents are at the 25th percentile of the national parent income distribution” (p.7). Nationally, this measure is closely associated with the rank-rank slope but at local level a child’s rank in the national earnings distribution is efficiently an absolute result because results in a certain region have small effect on the national distribution.
The data needed to complete the project would be absolute and relative mobility for core sample (1980-82 birth cohorts) of both genders by commuting zones within the U.S. and commuting zone characteristics. The data are collected on these variables to perform
statistical analysis. There are 741 commuting zones (CZs) of the United States in the dataset with more than 250 children in the relevant samples which are used for performing the adequate analyses. In overall, each CZs includes 4 countries with population 380,000.
Regarding the commuting zone (CZ) characteristics, they are CZ-level data on each of the variables that will be used in the second part of the research objective. These variables represent different factors that are linked with the sociology and economics literature, like segregation, income inequality, tax, education, college, local labour market, migration, social capital, and family structure. The data of these variables can be found from the 2000 Census and other publicly accessible datasets. A short description of them can be found in the figure 1 (Chetty et al., 2014).
Research Objective 1
The paired t-test is a descriptive approach to determine if females are generally more or less relative/absolute mobile than males. The dependent sample t-test, as it is called as well, is a statistical method applied to investigate whether the mean difference between two groups of observations is zero. In other words, it tests if the difference between the mean RM of males and the mean of RM of females is equal to 0 (and respectively the mean of AM of males and the mean of AM of females is equal to 0). We do the t-test between the variables RM males and RM females and between the variables AM males and AM females.
Research objective 2
“Why do some commuting zones of the U.S. show higher differences of RM (AM respectively) between the two genders than others?”
The multiple OLS regression analysis is chosen to be performed to determine the correlation of the difference between RM of males and RM of females (and between AM of males and AM of females respectively) with commuting zone characteristics. From the basic multiple regression model, the study designed two regressions:
𝑌𝑖 = 𝛽0+ 𝛽1𝑋𝑖1+ 𝛽2𝑋𝑖2 +. . . + 𝛽𝑝𝑋𝑖𝑝+ 𝜀𝑖
Besides that, it is essential to mention some important specifications regarding the regression analysis. Firstly, we normalize all the dependent and independent variables before running the analyses. Hence, regressions carried out on standardized variables produce standardized coefficients. Standardized coefficients give the effect size of independent
variables on our independent variable given that all variables are measured in the same scale. Secondly, state fixed effects variables are included to limit selection bias across states. The two initial regression analyses are clustered at the state level in order to examine the results of the linear regressions in the level of accuracy.
Next, we run the regression models based on robust standard errors. Also, another robustness check is conducted by removing the extreme 5% (lowest and highest) of
relative/absolute mobility difference and running the regression (p5-p95) RM/AM difference on all the set of explanatory variables as presented in Table 3 based on robust standard error. Despite that, including all the set of proxies for each category undermines the
multicollinearity problem. Therefore, we run the regression RM/AM difference between the two genders on the explanatory variables that are statistically significant in our main
regression model based on robust standard error.
3.2 Results
Results of Research Objective 1
Table 1. Paired t test : RMmales RMfemales
obs Mean1 Mean2 dif St_Err t_value p_value RMmales -
RMfemale~
665 .321 .335 -.013 .002 -7.85 0
Table 2. Paired t test : AMmales AMfemales
obs Mean1 Mean2 dif St_Err t_value p_value AMmales -
AMfemale~
665 42.8 44.717 -1.917 .077 -24.8 0
Two paired t-tests were run on 665 CZs with at least 250 children in the core sample to define whether there is a statistically substantial mean difference between the RM (AM respectively) of the two genders. Although, there are 76 missing observations due to absent values for the relative/absolute mobility. As shown in Table 1, there is a significant difference in the RM for males (M=0.32, SD=0.064) and females (M=0.33, SD=0.068); t (664) = -7.86, p=0.00. Since, the p-values of the mean difference of RM (males-females) is 0.000 which is smaller than alpha value of 0.01, we have enough evidence to reject the null hypothesis H0. By observing the RM mean of the two genders, we conclude that females are more mobile than males by 0.013.
males by 1.91. This implies that females show higher propensity, compared to males, to have higher income than their parents in real terms.
Results of Research Objective 2
Table 3. Correlates of the RM/AM difference between the two genders across CZs-clustered
CollegeGrad_R~v 0.090 -0.024 (0.083) (0.069) LaborForce_Pa~v -0.104 -0.017 (0.105) (0.087) Manufacturing~v 0.182 0.001 (0.124) (0.092) Growth_Chines~v -0.113*** 0.000 (0.039) (0.027) Teenage_Labou~v 0.066 -0.189 (0.134) (0.149) Migration_Inf~v 0.071 -0.078 (0.117) (0.085) Migration_Out~v -0.220* -0.014 (0.131) (0.084) Fraction_Fore~v -0.077 -0.016 (0.097) (0.075) SocialCapital~v -0.470** -0.039 (0.191) (0.099) Fraction_Reli~v 0.149 -0.027 (0.113) (0.078) Crime_R_sv -0.272** 0.227** (0.112) (0.107) Children_Sing~v -0.492 0.232 (0.295) (0.217) Fraction_Adul~v 0.275* -0.121 (0.145) (0.109) Fraction_Adul~v -0.009 0.088 (0.149) (0.125) IncomeGrowth_sv -0.089 0.215** (0.125) (0.082) _cons -0.550** -0.111 (0.254) (0.324) Obs. 414 414 R-squared 0.290 0.580 statedummy YES statedummy YES
Standard errors are in parenthesis
*** p<0.01, ** p<0.05, * p<0.1
Notes: OLS regression results of RM/AM difference between the two genders on all the set of
explanatory variables as pictured in Figure 1 clustered at the state level. Dependent and
independent variables are standardized to have mean 0 and standard deviation 1 in the assessment sample.
and CZ-level characteristics. We included all the characteristics in the regression analyses due to the fact that they reflect proxies for broader factors (race, segregation, income distribution, tax, education, college, local labour market, migration, social capital, family structure, and income growth). There are 327 missing observations due to absent values for the relative/absolute mobility and local area characteristics.
The main focus is on the regression models with standard errors clustered at the state level to analyse the relationship between independent and dependent variables. The
coefficient of determination (R-Squared) of the diffRM/diffAM cluster model is 0.290 and 0.580 accordingly. Accordingly, the 29% of the variation in diffRM is explained by the variation of independent variables and the 58% of the variation in diffAM is described by the variation of independent variables. Apparently, the data fit better on the regression model of diffAM than diffRM.
The most statistically significant factors that correlate with the mobility gap of the two genders are race, segregation, tax, education, local labour market, social capital, and income growth. While, family structure, migration, income inequality, and college are weakly correlated with the difference in the mobility of males-females. A more detailed analysis of the results is offered below:
Race The major measure of the factor is the fraction of black. The correlation coefficient of fraction black on the RM difference between the two genders is 0.550. The significance level (p<0.05) in coefficient estimate indicates that this is an important result since there is enough evidence to reject the null hypothesis 𝐻0. On the other hand, the
Segregation Various measures of segregation had been estimated to evaluate the impact that may have on the mobility difference of the two genders. Firstly, the racial segregation estimate on the differences in the RM of females and AM of males-females is close to 0. Secondly, the income segregation estimate has an effect on the difference in RM of males-females equal to 0.266 and on the difference in the AM of the genders equivalent to -0.791. More specifically, segregation of poverty has a weak impact on the mobility difference between the two genders, whereas segregation of affluence differs slightly. The last measure of segregation is the fraction of individuals who commute to work at less than 15 mins in the CZ. The correlation of the fraction of commute less than 15 mins with the difference in the RM of males-females is -0.0653, whereas with the AM gap of males-females is statistically significant 0.456.
Tax This factor is measured in four different ways: local tax rate, local government expenditures per capital, tax progressivity, and state EITC exposure. The tax progressivity has a highly positive significant correlation with the RM difference equal to 0.484 and negative one with the AM difference equal to -0.273. Moreover, the local tax rate and state EITC exposure are strong associate of relative mobility gap with estimates of 0.210 and -0.329 and absolute mobility gap with coefficients equal to 0.281 and 0.245 respectively. The last measure, local government expenditures, is uncorrelated with the mobility gap of males-females.
Education The correlation between school quality and mobility gap of the two genders is critical. The first and most important measure is focusing on school expenditures per student. The statistically significant association between school expenditures and relative mobility gap between the two genders is scored 0.456. However, the impact of school
mobility between the two genders. The rest of the measures of school quality are uncorrelated with the mobility gap between the two genders.
Local Labour Market The growth in Chinese imports, which is one of the fourth measures (labour force participation, share working in manufacturing, and teenage labour participation) of local labour market, has a statistically substantial negative relationship with the relative mobility gap of males-females equal to -0.113. The rest of the measures do not correlate with the dependent variables.
Social Capital The social capital index, the fraction of religious, and the violent crime rate are considered the some of the measures of social capital. The social capital index, developed by Rupasingha and Goetz (2008), has statistically important effect on relative mobility gap between the two genders equivalent to -0.470. Interestingly, the magnitude of the correlation between the violent crime rate and the difference in RM (AM) of males-females is -0.272 (0.227) with significance level less than 5%. Lastly, the association between the fraction of religious people and relative mobility gap is 0.149 in respect of -0.027 absolute mobility gap.
Income Growth The effect of income growth, whose calculation was done as the yearly growth rate implied by the variation in income over the 8-year period, on the relative mobility gap is 0.215. The significance level of the correlation coefficient is less than 5% (p<0.05).
absolute mobility gap (-0.121). Yet, the fraction of married adults does not play an important role in the mobility gap of the two genders.
Migration The migration outflow rate, which is the migration out of a commuting zone from other commuting zones, is negatively associated with the difference in the relative mobility of males-females (-0.220). Regarding the other two estimates of migration, the migration inflow rate and the fraction of foreign born, they do not play any important role in the mobility gap of males-females.
Income Inequality The mean household income level of a CZ and the
(relative/absolute) mobility gap of the two genders correlate positively (0.229/0.0155). Another measure of income inequality, Gini, which measures the income distribution within each CZ, is an omitted variable in the regression models because of collinearity with the other two measures of income inequality. However, the next measure of inequality is the top 1% of Earners which has a weaker impact on our dependent variables compared to the bottom 99% of Earners. Within the bottom 99% income shares belongs the middle class. The correlation between the size of the middle class and the mobility gap is close to 0. In total, the effect of different measures of income inequality on relative mobility gap of the two genders is weakly positive, whereas on absolute mobility gap varies.
College We define the factor college as the number of colleges, the college tuition, and the college graduation rate. All these measures show that college is uncorrelated with the relative or absolute mobility gap between males and females. Thus, the correlation
coefficients are close to 0.
identical results. Furthermore, another robustness check was carried out by precluding the extreme 5% (lowest and highest) of relative/absolute mobility difference. We investigated the effect of all explanatory variables on the (p5-p95) RM/AM difference.
A summary of the robustness check results can be seen in Appendix 2. The factors which have a statistically significant correlations with the (p5-p95) mobility difference of males-females are race (fraction of black), segregation (fraction of commute<15mins), income inequality (Gini bottom 99%), tax (tax progressivity, state EITC exposure), education (school expenditures, test score), local labour market (teenage labour force participation), social capital (violent crime rate), family structure (fraction of divorced adults) and income growth. Consequently, the robustness check results were quite close with the correlates findings based on standard errors clustered at a state level.
Last but not least, the incorporation of all the set of proxies for each category undercut the multicollinearity problem. Consequently, the regression analysis between RM/AM
difference of males-females and the explanatory variables that are statistically significant in Table 3 was conducted based on robust standard error (Appendix 3). Notably, the statistically insignificant CZ characteristics precluded were racial segregation, income segregation, segregation of poverty<p25, segregation of affluence>p75, household income, top 1% income share, Gini bottom 99%, fraction middle class, local government expenditures, student teacher ratio, test score, high school dropout rate, number of colleges, college graduation rate, labour force participation, share working in manufacturing, teenage labour force participation, migration inflow rate,fraction foreign born, fraction religious, fraction single mothers, fraction of married adults. Even though factor college did not have
Discussion
In this study, we examined the relationship between gender inequality and social mobility to gain a greater understanding of how the social mobility patterns of females and males vary and what factors associate with the mobility gap of the two genders. At first point, women were excluded from mobility studies due to the fact that they were not a part of the workforce and social classes were represented by the class positions of the husbands (Li & Singelmann, 1998). Over times, women started to integrate into the job market attempting equal opportunities with men (Goldin, 2006). The United States applied anti-harassment policies and laws to encourage equal chances for both genders (Blau & Kahn, 1996).
Even though there is less or no extended research on the social mobility of the two genders, our findings are quite encouraging and provide us information on which of the two genders is more mobile. In overall, there is a significant mean difference between the (relative/absolute) mobility patterns of males-females. The mean social (relative/absolute) mobility of females exceeds the mean social (relative/absolute) mobility of males. In other words, males show a lower tendency than females to occupy a higher position in the income distribution compared to their families (relative mobility). Also, males in contrast to females have lower inclination for getting higher income than their parents in real terms (absolute mobility).
declare that even though there has been a shrink of income inequality between the two genders, women are the ones that get the lowest paying jobs (Blau & Kahn 2017, Barron et al. 1993, Mincer & Polachek 1974).
Despite that, we investigated various factors that correlate with the social mobility gap of the two genders across commuting zones. In general, few studies have focused on the factors that associate with the social mobility, and especially with the mobility gap of the two genders. Chetty et al. (2014) found that social mobility is mostly associated with five
categories: segregation, income inequality, education, social capital, and family structure. In addition, Güell et al. (2018) confirmed that intergenerational mobility is higher in regions where there is high economic activity, educational achievement, social capital, and low inequality.
Some of these categories are consistent with our findings such as education, economic activity, social capital, and segregation. In particular, the factors that were identified as the biggest predictors of the mobility gap of males-females are race, segregation, tax, education, local labour market, social capital, and income growth. However, the other categories (family structure, migration, income inequality, and college) have weaker associations and will not be discussed extensively.
Based on the results, the proxies that have the greatest correlation with the mobility difference of the two genders in each category: fraction of black (race), fraction of
Race The fraction of black correlate positively with the RM difference of the two genders. This means that the RM gap is larger in areas with high number of individuals categorized as black. Consequently, an increase of 1 standard deviation (SD) in African-American populations, the difference in the relative mobility of the two genders rises by 0.550 SD. On the other hand, the AM difference between the two sexes is smaller in cities with large African-American populations by -0.360. To sum up, regardless of a person’s ethnicity living in black neighbourhoods correlates with the mobility gap of the two genders. Some reasons for these correlations can be the differences in the institutions or industries that are built in those areas.
Segregation The most statistically important form of segregation that associate with the relative mobility gap is the fraction of individuals who commute to work at less than 15 mins in the CZ. Regions with short commute significantly increase the relative mobility difference of males-females (0.456). Overall, the positive effect of segregation can make the relative mobility gap of the two genders more amplified.
Tax This factor is identified to be the most substantial predictor of the mobility gap between the two genders. The tax progressivity associate positively with the relative mobility gap and negatively with the absolute mobility gap. Additionally, the local tax rate and state EITC exposure are defined as the second significant associate of mobility gap. If the local tax rate and state EITC exposure rises by 1 SD, the difference in RM is less by 0.210 SD and -0.329 SD respectively and the difference in AM expands by 0.281 SD and 0.245 SD accordingly.
absolute mobility gap of the two genders decreases by 0.269 SD. All in all, the direction of the correlation between school spending and mobility gap is undefinable.
Local Labour Market The growth in Chinese imports does affect significantly only the relative mobility difference of females. The gap in the relative mobility of males-females is going to drop by 0.133 SD if the growth in Chinese imports increases by 1 SD. Social Capital The social capital index and the violent crime rate correlate negative with the relative mobility gap. This means that if social capital index/crime rate increases by 1 SD, the relative mobility gap will decrease by 0.470/0.272 respectively. In contrast, the rate of violent crime is highly positive correlated with the difference of absolute mobility.
Income Growth The effect of income growth on the relative mobility gap is
significantly positive. If the income growth in a CZ increases by 1 SD, the relative mobility gap expands by 0.215 SD.
The robustness checks presented in Appendices 1 and 2 demonstrate that the
findings of Table 3 are almost robust. The estimates of the regression models based on robust standard errors (Appendix 1) are similar with the Table 3. Race, segregation, tax, education, local labour market, social capital, and income growth are still the most statistically
significant correlates factors of the mobility difference between the two genders. On the other hand, the estimates of the regression analyses of (p5-p95) RM/AM gap on all the explanatory variables (Appendix 2) differ slightly compared to Table 3. Specifically, removing the
difference in the mobility patterns of males-females while income inequality, family structure and local labour market weakly.
However, the fact that we involved all the proxies from each category in the multiple regression models (Table 3) undermines the problem of multicollinearity. Some of the
independent variables in each category may be highly related such as Gini, Gini bottom 99%, and top 1% income share. Multicollinearity affects mostly the prediction of each variable, but it does not affect the reliability of the model as a whole. To rearrange the multicollinearity problem and compare the results with the outcomes of Table 3, we run the multiple regression models by abstracting the statistically insignificant independent variables from each category (Appendix 3). The category ‘college’ has no effect on the mobility gap of the two genders in the initial regression models (Table 3) and instead of excluding all its proxies we include only college tuition to get its impact on the independent variables.
The exclusion of statistically insignificant CZ characteristics helps us manage to reorganise the multicollinearity problem. Multicollinearity affects only the independent variables that are associated. It can change substantially the precision of the coefficient estimates and p-values can not be trusted. For example, we place in the models (Appendix 3) only the Gini variable from the ‘income inequality’ category and we see that this time the effect on the mobility gap of the two genders is statistically important.
are the ones that seem to correlate dominantly with the mobility patterns gap of the two genders either in the initial multiple regression analysis (Table 3) or in the robustness check tables (Appendices 1 & 2) or in the multicollinearity rearrangement (Appendix 3).
Conclusion
References
Abrams, J. A. (2012). Blurring the lines of traditional gender roles: beliefs of African American women (dissertation).
Aguiar, M., & Hurst, E. (2007). Measuring Trends in Leisure: The Allocation of Time over Five Decades. The Quarterly Journal of Economics, 122(3), 969-1006. Retrieved April 15, 2020, from www.jstor.org/stable/25098866
Andrews, D., & Leigh, A. (2009). More inequality, less social mobility. Applied Economics Letters, 16(15), 1489-1492. doi:10.1080/13504850701720197
Barron, J. M., Black D. A., and Loewenstein M. A., (1993). “Gender Differences in Training, Capital, and Wages.” Journal of Human Resources 28 (2): 343–64
Becker, Gary S. (1971). The Economics of Discrimination, Second edition. Chicago and London: University of Chicago Press, 1957.
Behrman, J. (2000). “Social Mobility: Concepts and Measurement in Latin America and the Caribbean.” In: N. Birdsall and C. Graham, editors. New Markets, New Opportunities? Economic and Social Mobility in a Changing World. Washington, DC, United States: Brookings Institution/Carnegie Endowment for International Peace.
Bertaux, D. (1977). Destins personnels et structure de classe : Pour une critique de l'anthroponomie politique (1. éd ed., Politiques). Paris: Puf.
Blau, F. D., and Kahn L. M., (2008). “Women’s Work and Wages.” In The New Palgrave Dictionary of Economics, Second edition, edited by Steven N. Durlauf and Lawrence E. Blume, 762–72. London: Palgrave Macmillan.
Blau, F., & Kahn, L. (1996). Wage Structure and Gender Earnings Differentials: An International Comparison. Economica, 63(250), new series, S29-S62. doi:10.2307/2554808
Burtless G. and Jencks C., (2003). American inequality and its consequences, in Agenda for the Nation. H. Aaron, J. M. Lindsay, and P. S. Nivola (eds), Brookings, Washington DC, pp. 61–108.
Chetty, R., Hendren, N., Kline, P., & Saez, E. (2014). Where is the Land Opportunity? The Geography of Intergenerational Mobility in the United States. The Quarterly Journal of Economics, 129(4), 1553-1624. doi:10.2307/26372582
Clark, G. with N. Cummins and Y. Hao and D. Diaz Vidal (2014). The Son Also Rises:
Surnames and the History of Social Mobility, Princeton, NJ: Princeton University Press. Corak M., (2013). "Income Inequality, Equality of Opportunity, and Intergenerational
Mobility." Journal of Economic Perspectives, 27 (3): 79-102.
Dahl, Molly W. and Thomas DeLeire. 2008. The association between children’s earnings and fathers’ lifetime earnings: estimates using administrative data, University of Wisconsin-Madison, Institute for Research on Poverty.
Economic Mobility Project. (2014). “Women’s Work: The Economic Mobility of Women Across a Generation”. Washington DC: Pew Charitable Trusts.
https://www.pewtrusts.org/en/research-and-analysis/reports/2014/04/01/womens-work-the-economic-mobility-of-women-across-a-generation
England, P. (2010). The Gender Revolution: Uneven and Stalled. Gender and Society, 24(2), 149-166. Retrieved April 15, 2020, from www.jstor.org/stable/27809263
Goldin C. and Rouse C., (2000). “Orchestrating Impartiality: The Impact of ‘Blind’ Auditions on Female Musicians.” American Economic Review 90 (4): 715–41.
Gramlich J. (2017, December 28). Key findings on gender equality and discrimination in the U.S. Pew Research Center. https://www.pewresearch.org/fact-tank/2017/12/28/10-things-we-learned-about-gender-issues-in-the-u-s-in-2017/
Güell, M., Pellizzari, M., Pica, G., & Rodríguez Mora, J. (2018). Correlating social mobility and economic outcomes. The Economic Journal, 128(612), 403. doi:10.1111/ecoj.12599 Hayes, B., & Miller, R. (1993). The Silenced Voice: Female Social Mobility Patterns with
Particular Reference to the British Isles. The British Journal of Sociology, 44(4), 653-672. doi:10.2307/591415
Kline, P., & Moretti, E. (2014). People, places, and public policy: Some simple welfare
economics of local economic development programs. Annual Review of Economics, 6, 629-662.
Kurtulus, F. (2012). Affirmative action and the occupational advancement of minorities and women during 1973-2003. Industrial Relations: A Journal of Economy and
Society, 51(2), 213-246. doi:10.1111/j.1468-232X.2012.00675.x
Lee, C., & Solon, G. (2009). TRENDS IN INTERGENERATIONAL INCOME
MOBILITY. The Review of Economics and Statistics, 91(4), 766-772. Retrieved April 15, 2020, from www.jstor.org/stable/25651375
Lee, H., Chua, C., Miska, C., & Stahl, G. (2017). Looking out or looking up: Gender differences in expatriate turnover intentions. Cross Cultural and Strategic Management, 24(2), 288-309. doi:10.1108/CCSM-02-2016-0046
Marginson, S. (2016). The worldwide trend to high participation higher education: Dynamics of social stratification in inclusive systems. Higher Education, 72(4), 413-434. Retrieved April 15, 2020, from www.jstor.org/stable/26447553
McGinn, K., & Oh, E. (2017). Gender, social class, and women's employment. Current Opinion in Psychology, 18, 84-88. doi:10.1016/j.copsyc.2017.07.012
Mincer J. and Polachek S., (1974). “Family Investments in Human Capital: Earnings of Women.” Journal of Political Economy 82 (2 Part 2): S76–108.
Mulligan, C., & Rubinstein, Y. (2008). Selection, Investment, and Women's Relative Wages over Time. The Quarterly Journal of Economics, 123(3), 1061-1110. Retrieved April 15, 2020, from www.jstor.org/stable/25098924
Niederle, M., & Vesterlund, L. (2011). Gender and Competition. Annual Review of Economics, 3, 601-630. Retrieved April 15, 2020, from www.jstor.org/stable/42940201
OECD. (2011a). Divided We Stand: Why Inequality Keeps Rising. Organization for Economic Cooperation and Development.
Rivera, L., & Tilcsik, A. (2016). Class Advantage, Commitment Penalty: The Gendered Effect of Social Class Signals in an Elite Labor Market. American Sociological Review, 81(6), 1097-1131. Retrieved April 15, 2020, from www.jstor.org/stable/44245496
Rupasingha, Anil and Stephan J. Goetz. 2008. “US County-Level Social Capital Data, 19902005.” The Northeast Regional Center for Rural Development, Penn State University, University Park, PA.
Soll, J., & Klayman, J. (2004). Overconfidence in interval estimates. Journal of Experimental Psychology. Learning, Memory, and Cognition, 30(2), 299-314.
Appendices
Appendix 1
College_Tuiti~v -0.064 -0.016 (0.058) (0.046) CollegeGrad_R~v 0.090 -0.024 (0.072) (0.057) LaborForce_Pa~v -0.104 -0.017 (0.145) (0.090) Manufacturing~v 0.182* 0.001 (0.110) (0.082) Growth_Chines~v -0.113** 0.000 (0.055) (0.045) Teenage_Labou~v 0.066 -0.189 (0.145) (0.127) Migration_Inf~v 0.071 -0.078 (0.105) (0.072) Migration_Out~v -0.220* -0.014 (0.114) (0.076) Fraction_Fore~v -0.077 -0.016 (0.119) (0.094) SocialCapital~v -0.470*** -0.039 (0.166) (0.123) Fraction_Reli~v 0.149 -0.027 (0.129) (0.093) Crime_R_sv -0.272** 0.227** (0.106) (0.095) Children_Sing~v -0.492** 0.232 (0.244) (0.175) Fraction_Adul~v 0.275* -0.121 (0.146) (0.104) Fraction_Adul~v -0.009 0.088 (0.146) (0.118) IncomeGrowth_sv -0.089 0.215** (0.139) (0.099) _cons -0.550* -0.111 (0.308) (0.263) Obs. 414 414 R-squared 0.290 0.580 statedummy YES statedummy YES
Standard errors are in parenthesis
*** p<0.01, ** p<0.05, * p<0.1
Notes: OLS regression results of RM/AM difference between the two genders on the full set of
Appendix 2
Table A2. Correlates of the [p5-p95] RM/AM difference between the two genders across CZs-robust standard error
Manufacturing~v 0.123 0.017 (0.080) (0.072) Growth_Chines~v -0.044 0.035 (0.060) (0.030) Teenage_Labou~v 0.137 -0.182* (0.116) (0.106) Migration_Inf~v 0.023 -0.090 (0.078) (0.065) Migration_Out~v 0.015 0.015 (0.083) (0.067) Fraction_Fore~v 0.037 -0.089 (0.082) (0.085) SocialCapital~v -0.115 0.027 (0.123) (0.101) Fraction_Reli~v 0.124 -0.027 (0.081) (0.082) Crime_R_sv -0.139* 0.199** (0.077) (0.083) Children_Sing~v -0.283 0.003 (0.193) (0.154) Fraction_Adul~v 0.204* -0.085 (0.112) (0.087) Fraction_Adul~v -0.090 -0.092 (0.099) (0.110) IncomeGrowth_sv 0.022 0.162** (0.099) (0.081) _cons -0.318 -0.325 (0.227) (0.221) Obs. 383 387 R-squared 0.254 0.514 Statedummy
Statedummy YES YES
Standard errors are in parenthesis
*** p<0.01, ** p<0.05, * p<0.1
Notes: OLS regression outcomes of [p5-p95] RM/AM difference between the two genders on the
full set of explanatory variables as represented in Table 3 based on robust standard error. By [p5-p95], we mean that the extreme 5% (lowest and highest) of relative/absolute mobility difference is removed before running the regression models. Dependent and independent variables are
Appendix 3
Table A3. Correlates of the RM/AM difference between the two genders across CZs with statistically significant CZ characteristics-robust standard error
(1) (2) diffRM_sv-rse diffAM_sv-rse Fraction_Blac~v 0.178*** -0.182*** (0.059) (0.055) Fraction_Comm~v -0.199* 0.451*** (0.102) (0.069) Gini_sv -0.120 -0.215*** (0.080) (0.062) Local_Tax_R_sv -0.111 0.204*** (0.094) (0.078) Tax_Progressi~v 0.225** -0.193*** (0.090) (0.058) State_EITC_Ex~v -0.206** 0.226*** (0.104) (0.049) School_Expend~v 0.281** -0.249*** (0.124) (0.069) College_Tuiti~v 0.033 -0.023 (0.040) (0.030) Growth_Chines~v -0.091* 0.006 (0.053) (0.034) Migration_Out~v -0.124** -0.129*** (0.061) (0.036) SocialCapital~v -0.171 -0.216*** (0.122) (0.081) Crime_R_sv -0.190** 0.187** (0.086) (0.078) Fraction_Adul~v 0.080 -0.032 (0.078) (0.054) IncomeGrowth_sv 0.022 0.052 (0.104) (0.075) _cons -0.249 -0.109 (0.261) (0.159) Obs. 534 534 R-squared 0.180 0.555 statedummy YES statedummy YES
Standard errors are in parenthesis
*** p<0.01, ** p<0.05, * p<0.1
Notes: OLS regression results of RM/AM difference between the two genders on the explanatory
Figures
Variable Definition Source
(1) (2) (3)
Fraction Black Number of individuals who are black alone divided by total population 2000 Census SF1 100% Data Table P008 Racial Segregation Multi-group Theil Index calculated at the census-tract level over four groups:
White alone, Black alone, Hispanic, and Other
2000 Census SF1 100% Data Table P008
Income Segregation Rank-Order index estimated at the census-tract level using equation (13) in Reardon (2011); the δ vector is given in Appendix A4 of Reardon's paper. H(pk)
is computed for each of the income brackets given in the 2000 census. See Appendix D for further details.
2000 Census SF3 Sample Data Table P052
Segregation of Poverty (<p25) H(p25) estimated following Reardon (2011); we compute H(p) for 16 income groups defined by the 2000 census. We estimate H(p25) using a fourth-order polynomial of the weighted linear regression in equation (12) of Reardon (2011).
2000 Census SF3 Sample Data Table P052
Segregation of Affluence (>p75) Same definition as segregation of poverty, but using p75 instead of p25 2000 Census SF3 Sample Data Table P052 Fraction with Commute < 15 Mins Number of workers that commute less than 15 minutes to work divided by total
number of workers. Sample restricts to workers that are 16 or older and not working at home.
2000 Census SF3 Sample Data Table P031
Household Income per Capita Aggregate household income in the 2000 census divided by the number of people aged 16-64
2000 Census SF3 Sample Data Table P054
Gini Gini coefficient computed using parents of children in the core sample, with income topcoded at $100 million in 2012 dollars
Tax Records, Core Sample
Top 1% Income Share The fraction of income within a CZ going to the top 1% defined within the CZ, computed using parents of children in the core sample
Tax Records, Core Sample
Gini Bottom 99% Gini coefficient minus top 1% income share Tax Records, Core Sample
Fraction Middle Class (between p25 and p75)
Fraction of parents (in the core sample) whose income falls between the 25th and 75th percentile of the national parent income distribution
Tax Records, Core Sample
Local Tax Rate Total tax revenue per capita divided by mean household income per capita for working age adults (in 2000)
1992 Census of Government county-level summaries
Local Govt Expenditures Per Capita Total local government expenditures per capita 1992 Census of Government county-level summaries
Tax Progressivity The difference between the top state income tax rate and the state income tax rate for individuals with taxable income of $20,000 in 2008
2008 state income tax rates from the Tax Foundation
State EITC Exposure The mean state EITC top-up rate between 1980-2001, with the rate coded as zero for states with no state EITC
Hotz and Scholz (2003)
School Expenditure per Student Average expenditures per student in public schools NCES CCD 1996-1997 Financial Survey Student Teacher Ratio Average student-teacher ratio in public schools NCES CCD 1996-1997 Universe Survey Test Score Percentile (Income
adjusted)
Residual from a regression of mean math and English standardized test scores on household income per capita in 2000
George Bush Global Report Card
High School Dropout Rate (Income adjusted)
Residual from a regression of high school dropout rates on household income per capita in 2000. Coded as missing for CZs in which dropout rates are missing for more than 25% of school districts.
NCES CCD 2000-2001
Number of Colleges per Capita Number of Title IV, degree offering insitutions per capita IPEDS 2000 College Tuition Mean in-state tuition and fees for first-time, full-time undergraduates IPEDS 2000 College Graduation Rate (Income
Adjusted)
Residual from a regression of graduation rate (the share of undergraduate students that complete their degree in 150% of normal time) on household income per capita in 2000
IPEDS 2009
Labor Force Participation Share of people at least 16 years old that are in the labor force 2000 Census SF3 Sample Data Table P043 Share Working in Manufacturing Share of employed persons 16 and older working in manufacturing. 2000 Census SF3 Sample Data Table P049 Growth in Chinese Imports Percentage growth in imports from China per worker between 1990 and 2000,
scaled as an annualized rate times 10
Autor, Dorn, and Hanson (2013)
Teenage (14-16) Labor Force Participation
Fraction of children in birth cohorts 1985-1987 who received a W2 (i.e. had positive wage earnings) in any of the tax years when they were age 14-16
Tax Records, Extended Sample
Migration Inflow Rate Migration into the CZ from other CZs (divided by CZ population from 2000 Census)
IRS Statistics of Income 2004-2005
Migration Outlflow Rate Migration out of the CZ from other CZs (divided by CZ population from 2000 Census)
IRS Statistics of Income 2004-2005
Fraction Foreign Born Share of CZ residents born outside the United States 2000 Census SF3 Sample Data Table P021
Social Capital Index Standardized index combining measures of voter turnout rates, the fraction of people who return their census forms, and measures of participation in community organizations
Rupasingha and Goetz (2008)
Fraction Religious Share of religious adherents Association of Religion Data Archives
Violent Crime Rate Number of arrests for serious violent crimes per capita Uniform Crime Reports Fraction of Children with Single
Mothers
Number of single female households with children divided by total number of households with children
2000 Census SF3 Sample Data Table P015
Fraction of Adults Divorced Fraction of people 15 or older who are divorced 2000 Census SF3 Sample Data Table P018 Fraction of Adults Married Fraction of people 15 or older who are married and not separated 2000 Census SF3 Sample Data Table P018
Local Labor Market Migration Social Capital Family Structure Segregation Income Inequality Tax K-12 Education College
Figure 1. A short description of CZ characteristics.
Source: From “Where is the Land of Opportunity? The Geography of Intergenerational Mobility in
