Title: Educational Selectivity in US Immigration: How do immigrants compare to those left behind?
Author: Cynthia Feliciano
Source: 2005 Feb, 131-152, Demography
Category and Keyword: Migration, Education, Global, US Immigration
To understand the kinds of immigrants that come to the U.S. better, helping explain socioeconomic differences among immigrant groups in the U.S.
How do immigrants' characteristics compare to those of persons who remain in the sending society (Gans2000)?
Immigrants from sending countries are not random samples, as previous studies have not challenged (Borjas, 1999).
Why important? Remaining population may be “drained” of important resources
Educational selectivity may affect how will adept to the U.S.;
May explain why others are more successful than others. Understanding the relative position of immigrants in their country of origin is necessary to test theories of assimilation that predict upward or downward mobility in the United States among the second generation.
Old View (B. Franklin): they escape desperate poverty and unemployment (see Portes & Rumbaut 1996).
Newer: All immigrants are self-selective, i.e. more ambitious/willing to work, have some education / exposure (also see Portes & Rumbaut 1996).
Push and Pull factors (Lee 1966) – Pull = immigrants leave because of “plus factors” = positively selected; Push = “minus factors” in the sending society.
Borjas (economic): more positively selected in more egalitarian societies; negatively selected in more unequal societies like Mexico (sends unskilled here, skilled have jobs)
Even if immigrants are all positively selected, there may be substantial variability in the level of selectivity by origin country, such that immigrants from some countries are more positively selected than others.
Given the greater costs associated with migrating long distances, migrants from countries that are farther from the United States should be more highly selected.
Massey (1987b, 1999) contended that although migrants tend to be positively selected initially, they become less highly selected over time as successive waves migrate from a particular country. Social capital is a major force in perpetuating migration;
Using Mexican and U.S. census data, I examined whether successive waves of immigrants from Mexico were less educated relative to Mexican nonmigrants than were those who immigrated earlier.
Most of the country-of-origin data were available from UNESCO (1975-1997), which compiles census data from the various countries and presents the data in comparable ways.
To summarize educational attainment on the receiving side of migration, I created extracts of census data on U.S. immigrants from each of the 32 countries from the Integrated Public Use Micro Samples.
I combined the Mexican and U.S. census samples for 1960-2000 to create a data set for each year that consisted of a large sample of Mexicans in Mexico and Mexican immigrants in the United States.
Table 1: Summarizes the variation in educational selectivity by country of origin for all immigrants, as well as for women and men separately.
The findings show that even though all immigrant groups are positively selected, the degree of positive selectivity varies considerably by country of origin. Immigrants from Asia tend to be more positively selected than those from Latin America or the Caribbean.
although educational selectivity often differs between male and female immigrants from the same country, these differences are generally not great.
Table 2: Relationships Between Select Factors and Immigrants' Educational Selectivity (Net Difference Indexes.
correlations between the included variables and immigrant selectivity (NDI), bivariate regression coefficients for the NDI regressed on each variable, and multivariate regression coefficients for a model including all significant bivariate relationships.
immigrants from highly educated populations are less likely to be as highly positively selected as those from less-educated populations.
Greater distance from the United States is associated with greater positive selectivity
immigrants from countries who only recently began migrating to the United States tend to be more positively selected than those who came primarily in the 1960s or 1970s; however, this relationship is not statistically significant in the bivariate model. Thus, these findings challenge the popular perception that immigrants' skills have declined as the regional origins of immigrants have changed over time.
Some support for Borja: but: immigrants from more-egalitarian countries are more positively selected, income inequality is not a statistically significant predictor of selectivity in the bivariate model. This finding contradicts the theory that immigrants from highly unequal societies are less likely to be positively selected.
Political reasons and immigration = +selection
Figure 1: Educational Selectivity of Migrants to the United States, by Average Decade of Migration and Region
Most immigrant groups whose major waves arrived in the 1960s and 1970s are from Europe, whereas most immigrant groups who arrived more recently are from Asia, Latin America, and the Caribbean.
This figure suggests that immigrant groups today, especially those from Asia, are actually more likely than were earlier immigrants to come from the top of the educational distribution in their countries of origin. Thus, any suggestions that immigrants are currently less selective than in the past owing to their changing regional origins are overstated.
Table 3: Coefficients of the Determinants of Years of Schooling Among Mexicans in the United States and Mexico, Aged 25-64
those who migrated within the past five years, there is a pattern of strong positive selection from 1960 to 2000. Migrants consistently averaged more than one additional year of schooling than nonmigrants
Overall, the findings do not indicate substantial changes in selectivity. These mixed results may have to do with the changing nature of migration from Mexico. While migration from Mexico has historically been dominated by migrants from rural Mexico, in recent years, a growing number have come from urban areas.
Questions Massey’s thesis of declining selectivity over time. // Thus, it may be possible that selectivity is declining among migrants from rural areas, where social capital mechanisms operate most strongly in reducing the costs of migration, whereas urban-origin migrants, who are more educated, may be responding to a different set of factors (Fussell and Massey 2004).
I found that there is substantial variation in the degree of educational selectivity depending on the country of origin and the timing of migration from a particular country, but that nearly all immigrant groups are more educated than their nonmigrant counter- parts.
Although scholars have agreed that immigrants do not represent a random sample of their home countries' populations, from the vantage point of the average U.S. native, who sees only immigrants and not those who remain in the homeland, it is easy to attribute immigrants' characteristics to an entire national group
3 Demography Items, including Adaptation
Title: Immigrant residential segregation in US metropolitan areas, 1990-2000
Author: Iceland, J. & Scopilliti, M.
Source: Demography; 42(1) 131-152
Category and Keyword: Immigrants, Residential Segregation, 1990-2000
To examine the extent of spatial assimilation among immigrants of different racial and ethnic origins.
1990-2000 restricted US Census Data
CALCULATE: Levels of Dissimilarity by 1) Race 2) Hispanic origin 3) Nativity 4) Year of Entry
Multivariate Models are then Run
Broad support for spatial assimilation theory.
Segregation: foreign-born Hispanics, Asians and Blacks are more segregate from non-Hispanic whites than are the US born of these groups
Those in the US longer are less segregated than those who more recently arrived.
BUT: Blacks have much higher levels of immigration than Asians, Hispanics and Whites
ODDITY: Black immigrants have higher SES than native blacks in U.S. – SO the above things that EXPLAIN for other race segregation, doesn’t explain it for Blacks
Some support for segmented assimilation theory for Black immigrants – but not unequivocal.
Support for Spatial assimilation model? Overtime, they are less segregated;
Why do we have cautions?
1) Blacks and Black immigrants continue to be “considerably” segregated from whites.
2) Short and medium run, we expect high levels of Hispanic immigration, of low SES, and therefore longer periods of time of segregation. Will this polarization close over time?
3) need to look at intra-group variation
3 Demography Items, including Adaptation
Title: Growing up American: The Challenge Confronting Immigration Children and Children of Immigrants
Author: Zhou, Min
Source: ARS 23:63-95
Category and Keyword: Immigrants, Children, Family
The purpose of this article is to pull together existing studies that bear directly or indirectly on children’s immigrant experiences and adaptational outcomes and to place these studies into a general framework that can facilitate a better understanding of the new second generation.
How will children be assimilated? Classical theory (straight line assimilation) or other?
Period of Study:
Looking at the children of immigrants SINCE 1980s –
Findings / Other Researchers:
Ruben Rumbaut: coined the term, 1.5 generations, to refer to children who straddle both worlds but are not part of either of the two.
Broken down 6-13 for 1.5 generation
13-17 aged children are similar to first generation children;
BUT 0-4 foreign born children put with second generation children – often.
Page 70: reviews THEORIES, i.e. classical assimilation (Park 1928) then with added variation, but maintaining basic thesis (Gordon 1964); e.g. cultural or behavioral, structural, identificational, attitude- receptional, behavior-receptional, and civic assimilation. AND more recently: social mobility across generations of immigrants and increasing rates of intermarriages, as determined by educational attainment, job skills, length of stay since immigration, English proficiency, and levels of exposure to American cultures (Alba 1985 .. Greeley 1976);
Specifically, children of highly educated immigrants consistently fare much better in school than do fourth- or fifth- generation descendants of poorly educated ancestors regardless of religio-ethnic backgrounds. (e.g. Hirschman)
GANS (1992): immigrant with low education continue to pass it on 3rd / 4th generation;
SEE OTHER Article on THIS by ZHOU
The pluralist perspective offers an alternative way of viewing the host society, treating members of ethnic minority groups as a part of the American population rather than as foreigners or outsiders and presenting ethnic or immigrant cultures as integral segments of American society.
However, the questions of “second-generation decline” and “second-generation revolt” have been unanswered within this theoretical framework.
Measurement problems: Perlmann & Waldinger (1996) note that, because of high rates of intermarriages in the third generation, the respondent’s choice of ethnic identity is selective, making it difficult to accurately predict the independent effect of ethnic origin on intergenerational mobility.
No contextual measures in the US Census: the census data do not have any direct measures for contextual effects of the family, the school, the neighborhood, and the ethnic community, nor do they have detailed information on school performance.
NELS has it; and Add Health;
3 Demography Items, including Adaptation
Title: Neighborhood Selection and the Social Reproduction of Concentrated Racial Inequality
Author: Robert Samson and Patrick Sharkey
Source: Demography 2008 V45 N1
Category and Keyword: Racism, Segregation, Household units, Chicago, also Spatial (neighborhood structures)
To show how Neighborhood Selection is a social process central to the reproduction of racial inequality in neighborhood attainment.
Point of study:
(Dr. Raley): Boundaries are discreet. In this context we are talking about living near: 1994 Farley ARS/AJS - gave whites a picture with a neighborhood; measure of tolerance for immigration; whites fairly quickly felt discomfort; blacks wanted more integration; bottom line: I don't think segregation is driven by choices by Black people, but rather the white population;(and Hispanics) who move OUT of Black neighborhoods;
All clusters of Chicago’s 343 neighborhoods – then 80 were randomly selected (first part of study = 4,000 households that were followed over a period of 7 years); (PHDCN study);
Residential stratification falls powerfully along racial/ethnic lines and socio-economic location, especially income and education. // for the most part only surviving factors that explain a significant portion of the variance in neighborhood attainment conditions.
Only 10% explained – so individual characteristics only go so far to explain neighborhood stratification.
Sorting as a social process (whites / Hispanics leave city where there are growing non-whites)
Decisions made by movers and stayers.
Thus poverty traps are likely to remain unless there is state interventions **
Therefore, neighborhood selection is a process of stratification.
3 Demography Items, including Adaptation
Title: Migration and Spatial assimilation among U.S. Latinos: Classical versus Segmented Trajectories
Author: Scott J. South, Kyle Crowder, and Erik Chavez
Source: Demography 2005 V42 N3
Category and Keyword: Migration, Spatial Assimilation, Segmented Theory, Assimilation Theory
To study Hispanic Immigrant Mobility in the U.S.
Latino National Political Survey / Panel Study of Income Dynamics / US Census
1990-1995 – 2,074 US residents from: Mexico, Puerto Rico and Cuba
Support central tenants of Spatial Assimilation Theory
Latino residential mobility into neighborhoods that are inhabited by greater percentages of non- Hispanic whites (i.e., Anglos) increases with human and financial capital and English-language use.
Supports Segmented Assimilation Theory, too
Puerto Ricans are less likely than Mexicans to move to neighborhoods with relatively large Anglo populations.
He generational and socioeconomic differences that are anticipated by the classical assimilation model emerge more strongly for Mexicans than for Puerto Ricans or Cubans. (Harder for Mexicans)
Among Puerto Ricans and Cubans, darker skin color inhibits mobility into Anglo neighborhoods.
Many versions of this classical model of ethnic assimilation view spatial assimilation -- an ethnic group's geographic proximity to the majority group – as an especially salient dimension of this general process (Massey 1985).
Classical Assimilation theory affirmed: For example, among Mexicans, residential mobility out of origin neighborhoods and into neighborhoods that are occupied by larger percentages of Anglos is greater among later generations than among the first generation.
AND Latino migration into more-Anglo neighborhoods tends to increase with human and financial capital and with English-language use.
Segmented Assimilation theory affirmed: Puerto Ricans stand out as having both low rates of intertract (interneighborhood) mobility overall and, conditional upon moving, low rates of mobility into more-Anglo neighborhoods compared with Mexicans.
It appears that Cubans move to neighborhoods that are less Anglo than the neighborhoods that Mexicans move to primarily because Cubans are concentrated in metropolitan areas that contain comparatively few Anglos.
Among Cubans and especially among Puerto Ricans, residential mobility into Anglo neighborhoods is sharply segmented by skin color, with dark skin a substantial impediment to moving to neighborhoods that are inhabited by the non-Hispanic white majority.
Other, not in either theory: For all three Latino ethnic groups, residential mobility into more-Anglo neighborhoods increases sharply as the percentage of the metropolitan area population that is Anglo increases.
Future research examples:
Furthermore, research designs that trace the internal residential mobility experiences of immigrants upon their entry to the United States -- and the mobility experiences of their children as they leave the parental home-hold promise for providing a more complete description of intergenerational patterns of spatial assimilation.
Finally, future research on ethnic residential mobility and assimilation will need to deal with the growing complexity and temporal dynamics of U.S. neighborhoods.
** Send to Mauritius: ZHOU 1997 **
4 Inequality, Social Mobility and Labor
Title: Is Rising Earnings Inequality Associated With Increased Exploitation? Evidence For U.S. Manufacturing Industries, 1971–1996
Category and Keyword: exploitation, earnings inequality, manufacturing industry
Investigate whether or not there is exploitation among workers in the manufacturing industry.
Exploitation of workers has not been empirically investigated since Marx.
Definition of Exploitation: the extent to which the earnings of various groups in the labor force are underpaid relative to the market value of their productivities.
Concept of exploitation different from the process that generates it;
“the value of what one produces”: no consensus yet on how to conceptualize
For this study: The “value of what is produced by labor,” on a per-unit basis, is then defined as the increment in the total output value that accrues by employing another unit of labor of a particular type, holding constant the other factors of production by way of the multivariate production function.
Exploitation: We then define exploitation as the underpayment of a particular category of workers relative to its marginal revenue product.
What they Add to this Science: Our analysis can empirically identify which groups (if any) may receive some of the surplus generated by the exploitation of some types of workers. This approach contrasts with the Marxist view that begins with the theoretical assumption that only capitalists benefit from exploitation and that no groups of employees can ever be exploitative (Wright 2000).
Weakness: do not use capital incomes (but note: bulk of income comes via wages and salaries.)
Durkheim: functionalist view of stratification; Weber: none; Marx: between classes/capital income. Weber used: Weber’s concept of class as “market situation” is compatible with the idea that exploitative economic relations can be generated by processes relating to market closure affecting economic actors in the same class category. Our investigation of exploitation is consistent with Weber’s general perspective on social stratification.
Variables included are 1) occupational structure (Grusky 2005); 2) Demographic and ascribed sources of inequality (external characteristic used to exclude, i.e. Weber); 3) Gender segregation in lower sector jobs; 4) Labor market discrimination against African Americans; 5) Education (more means more productivity); 6) Marital status (married people have higher earnings, esp. men); 7) Age (correlates with developmental work skills);
Authors use a model of productivity and earnings differentials
Limited to the Manufacturing Sector
DATA: A) annual information on productivity and other related factors for manufacturing industries is the Manufacturing Industry Data- base (NBER-CES MID), which is compiled by the National Bureau of Economic Research and the Center for Economic Studies of the U.S. Census Bureau. B) this database does not provide any information on workers’ characteristics in each manufacturing industry. We therefore use the NBER-CES MID in conjunction with other annual data from the Current Population Survey (CPS).
Unit of Analysis: We matched the NBER-CES MID data on productivity (and other related factors) and the CPS data on workers’ characteristics on the basis of three-digit Standard Industrial Classification (SIC) codes. That is, our unit of analysis is the three-digit SIC code. Panel data set of 1,456 CASES consisting of 56 manufacturing industries, observed annual from 1971 to 1996.
Two Regression models:
DV: Total earnings paid to all workers in the particular industry during a given year.
IVs for both: include the proportions indicating the various characteristics of workers in that industry during a given year.
Measuring inequality: As explained above, when a particular group has a higher relative marginal productivity compared to the reference group (on average), then its ∂ will be greater than one, while in the opposite case, it will be less than one. When a particular group has higher earnings compared to the reference group (on average), then its ∂ will be greater than one, while in the opposite case, it will be less than one. When workers of a particular category are paid exactly according to their marginal productivities (on average), then there should be no difference between ∂ and ∂ for that group.
Workers relative in the distribution of earnings: To the extent that the relative earnings differential for workers in a particular quintile is matched by a corresponding increase in its relative marginal productivity, then ∂ should equal , which can be again assessed with a Wald chi-square test.
Table 1: Descriptive Statistics; Value of earnings/capital investment in terms of 1984 dollars: 1971: $13,913 – 1996: $19,913 millions. Mean # of workers: 313 to 309 thousand (shows manufacturing productivity increased);
Table 2: Nonlinear Least Squares Estimates of Productivity and Earnings Model
Overpayment for various groups, based on what one group is paid compared to another group (i.e. reference group); These estimates of differentials in productivity, earnings, and exploitation refer to proportionality contrasts relative to a particular reference category.
i.e. women 39% underpaid compared to men (women are only 85% of productivity than men, but earn only 47% of what men earn, i.e. 85.5% - 46.8%)
Hispanics: underpaid by 29%
African Americans: underpaid by 23%
Overpayments: Middle aged workers of 40% and Old Age 47%
Table 3: Nonlinear Least Squares Estimates of Productivity and Earnings Models of the Effects of Relative Position in the Distribution of Earnings
Overall, the findings reported in Table 3 indicate that exploitation increased in the later period relative to the earlier period. In the period from 1971 to 1982, the statistical significance tests confirmed that only one quintile was clearly overpaid (i.e., the fifth) and only one quintile was undoubtedly underpaid (i.e., the first). In the period from 1983 to 1996, however, the statistical significance tests indicated that two quintiles were overpaid (i.e., the fourth and the fifth) while two quintiles were underpaid (i.e., the first and the second). Thus, exploitation increased in the later period because the second and fourth quintiles were no longer being paid according to their marginal productivities but instead had become either exploited or exploiting groups. Although the level of exploitation of the first quintile changed only slightly in the later period compared to the earlier period, the exploitative over- payment of the fifth quintile increased substantially from 134 percent to 162 percent (i.e., an increase of 28 percentage points).
Explaining the result via power: This trend towards increasing exploitation may be interpreted in terms of Hirsch and De Soucey’s (2006) discussion regarding the role of power in influenc- ing the economic outcomes associated with organizational restructuring in recent years. Rather than reflecting only a rising organizational imperative for greater ef- ficiency, the increased inequality that has occurred is at least partly a reflection of bargaining power differentials between workers. The decline in unions, the falling real value of the minimum wage, the increased number of part-time workers, and the dismantling of internal labor market practices have all increased inequality but may be less important to improving productivity than is usually assumed in popular discussions (Kim and Sakamoto 2008b).