Between 1942 and 1946, more than 110,000 Japanese Americans in the western part of the United States were evacuated and relocated by the War Relocation Authority. I estimate the effects of this experience on weekly earnings of Japanese-American men in the states of California, Oregon and Washington using the micro-data from the 1940 and 1950 U.S. Censuses. Empirical results demonstrate that the forced withdrawal from the labor market had considerable adverse impacts on earnings of prime-age Japanese men. The loss of earnings was magnified as during this decade other minority groups made tremendous advances in narrowing the wage gap with whites. Between 1940 and 1950, the average weekly earnings for those with 12 years of education grew by 24 percent for Japanese Americans, whereas blacks and other Asians enjoyed increases of 50 percent and 57 percent, respectively. Estimation results also indicate that the effects of the internment were different depending on the individuals education level. For example, the return on skills, measured by extra earnings for an additional year of experience, deteriorated considerably for Japanese Americans with higher education, while there were no noticeable differences among racial/ethnic groups with less than a high school education. Difference-in-differences estimation results indicate that the fall in earnings of Japanese Americans was largely due to changes in industry and occupation choices rather than depreciation of human capital within the same industry or occupation.

This paper examines the relationship between interethnic marriages and economic assimilation among immigrants in the United States. Two competing hypotheses are evaluated: the productivity hypothesis, according to which immigrants married to native-born spouses assimilate faster than comparable immigrants married to foreign-born spouses because spouses play an integral role in the human capital accumulation of their partners; and the selection hypothesis, according to which the relationship between intermarriages and assimilation is spurious because intermarried immigrants are a selected subsample from the population of all married immigrants. These two hypotheses are analyzed within a model in which earnings of immigrants and their interethnic marital status are jointly determined. The empirical evidence favors the selection hypothesis. Non-intermarried immigrants tend to be negatively selected, and the intermarriage premium obtained by the least squares completely vanishes once we account for the selection.

The segregation of cities in the United States has been of considerable interest in how it affects the economic outcomes of blacks and whites. Previous literature has examined the negative effects of segregation on blacks educational attainment, earnings, idleness, and rate ofsingle motherhood. In highly segregated cities with ghettos, blacks have unequal access to quality jobs, schooling, and role models. Over time, new laws, better communication, and better transportation should break some of the barriers caused by segregation, diminishing its negative effects.This paper examines the effects of segregation on an individuals earnings over time, focusing on variations across regions. Using Census data from 1950 to 1990, it illustrates how segregation in an individuals city impacts his or her earnings. It concludes that the effects ofsegregation on earnings depend on region and time, and are largely a factor of the level of discrimination against blacks.

Over forty to fifty years, there are large shifts in the survey methods employed by federal surveys like the Current Population Survey (CPS) and the National Health Interview Survey (NHIS) that make analytical harmonization difficult (but not impossible). Changes present a special challenge to researchers interested in constructing an analytically consistent analysis of change over time in a phenomenon of interest. Sampling changes, instrumentation changes, and data processing changes are the three most common challenges causing potential gaps in the analysis of trends over time. Changes in sampling of people within households and households within Primary Sampling Units can cause problems for variance estimation. Instrumentation changes can cause estimates to vary for no other reason. These include question wording, question universe definitions, and computer-assisted instruments versus paper and pencil. Data processing changes can also alter estimates over time. These include how missing data is handled, how out of range values are edited, and how universe definitions are enforced. As part of the integrated public use microdata projects at the Minnesota Population Center, we have developed a variety of methods for bridging the gaps in surveys. First and foremost, we tightly integrate information about these major changes into our documentation, so data users know about the problems when they download the data from our system. Knowledge of these changes is important for performing appropriate analyses. Second, we are developing various statistical approaches to bridging the gap and testing their performance for analyzing trends. We will highlight examples from our NHIS and CPS data projects to extract basic rules for dealing with these gaps. We will also point out instances in which the gaps are just too big to bridge for most analytical purposes.

This paper presents intergenerational evidence in favor of the hypothesis that a significant factor explaining the increase in female labor force participation over time was the growing presence of men who grew up with a different family model--one in which their mother worked. We use differences in mobilization rates of men across states during WWII as a source of exogenous variation in female labor supply. We show, in particular, that higher WWII male mobilization rates led to a higher fraction of women working not only for the generation directly affected by the war, but also for the next generation. These women were young enough to profit from the changed composition in the pool of men (i.e., from the fact that WWII created more men with mothers who worked). We also show that states in which the ratio of the average fertility of working relative to non-working women is greatest, have higher female labor supply twenty years later.

This paper is about non-approximate acceleration of high dimensional nonparametric operations such as k nearest neighbor classifiers and the prediction phase of Support Vector Machine classifiers. We attempt to exploit the fact that even if we want exact answers to nonparametric queries, we usually do not need to explicitly find the datapoints close to the query, but merely need to ask questions about the properties about that set of datapoints. This offers a small amount of computational leeway, and we investigate how much that leeway can be exploited. For clarity, this paper concentrates on pure k-NN classification and the prediction phase of SVMs. We introduce new ball tree algorithms that on real-world datasets give accelerations of 2-fold up to 100-fold compared against highly optimized traditional ball-tree-based k-NN. These results include datasets with up to 106 dimensions and 105 records, and show non-trivial speedups while giving exact answers.

As the U.S Census Bureau moves toward full implementation of the American Community Survey (ACS) to replace the Census Long Form Questionnaire, demographers must determine how best to collect data on the population living in Group Quarters (GQs). In the past, the Census Bureau has spent much time and many resources to collect detailed social and economic data using Long Form Questionnaires to sample GQ residents. Currently, because of the cost and labor intensity of detailed GQ data collection, the ACS has delayed collecting data on GQ populations until 2006. This presentation explores the utility of collecting detailed information in GQs. Using data from Census 2000 Summary File 3 and PUMS, we find that Long Form data on GQs suffer from high allocation rates (about 75%); and that the output has only marginal utility to researchers and policy makers due to limited coding of GQs by type. For instance, data output does not distinguish between different types of institutions (like prisons, nursing homes, or hospitals) nor types of non-institutional settings (like shelters, hotels, and half-way houses). The Long Form (PUMS) does provide potentially useful social and economic characteristics of residents in college dormitories and military barracks.

This article offers a new approach for deepening our understanding of the immigrant experience. It describes how and explains why a historically grounded cohort analysis brings to the fore aspects of migr views and involvements, including within a single immigrant generation, other approaches leave undocumented and unexplained. Differences in pre-migration experiences are shown to shape both how immigrants adapt to their new country of settlement and how they relate to their homeland. The utility of the approach is illustrated by contrasting the experiences of different cohorts of Cuban immigrants: how they have adapted here and the nature of their transnational ties.

Artists continue to sort themselves out among American cities. In the 1990s, they reversed a trend of several decades and gravitated in largernumbers towards three premier centers of tourism, entertainment and creative work: Los Angeles, New York and San Francisco. Theyalso favored a set of second tier metros Washington DC, Seattle, Boston, Orange County, Minneapolis-St. Paul, San Diego and Miami over nineteen other large metro areas. Neither sheer metropolitan workforce size nor recent growth rates explain these divergent patterns. A combination of amenities, regional support for the arts, informal networks among artists and synergy with particular industries appear to explain their presence and persistence. In this update of our 2003 study, The Artistic Dividend: The Hidden Contributions of the Arts to the Regional Economy, we explore the results of the 2000 Census to update our depiction of artistic prowess city by city, expanding our analysis to the twenty-nine largest U.S. metros. We confirm the tendency for different metros to specialize in artistic suboccupations. Performing artists, visual artists and writers sort themselves out in distinctive spatial patterns rather than replicating each others preferences. We add two arts-related occupations, architects and designers, to our analysis, showing how members of these groups, more prosperous and less likely to be self-employed, exhibit yet different urban patterns. We explore the relationship between occupation and industry in one case, advertising, to underscore the desirability of a dual analysis in understanding emerging urban economies. We also probe the significance of selfemployment among artists, which we find to be quite high. We find that many more individuals report artistic work as their occupation in the Census than do employer based data sources. We show that in some metros, artists are more likely to be formally employed than in others. Furthermore, some, especially musicians, pursue their income earning artistic activity as a second occupation. The number of artists in a region is greatly undercounted, in many subgroups by more than 100%, when non-Census sources are relied upon. We reassert our preference for using Census data for assessing the size of the artistic dividend. This update should be read along with The Artistic Dividend, our initial work. In that study, we reject the view that the arts are a discretionary element in a regional economy, disconnected from the competitive forces shaping its growth and stature. We articulate the various ways in which self-employed and other undercounted artists contribute to the economy through direct export of their work and services, through contractual work for area businesses, and by instigating innovation on the part of their suppliers. We explain the occupational approach to gauging the size of a metros artistic dividend and make the case for artists choice of a place to live and work independent of a particular employer or job offer. And we probe the policy implications for artists, private sector businesses, non-profits and state and local governments who wish to enhance the artistic sector of their economies, most of them modestly-priced initiatives that will augment artistic networks and learning, prompt greater artistic entrepreneurship, thicken the ties between non-artistic businesses and artists, and nurture diversified, decentralized artistic live and work spaces across metropolitan neighborhoods.

This article offers a new approach for deepening our understanding of the immigrant experience. It describes how and explains why a historically grounded cohort analysis brings to the fore aspects of migr views and involvements, including within a single immigrant generation, other approaches leave undocumented and unexplained. Differences in pre-migration experiences are shown to shape both how immigrants adapt to their new country of settlement and how they relate to their homeland. The utility of the approach is illustrated by contrasting the experiences of different cohorts of Cuban immigrants: how they have adapted here and the nature of their transnational ties.

The new immigrants, from Third World countries, often retain homeland ties. Theories that focus on assimilation accordingly do not adequately capture their experiences. More useful is a transnational conceptual frame.But analyses premised on transnational as well as assimilation perspectives typically conceptualize immigrant generations similarly. The big social divide that they point to is between first generation immigrant parents and their children born and raised where they resettle. That is, both perspectives focus on differences between generations genealogically defined. Yet, generations take on distinctive meaning depending on historical context. Though not in reference to immigration, Karl Mannheim (1952), and scholars influenced by him (c.f. Eisenstadt 1956, Zeitlin 1970, and Rumbaut 2004), for example, have addressed how generational political outlooks, shaped by key historical experiences, may be longstanding in their impact. Shared experiences give rise to shared worldviews influencing subsequent involvements and attitudes.

This research focuses explicitly on the life chances of biracial black/whites. I contrast the "new" biracials, those born to interracial parents in the post Civil Rights era, to the "old" biracials, the lighter-skinned descendants of the original mulatto elite. Both groups have occupied privileged positions relative to monoracial blacks within educational and occupational institutions. For the old biracials, this privilege derives both from the inherited advantages of the mulatto elite and from the independent signifcance of skin tonewithin the black community. I show that the skin tone privileges of lighter-skinned blacks declined for cohorts coming of age during and after the Civil Rights era. This decline marked the end of a system of stratifcation which characterized the black population for over a century and indicates that the new biracial advantage over monoracial blacks in educational outcomes is not a result of skin tone privileges within the black population. These new biracials differ from the old in that they have access to intimate white relatives within their family networks. The new biracial advantage may potentially result from white cultural resources and racial ambiguity. On the other hand, the new biracial advantage may stem from the selectivity of highly-educated parents into interracial unions. I show that the new biracial advantage over monoracial blacks in educational outcomes is largely explained by their relatively privileged family backgrounds. These advantages, and not racial ambiguity, result in higher grades and lower rates of drop out and grade retention, although they do not fully explain differences in standardized test scores. In order to understand the new biracial advantage, we must understand the dynamics of class/race assortative mating in the immediately prior generation. This pattern of interracial union formation is most accurately described as one of lower-class isolation. While traditional models of status exchange, educational homogamy, and differential propensity are all supported to some extent by the data, these patterns work to produce an overall pattern which universally excludes blacks with a high school degree or less from interracial unions, regardless of their potential partner's education. This finding points to the possibility of greater isolation for lower class blacks as interracialunions increase and to a generational bifurcation of the black class structure directly tied to issues of racial identity.

Recent work about the method of family reconstitution and economic history raises serious doubts about the demographic and economic premises that underlie much of the existing scholarship about early American family history. As a result, early American family historyone of the new social history's crowning achievements during the 1960sis now in disarray. Some scholars see the new microhistorical studies of the colonial family as an effort to sidestep these difficulties by ignoring demographic and materialist perspectives. However, such cultural approaches may well intensify the crisis by challenging the image of the early American family as a loving institution incapable of violent conflict.

Many studies document a clear relationship between the supply of marriageable men and marriage rates, but few studies find that the supply of marriageable men affects the number of women who choose to be single mothers. The model presented here addresses this puzzle. Many women view either marriage or single motherhood as an inframarginal choice because a third option, remaining single without children, is relatively attractive to them. Regression models that implicitly treat all women as potential mothers, who simply choose whether to raise children inside or outside marriage, may yield false inferences concerning the relationship between marriage markets prospects and family structure choices.

The paper announces the imminent availability of a major extension of the well-known Bateman-Foust sample. This new resource will contain linked agricultural and population census data between 1850 and 1880 for thousands of individuals in an expanded group of townships including some from the Pacific Coast states and Massachusetts and has in excess of 220, 000 person-observations. The paper discusses a number of problems and complications associated with the creation of this retrospective panel database. It also shows how these data may be linked to other computer-searchable databases and resources (such as land records) and discusses the impact of personal and family characteristics on persistence and the likelihood of record linkage within the sample townships using a panel from a completed subsection of the project.

Since the mid-1960s, black/white interracial marriage rates in the United States have in-creased rapidly. Although such marriages are still only a small fraction of total marriages,their prevalence continues to grow. This rise has renewed interest among social scientistsin classic theories which attempt to understand how the context of interracial unions a ectsmatching on other status characteristics.Various theories of interracial marriage patterns have been developed, and these theorieshave been implemented in empirical research using an equally-wide variety of methods. Whilemost researchers have developed their own theories and methods in contrast to other research,there has been little explicit comparison between competing models. In the research outlinedhere, I develop log-linear models from various theoretical perspectives and compare themusing a consistent data source and consistent methodological techniques. The results suggestthat educational homogamy fails to capture important patterns in interracial unions, andthat these patterns are best captured not by the prominent caste-status exchange theory butby a model which emphasizes the isolation of lower-class blacks.

(3) the pull of employment growth becamestronger and more industry-specific from thelate 1980s to the late 1990s; and (4) the pull ofservice employment growth, especially for theleast-educated Hispanic immigrants, becamemuch stronger in the later period. In thecontext of the progressive entrenchment ofneoliberalism and the major changes inimmigration policies, our empirical findingssuggest that the ethnically selective dispersalof immigrants in the late 1990s is probably thebeginning of a new trend.