This paper analyzes the impact that various public policies are likely to have in achieving a desegregation strategy that focuses on housing as outlined in United States et al. v. Board of School Commissioners of the City of Indianapolis, et al. Household level 1990 Census data for the 1990 Indianapolis Metropolitan Statistical Area, except for Madison County, is used to determine the likelihood that such policies will be successful. The analytical model is based on Gabriel and Rosenthals 1989 paper Household Location and Race: Estimates of a Multinomial Logit Model and McFaddens Choice Model (i.e. conditional logit model with both individual and choice-specific independent variables).There are three primary results from this paper: Gabriel and Rosenthals model can be expanded with McFaddens Choice Model to include variables representing the characteristics of the outcomes in addition to the characteristics of households, thereby, allowing for greater policy analysis economic variables (i.e. their coefficients) are statistically significant in household decision making regarding location race continues to play an important role in the locational choices of householdsThe housing related policies had very little favorable impact on the segregation indexes when viewed from a regional perspective, but were more effective when considering Marion County alone. From a regional perspective, it is possible that these policies would increase segregation because the coefficients of housing tenure and housing cost are not statistically significant for non-Caucasians, but are statistically significant for Caucasian households. On the issue of race, this analysis does not positively prove that racial discrimination or self-segregating preferences exists, but it rules out many other likely causes. If such race-based decision making is occurring, the fair housing and anti-discrimination policies contained in the desegregation order may have the greatest impact on desegregation of any policy in the desegregation agreements.

The latest data from Census 2000 paints a picture of changing New York City that is stunning, exciting and in some ways disturbing.There are more foreign-born New Yorkers than at any time since 1910. New York is a better-educated city, but it is also a less English-speaking city. For the affluent and extremely affluent it is a substantially richer city; for the poor and middle ranks, it is a poorer city. Most workers in New York depend on public transit to get to work, and it takes them much more time to get to work than workers elsewhere. During the 1990s, the travel time for workers increased. For renters, it is a somewhat more expensive city to live in that it used to be, but for homeowners, the expense is astronomical. New York City has always been much different from the rest of the state and the nation. In the 1990s, it diverged even more.

This dissertation addresses the tension between multiculturalism and national unity within the context of Spanish and English language usage in the United States. Will moves to promote bilingualism and dual-language school programs lead to a fragmented society? Under what circumstances might bilingualism be compatible with American identity? Where bilinguals have high status and Hispanics have political power, I find that English-speaking Hispanics are much more likely to maintain Spanish. The project first documents the role that language has played in national identity formation and, in turn, how beliefs about what it means to be American have shaped school policies relative to second language learning and the education of immigrant children. Then -using the Census, NCES, and other data sources - I develop and quantitatively test models of the macro- and micro-level contexts that influence (1) the level of bilingualism among Hispanic adults, and (2) the adoption of Spanish-English dual-language programs in public schools. The latter is one of several indicators of a more generalized valuation of bilingualism that extends beyond Hispanics and Hispanic communities. Qualitative case studies augment my findings by elucidating the motives and process behind decisions to institute dual-language school programs. The U.S. is not moving toward a bilingual norm, but my findings provide evidence that, under certain circumstances, Spanish-English bilingualism has become a compatible and stable part of what it means to be an American.

Much of the research examining the question of how, federal financial aid affects decisions to enroll in college has focused on the behavior of students in the relatively narrow range immediately following high school graduation, leaving unanswered the question of how, changes in the availability of aid affect the behavior of older students. This analysis examines the question of how changes in the means-tested federal Pell grant program affects enrollment decisions of potential students in their twenties and thirties. Our results indicate sizable effects of the introduction of the Pell grant program on college enrollment decisions,for older students.

Using the 5% public use micro sample of the 1990 U.S. census, we find that observationally equivalent workers in the manufacturing sector earn higher wages when they are in urban labor markets that have a larger share of national or metropolitan employment in their same occupation and industry groups. Quantitatively, the effect is large, with an elasticity (measured at the means) of between 1.2 and 3.6 for these effects. We interpret the willingness of firms to pay more for equivalent workers in dense markets as evidence of an agglomeration economy in urban labor.

Following the 1848 Treaty of Guadalupe Hidalgo, the citizenship and identity of individuals living in Southern New Mexico became a matter open to question. They were forced to make sense of competing notions about race, nation, and region. Questions about the racial and national implications of adopting/being assigned either a Mexican or an American identity became a critical issue in the newly transnational space. Two towns, Mesilla and Las Cruces, developed with competing ideas of race and nation. Although both towns were located in the United States, settlers in Mesilla frequently refused to acknowledge the boundary change and claimed their town as nationally Mexican Settlers in Las Cruces, conversely, attempted to integrate their town into the United States and conceived their Mexican identity as racial. This study explores the complexity of racial, national and regional ideologies in Southern New Mexico's Mesilla Valley from 1848 to 1912. The question of identity had longstanding repercussions on the lives of those living in the Mesilla Valley. This study uses the shifting national border in Southern New Mexico's Mesilla Valley to explore the complicated meanings of race, nation, and region along the nineteenth-century United States/Mexican border. Tensions and dissimilarities between the neighboring towns of Mesilla and Las Cruces make visible the complexity individuals faced when their local space was suddenly renamed either Mexican or American. To understand the ways that ideologies about race, nation, space and region developed during New Mexico's territorial period, this study considers areas in which changing circumstances perplexed local actors about the meaning of these categories. Changing conditions in Southern New Mexico confounded individuals who puzzled over the meanings of region, race, and nation at the boundary between the two nations. Mesillaros and Gringo Mexicans questions modern assumptions about how a Mexican/Mexican American identity emerged in the nineteenth century. This study disrupts commonly accepted knowledge about racial and national identities by suggesting that these identities have historically been dependent on the complexity of local contexts. By understanding how individuals deployed, disrupted and lived identities through local spaces we will begin to unravel the larger discourses surrounding national and racial identities.

Despite a once conspicuous presence in the Western states, little is known demographically about the Chinese in the late 19th and early 20th U.S. The widely accepted model of a declining male 'sojourner society,' beset by draconian restrictions on immigration and the impossibility of family formation, persists largely unexamined. This paper tests the sojourner hypothesis through the application of cohort-component projection on census data from 1880 through 1940, including data recently made available as part of the Integrated Public Use Microdata Sample (IPUMS), a historical data set constituted from U.S. census manuscripts. The results fail to support the sojourner model of passive population decline, suggesting instead that the Chinese actively engaged in a collective strategy of long-distance labor exchange to maximize economic productivity among Chinese workers in the U.S.

This study addresses some aspects of the financial impact on women under the Social Security benefits reform or redistribution. This paper presents a preliminary result of the study. A multiple decrement model (LL Model) is developed based on a proxy population of U.S. women and its demographicprojection. Social Security benefits under current policy rules are then assigned to each sample unit in the resulting proxy population. The authors then compare aggregate benefit entitlement figures under current policies to figures obtained when potential policy change is implemented. For example, one potential policy change is the proposed de-coupled allocation policy, which involves changing the current benefit loss (ranging from 3350%) upon spousal death to 40% in order to redistribute wealth and to help alleviate poverty in elderly widows. The analysis of the authors shows that this redistribution is more equitable for de-coupled allocation, of which the gross effect would not significantly increase Social Security payments. The redistribution also seeks to improve the financial condition of the American senior citizens who live below the poverty line.

This article divides the history of coming of age in the U.S. in the twentieth century into three distinct time periods: one that ran from 1880 until World War II, one that started in the 1940s and continued until the 1960s, and a third that began by 1970 and was clearly still in evidence in 1990. The story is based on data that recorded whether young people were living with one or both of their parents at the time of each of the decennial censuses of the U.S., from 1880 through 1990. The dramatic finding reported here is in the history of home-leaving ages up until the time of the second world war. Earlier studies assert that the age of home-leaving was declining from as early as it could be measured until 1970. These results are different. From 1880 until 1940 for males and 1950 for females, the age at leaving home did not decline, it rose. The decline came later, from 1940 (men) and 1950 (women). Beginning with 1970, the age of home-leaving rose again, reaching relatively high levels by 1990. The results also show that home leaving ages varied by race, sex, and region, among other factors.

This article divides the history of coming of age in the U.S. in the twentieth century into three distinct time periods: one that ran from 1880 until World War II, one that started in the 1940s and continued until the 1960s, and a third that began by 1970 and was clearly still in evidence in 1990. The story is based on data that recorded whether young people were living with one or both of their parents at the time of each of the decennial censuses of the U.S., from 1880 through 1990. The dramatic finding reported here is in the history of home-leaving ages up until the time of the second world war. Earlier studies assert that the age of home-leaving was declining from as early as it could be measured until 1970. These results are different. From 1880 until 1940 for males and 1950 for females, the age at leaving home did not decline, it rose. The decline came later, from 1940 (men) and 1950 (women). Beginning with 1970, the age of home-leaving rose again, reaching relatively high levels by 1990. The results also show that home leaving ages varied by race, sex, and region, among other factors.

This paper reviews the appropriateness for application to large data sets of standard machine learning algorithms, which were mainly developed in the context of small data sets. Sampling and parallelisation have proved useful means for reducing computation time when learning from large data sets. However, such methods assume that algorithms that were designed for use with what are now considered small data sets are also fundamentally suitable for large data sets. It is plausible that optimal learning from large data sets requires a different type of algorithm to optimal learning from small data sets. This paper investigates one respect in which data set size may affect the requirements of a learning algorithm-the bias plus variance decomposition of classification error. Experiments show that learning from large data sets may be more effective when using an algorithm that places greater emphasis on bias management, rather than variance management.