For a large number of data management problems, it would be very useful to be able to obtain a few samples from a data set, and to use the samples to guess the largest (or smallest) value in the entire data set. Min/max online aggregation, top-k query processing, outlier detection, and distance join are just a few possible applications. This paper details a statistically rigorous, Bayesian approach to attacking this problem. Just as importantly, we demonstrate the utility of our approach by showing how it can be applied to two specific problems that arise in the context of data management.

Secondary data is the data that researchers do not create themselves but use in their research. Compared to primary data that is generated over the course of fieldwork (e.g. measuring water quality or interviewing respondents), “secondary” data is already created by someone else. Secondary data providers include government agencies and private companies or such sources as published scientific studies, archives, or collections. Most commonly, the term “secondary data” refers to relatively large databases that individual researchers would not be able to gather themselves (e.g. census data, newspaper archives, inventories of resources, or satellite imagery). Although called “secondary,” this data informs a great deal of academic work and is central to entire subdisciplines in the social and environmental sciences. Moreover, the importance of secondary data in research and policy development is likely to increase with time. This is because information technologies have facilitated an explosion of a wide range of both environmental and socio-economic digital information as well as methods for its analysis. Widely available and accepted as legitimate, secondary data has come to influence in important ways what kind of knowledge we produce and how. The ubiquity of secondary data, especially within the global north, demands that we carefully evaluate its potentials and limitations before integrating it into any research project or using it to answer specific research questions. This chapter addresses some of the issues related to the use of secondary data by geographers. We point to the wide variety of secondary data and its many sources and we discuss the important advantages and limitations of secondary data. We then address issues of particular importance to geographers: ecological fallacy and the Modifiable Areal Unit Problem (MAUP) as they relate to secondary data. Finally, we illustrate the need to engage creatively and critically with secondary data by focusing on non-standard approaches to analysis that use “mixed” research methods. Throughout the chapter, we will use examples from our and our students’ work in urban geography (cases from Moscow and New York) as well as resource management (the case of fisheries in the Northeast US).

Child sexual abuse rates have fallen dramatically in the United States since the early 1990s. Between 1992 and 2000, substantiated sexual abuse reports dropped from 150,000 to 89,500 cases, a decline of 40 percent(Finkelhor and Jones 2004). Much of this decline may be attributable to a declining pool of older, previously unreported cases of child abuse that came to light in the late 1980s with rising public awareness but were largely investigated by the mid-1990s. However, similar declines in self-reported sexual abuse provide evidence that at least a portion of the change represents a real decline in child sexual abuse, rather than increased caution on the part of Child Protective Services (CPS) agencies that investigate and substantiate abuse allegations. Despite this encouraging national trend inchild sexual abuse, administrative data from Texas suggestthat the share of Latin American immigrant children in out-of-home care who were removed for sexual abuse isthree times as high as the share of children of nativesremoved for sexual abuse.

This study was centrally concerned with the role of immigration and ethnicity within discussions of interracial marriages. One of the goals was to examine the relationship between assimilation and intermarriage. Unlike many studies on intermarriage that focus on black-white couples, this study differed in three important ways: 1) I focused on immigrants from Asia and Latin America and their U.S.-born children, 2) I made a key distinction between interethnic and interracial relationships, and 3) I included not only married couples, but also cohabiting and dating couples. The questions that this research addressed were: Who enters into mixed relationships and why?, and 2) What is the role of mixed relationships in the overall processes of assimilation (often described as segmented or straight-line approaches)? I used the Current Population Survey, 2003-2006, and the Children of Immigrants Longitudinal Study, composed of survey questionnaires and 134 in-depth interviews. While race was important for these children of immigrants, ethnicity was a more important factor. The most common ways in which the respondents expressed cultural differences was in the form of language and religion. Thus, the lived experience of being in an interethnic versus an interracial relationship was different. Over time males and females were more likely to find themselves in coethnic relationships as they imagined a future of marriage and the role that extended family would play. These young adults talked about parental prejudices, language, religion, and other cultural clashes as the major factors they took into account. There were many females, however, who did not follow this pattern because of perceptions of patriarchy. They avoided coethnic relationships because they wanted a partner who would think of them as an equal rather than subordinates in a patriarchal relationship. Finally, in terms of assimilation, I argue that segmented paths of assimilation more closely represent the experiences of the respondents. The couples are part of institutions, such as higher educational institutions, are getting good jobs, and are becoming involved in politics and voting; yet they still have preferences for being with certain types of peoples when it comes to intimate relationships.

In the mid-nineteenth century, almost 70 percent of persons age 65 or older resided with their adult children; by the end of the twentieth century, fewer than 15 percent did so. Many scholars have argued that the simplification of the living arrangements of the aged resulted primarily from an increase in their resources, which enabled increasing numbers of elders to afford independent living. This article supports a different interpretation: the evidence suggests that the decline of coresidence between generations had less to do with the growing affluence of the aged than with the increasing opportunities of the younger generations. Using data from the Integrated Public Use Microdata Series (IPUMS), I examine long-run trends in the characteristics of both the older and the younger generations to gain insight into changing motivations for coresidence. In particular, I investigate headship patterns, occupational status, income, and spatial coresidence patterns. I also reassess the potential impact of the Social Security Program. I conclude that the decline of intergenerational coresidence resulted mainly from increasing opportunities for the young and declining parental control over their children.

When racial groups di ffer in terms of their political sensitivity-the propensity of group members to change their vote based on changes in redistribution promised by acandidate-the racial composition of a political jurisdiction a ects redistributive policy even when preferences are correlated only with income. Furthermore, the extent to which race and class are cross-cutting is important. Although the concept of cross-cutting cleavages is well-known in political science and sociology, its implications have yet to be analyzed formally. This paper formally defi nes, develops and tests a model of cross-cutting cleavages to better understand the complex relationships between race, class, preferences, political sensitivity and redistributive policy. I show that the effect of cross-cutting cleavages on redistributive policy depends critically on economic parity-the ratio of average black income to average income. The cross-cutting cleavages model makes six predictions that are all supported by data from large U.S. cities between 1942 and 1972.

In the mid-nineteenth century, almost 70 percent of persons age 65 or older resided with their adult children; by the end of the twentieth century, fewer than 15 percent did so. Many scholars have argued that the simplification of the living arrangements of the aged resulted primarily from an increase in their resources, which enabled increasing numbers of elders to afford independent living. This article supports a different interpretation: the evidence suggests that the decline of coresidence between generations had less to do with the growing affluence of the aged than with the increasing opportunities of the younger generation. Using data from the Integrated Public Use Microdata Series (IPUMS), I examine long-run trends in the characteristics of both the older and the younger generations to gain insight into changing motivations for coresidence. In particular I investigate headship patterns, occupational status, income, and spatial coresidence patterns. I also reassess the potential impact of the Social Security program. I conclude that the decline of intergenerational coresidence resulted mainly from increasing opportunities for the young and declining parental control over their children.

La question de savoir si les taux relatifs de mariages homogames ou assortatifs augmentent dans les démocraties riches fait l’objet d’une importante controverse. Premièrement, nous montrons que les résultats empiriques conflictuels qui alimentent le débat sont souvent un artéfact d’autres stratégies méthodologiques utilisées pour répondre à la question. Puis, en nous appuyant sur les données de recensement comparables pour le Canada et les États-Unis, nous examinons les tendances au chapitre de l’homogamie éducationnelle et du mariage de personnes de niveaux d’études différents au moyen de modèles log-linéaires pour tous les mariages chez les jeunes adultes de moins de 35 ans au cours de trois décennies. Nos résultats montrent une augmentation non ambiguë de l’homogamie éducationnelle, soit la tendance des personnes à épouser des personnes de même niveau d’études, dans les deux pays depuis les années 1970 sans signe du revirement dans les niveaux de mariage de personnes de niveaux d’études différents dont ont fait état certaines études comparatives antérieures. Les niveaux d’homogamie maritale à la hausse sont attribuables à la diminution des mariages de personnes de niveaux d’études différents aux deux extrémités de la distribution des niveaux. Toutefois, si les tendances pour les hommes et les femmes sont fort semblables au Canada, elles diffèrent sensiblement aux États-Unis. En effet, aux États-Unis, l’augmentation globale de l’homogamie maritale est contrebalancée en partie par une tendance plus marquée chez les femmes ayant fait des études collégiales partielles à épouser un homme de niveau inférieur dans la hiérarchie éducationnelle. Au Canada, le seul signe d’affaiblissement de la tendance à une plus grande homogamie éducationnelle est une légère augmentation des mariages d’hommes et de femmes ayant fait des études universitaires avec des personnes de niveaux d’études différents durant les années 1990.

This dissertation is quantitative in its orientation and utilizes multilevel analysis as well as geographic maps and descriptive tables to explore how the intersection of race and gender functioned to create the economic disadvantage experienced by black women over the course of a century. The data for this analysis is drawn from the Integrated Public Use Microdata Series (IPUMS).

This article seeks to describe and explain variation in voter turnout in American big city municipal elections using data from 332 mayoral elections in 38 large U.S. cities over 25 years. In my cross-sectional time-series analysis of turnout in mayoral elections, I find that city-level demographic factors are only weakly correlated with turnout. By contrast, institutional and campaign factors explain much of the variation. The effect of Progressive era reforms on depressing turnout is greatest in the most competitive elections. I conclude by discussing the implication of the overall downward trend in turnout and changes cities can make to increase participation.

The exodus of millions of southerners between 1910 and 1970 is largely attributed to the economic and social push factors in the south combined with pull factors in other regions. Researchers generally find that participants in the Great Migration were positively selected from their origin, in terms of educational attainment and urban status. Although a considerable amount of attention examines how these migrants fared in their destinations, to fully measure the success of migrants, a comparison to those who remained within the south is necessary. This paper compares inter-regional migrants (i.e. migrants who left the south) with their southern contemporaries they left behind. Our findings indicate that although positively selected, migrants who left the south did not benefit appreciably in terms of employment status, income, or occupational status. In fact, in many instances, inter-regional migrants fared worse than did southerners who moved within the south or who remained sedentary.

Skala migracji jest duża i ciągle rośnie, choć pozostaje niższa niż rozmiary handlu i przepływy kapitału, za sprawą czego migracje mogą być postrzegane jako obiecujący sposób materializacji ogólnoświatowych korzyści gospodarczych. Zdaniem autora, większa mobilność siły roboczej w skali międzynarodowej mogłaby podwyższyć produkcję i poprawić położenie ekonomiczne pracowników w krajach rozwijających się bardziej aniżeli innego rodzaju polityka związana z procesem globalizacji. Pewne korzyści przypadłyby w udziale także krajom przyjmującym imigrantów. Jednak z uwagi na potencjalne napięcia gospodarcze i kulturalne, jakie rodzi imigracja, jego zdaniem jest mało prawdopodobne, aby kwestia swobody migrowania uzyskała choć część poparcia, jakim cieszy się wolny handel. Zważywszy na to, iż korzyści związane z migracją przypadają głównie imigrantom pochodzącym z krajów o niskich dochodach, skłonienie obywateli krajów wysoko rozwiniętych do przychylniejszego spojrzenia na imigrację będzie wymagało przede wszystkim opracowania polityki, która pozwoli krajom otwierającym się na imigrantów czerpać większe korzyści z tego tytułu.

Using the U.S. Census Bureaus County-to-County MigrationFile and Public Use Microdata Sample File, we investigatedCalifornias interstate migrants between 1995 and 2000. Aftermapping and discussing the spatial patterns of migrants origin and destination counties, we calculated percentages of college graduates and detailed industries of employment for Non-Hispanic White, Latino, Black, and Asian migrants. To examine migrants social and economic fit at their destinations, we compared them to longer-term residents of both California and other leading destinations. In all cases inmigrants to California had higher percentages of college graduates than did their ethnic group in origin states, and in nearly all cases inmigrants raised the percentage of college graduates of their ethnic group in California. Migrants from California worked in destination-state industries at similar or slightly lower rates than non-migrant residents, but migrants to California were better represented inhigher-status industries than longer-resident Californians.

This paper studies the effect of state-level banking regulation on financial development and on components of state-level growth in the United States from 1900 to 1940. We use these banking laws to assess the findings of a large recent literature that has argued that financial development contributes to economic growth. We contend that the institutional mechanism leading to financial development is important in determining its consequences and that some types of financial development can even retard economic growth. For the United States from 1900 to 1940, we argue that the financial expansion induced by expanded bank branching accelerated the mechanization of agriculture and spurred growth in manufacturing. In contrast, financial expansions induced by state deposit insurance had negative consequences for both the agricultural and manufacturing sectors.

The goal of this paper is to build on the growing body of research on immigration and crime in two important ways. The first is toemploy more specific measures of immigration than have been used in previous analyses. Specifically, this analysis includes measuresof ethnicity, indicators that contain information about both nativity and country of origin, which have rarely been used in priorresearch. Using ethnic-origin as a means of classifying a neighborhood's foreign-born population will promote a more nuancedunderstanding of the differential impacts of immigration on levels of violent criminal offending. Additionally, this research advancescurrent knowledge on the link between immigration and crime by using more comprehensive crime indicators, including measures ofnon-lethal violence, which allows for a test of the degree to which the impact of immigration on violence varies across crime types.Using data for Miami and Houston, two immigrant destination cities, the results illustrate the need for researchers to be sensitive toethnic differences among foreign-born populations. The findings support the calls for a refinement of the disorganization theory, onethat is sensitive to the differences among the foreign-born population and one that does not assume that immigration is a causallyassociated with levels of criminal violence.

Toyboy marriages (where the female partner is at least 5 years older than her male partner) have grown threefold since the 1970s in the United States and Britain. This paper examines this phenomenon using an equilibrium search framework in which becoming successful in the labour market takes time and fitness decays with age. Our framework hinges on contract incompleteness in the marriage market and the assumption that the marginal gain to marrying someone rich is greatest for someone poor. With this structure we can explain why successful (older) types might marry fitter (younger) and less successful types. We show that toyboy marriages arise in equilibrium only when men and women have comparable labour market opportunities. U.S. and British data confirm this indicating that the probability that a woman is married to a toyboy increases by about 45 percent if, relative to her partner's, she is more educated and in a better paid job.

This paper explores new opportunities provided by the ongoing revolution in information technology, computer science, and Internet expansion for studies of exceptional human longevity. To this aim, the detailed family data for 991 alleged centenarians born between 1875 and 1899 in the United States were extracted from publicly available computerized family histories of 75 million individuals available at the Rootsweb site. To validate the age of the centenarians, these records were linked first to the Social Security Administration Death Master File records (for death date validation) and then to the records of the U.S. censuses for 1900, 1910, and 1920 (for birth date validation). The results of this cross-validation study demonstrated that computerized genealogies may serve as a useful starting point for developing a reliable family-linked scientific database on exceptional human longevity. The resulting database on centenarians with validated ages was used in the study of the predictors of exceptional human longevity, including familial factors and early-life living conditions. The comparison of households where children (future centenarians) were raised (using data obtained through linkage of genealogies to early U.S. censuses) with control households drawn from the Integrated Public Use Microdata Series for the 1900 U.S. census suggests that a farm background (farm ownership by parents in particular) and childhood residence in the Western region of the United States may be predictive for subsequent survival to age 100. These findings are consistent with the hypothesis that lower burden of sickness during childhood (expressed as lower child mortality in families of farm owners and families living in the West) may have far-reaching consequences for survival to extreme old ages. Analysis of familial factors suggests that there may be a link between exceptional longevity and a persons birth order. It was found that first-born daughters are three times more likely to survive to age 100, compared to later-born daughters of higher birth orders (7). First-born sons are twice more likely to become centenarians compared to sons having birth order between four and six. Further within-family comparison of centenarians with their siblings found that the protective effect of being first-born is driven mostly by the young maternal age at the persons birth (being born to a mother younger than 25 years). Being born to a young mother is an important within family predictor of human longevity, and even at age 75 it is still important to be born to young mother to survive to 100 years.