In this article, I rely on new estimates of nineteenth-century mortality and the Integrated Public Use Microdata Series to construct new estimates of white fertility in the nineteenth-century United States. Unlike previous estimates that showed a long-term decline in overall fertility beginning at or before the turn of the nineteenth century, the new estimates suggest that U.S. fertility did not begin its secular decline until circa 1840. Moreover, new estimates of white marital fertility, based on "own-children" methods, suggest that the decline in marital fertility did not begin in the nation as a whole until after the Civil War (1861-1865).

Americans have long been loath to describe themselves in terms of class. Compared to the British, for example, Americans are far less likely to say that their society is composed of haves and have nots.2 In many respects, American culture is notably egalitarian; for centuries, foreign visitors have remarked on how little deference common folk give to their social betters here. They also used to note the political equality among citizens in a former era when the United States was exceptionally democratic. But egalitarian style and universal rights coexist with profound differences in economic resources. Americans are and have always been divided economically. Indeed, America in 2000 was the most economically divided nation in the developed world it had the widest spread in wealth. These divisions not only challenge the self-image of Americans as egalitarian, they have further consequences. Research shows that nations and communities with relatively wide disparities in material standards of living tend also to have relatively high rates of social problems, civic alienation, and discontent. In this paper, we examine differences in Americans standards of living in 2000 and how differences evolved over the twentieth century, especially since World War II. We describe how Americans in 2000 were separated by their material circumstances and how those distinctions compare to ones earlier in the twentieth century, taking into account peoples annual incomes, financial assets, consumption, and subjective assessments of their wealth. For each of these dimensions, we first contrast the well-being of better-off to less well-off Americans. Second, we distinguish the living standards of people of different ages, ancestries, educational levels, and locations. Over most of the century, economic divisions among Americans narrowed considerably in both these ways affluent and indigent became less distinct, as did Americans of different regions and races. But that the convergence stalled and then reversed in roughly the last three decades, widening the economic differences among Americans, most notably dividing by levels of education.

This book provides students with an in-depth examination of the world of work at the beginning of a new century. It compares womens and mens work status, addressing contemporary issues within a historical perspective, incorporating comparative material from countries other than the U.S., and recognizing differences in the experiences of women and men from different racial and ethnic backgrounds, drawing on both quantitative and qualitative data.

Association rule discovery has become one of the most widely applied data mining strategies. Techniques for association rule discovery have been dominated by the frequent itemset strategy as exemplified by the Apriori algorithm. One limitation of this approach is that it provides little opportunity to detect and remove association rules on the basis of relationships between rules. As a result, the association rules discovered are frequently swamped with large numbers of spurious rules that are of little interest to the user. This paper presents association rule discovery techniques that can detect and discard one form of spurious association rule: trivial associations.

Many scholars have noted the theoretical importance of remarriage in twentieth-century American life (Burch 1995; Cherlin 1998; Furstenberg 1980; Glick 1980; Thornton 1977; Uhlenberg and Chew 1986), yet few historical studies have examined remarriage in the United States empirically. This gap in the literature is noteworthy for two reasons. First, the turn of the twentieth century seems to have marked a crossover in the remarriage transition of the United States, reflecting changes in the pool of persons eligible resulting from declines in mortality and the probability of widow(er)hood, followed by increases in remarriage resulting from higher divorce rates. The crossover in the transition was likely to have occurred when the pool of eligibles was at or near its nadir. Second, there is ongoing debate about the implications of remarriage for families and individuals (Booth and Dunn 1994), and about the impacts of remarriage on family functions (Cherlin 1978; Cherlin and Furstenberg 1994). In the light of these considerations,we believe it is important to examine remarriage and its consequences in the United States at the turn of the century so that we may better understand the ways that remarriage influences family life and shapes the life course of persons within families (see London and Elman 2001). In this article,we use data from the 1910 Integrated Public Use Microdata Series (IPUMS) to provide a social demographic account of remarriage in the United States at the turn of the twentieth century. According to Uhlenberg and Chew (1986), the remarriage process has three components: eligibility (who enters into the risk pool for remarriage, how they do so, and at what age), the propensity to remarry (who actually remarries, given eligibility), and the multiple sequelae of remarriage for adults, children, families, and households. No historical study has systematically examined all three aspects of the remarriage process in the United States, at least in part because of data constraints. Similarly, the cross-sectional IPUMS is not well suited to examine the dynamism inherent in the first two stages of the remarriage process (i.e., which prime-age adults came to be at risk of remarriage and why a subset of those at-risk remarried). However, the IPUMS data are rich enough to allow for estimates of the point prevalence of remarriage and rate differentials by age, gender, and ethnicity/nativity, and to examine the consequences of remarriage for families and individuals within families. In this article, we focus on different remarriage configurations (i.e., which partner had remarried or whether both had) and their effects on two specific kin functionsresource gathering and housing dependents.

In this paper, we focus our attention on the college attendance decisions of women by race and specifically whether they respond to the rate of return. Our results suggest that Hispanic women increase the probability of college attendance if they are faced with higher rates of return, as do white women when regional variation is considered. Black women do not respond to the rate of return, at the traditional 10% level of confidence. We provide further evidence that early childbearing reduces the probability of attending college for black women, even after controlling for family and individual background characteristics.

In 1997 Odense City Museums, Denmark, presented an exhibition entitled 'Danes on the Prairie'. The exhibition aroused considerable interest both with the public and the press. A number of books, articles and TV and radio programmes followed in the wake of the exhibition, and in 2000 the present author published 'Dannebrog on the American Prairie'. Fascination with these Danish settlers, with their new country, integration and fortunes of their many descendants was only reinforced with the completion of the book. During the work with this study of a Danish colony, it became clear to the author that there was a need for a complete overview of Danish immigrants to USA and their descendants. In 1971 the historian Kristian Hvidt published his dissertation 'The Flight to America', based on a comprehensive statistical study of Danish emigrants and their geographical, economic and social background. A corresponding study of Danish immigrants to USA would be able to provide a unique understanding of this entire exodus and integration to a new country and would create an understanding of the Danish element in American society.In 2000 a concerted project was outlined with the title 'Danes in the USA, 1850-2000, - a demographic, social and human geographic study of Danish immigrants and their descendants'. The aim of the project is to create a complete overview of these Danish immigrants and the lives and lifestyles of their families and their descendants primarily by means of a systematic examination of the 10 yearly census. Among the questions raised are those of the immigration's chronology, its geographical dispersion or concentration in colonies, age, sex and family structure, progression and integration. Light is thrown on the last of these, for example, by means of a survey of linguistic ability and mixed marriages.At the end of 2000 the Carlsberg Research Foundation agreed to fund the project, which is estimated to take 4 years. Until the middle of 2003 the project is primarily focused on the collection of data from the census. Starting in the summer 2003 these will be analysed and worked on and results will be compared with the history of other immigrant groups. Finally, the last period of the project will be used to work on the final thesis. The analysis and study of the data will take place during a year's research leave, when the author will be affiliated to the University of Wisconsin-Madison.The article concludes with a selection of the provisional results of the project's statistical survey. Examples are given of the language abilities of these Danish immigrants, the distribution of men and women, where Utah clearly stands out from other areas, mixed marriages, and finally the degree to which Danes have remained stationary up to the present day in those areas where they first settled in the state of Wisconsin.

Discretization is a popular approach to handling numeric attributes in machine learning. We argue that the requirements for effective discretization differ between naive-Bayes learning and many other learning algorithms. We evaluate the effectiveness with naive-Bayes clas-sifiers of nine discretization methods, equal width discretization (EWD), equal frequency discretization (EFD), fuzzy discretization (FD), entropy minimization discretization (EMD), iterative discretization (ID), proportional k-interval discretization (PKID), lazy discretization (LD), non-disjoint discretization (NDD) and weighted proportional k-interval dis-cretization (WPKID). It is found that in general naive-Bayes classifiers trained on data preprocessed by LD, NDD or WPKID achieve lower classification error than those trained on data preprocessed by the other discretization methods. But LD can not scale to large data. This study leads to a new discretization method, weighted non-disjoint discretiza-tion (WNDD) that combines WPKID and NDD's advantages. Our experiments show that among all the rival discretization methods, WNDD best helps naive-Bayes classifiers reduce average classification error.

When the minimum wage was first enacted in 1938, the fiercest opposition came from the South, where wages were considerably lower than in the industrial North. Today, that opposition is found to emanate from states that have right-to-work laws (regardless of location). Using census data from the Integrated Public Use Microdata Series (IPUMS) for the years 1940 to 1990, this paper looks at workers earnings around the minimum wage by region and industry. It shows that, controlling for education and industry type, wages tend to be depressed more in states with right-to-work legislation than in high union density states. These effects on the wage structure have implications for the distribution of income.

We show that households spend their marginal and their average dollar dif- ferently across sectors. Crucially, marginal expenditure is biased toward sectors employing high-MPC workers, revealing a new redistribution channel that ben- efits high-MPC households during expansions. We build a Multi-Sector, Two- Agent, New Keynesian model with non-homothetic preferences consistent with these findings. The new redistribution channel increases the fiscal multiplier by 10pp compared to an equivalent homothetic economy. The model also predicts steeper Phillips curves in sectors with high-MPC workers, a result we validate empirically with a novel identification strategy. The implied sectoral wage dy- namics strengthen the redistribution towards high-MPC households and raise the inflationary impact of the shock by over 70 percent.

This dissertation consists of three independent essays, all of which are empirical treatments of different types of investment. The first essay evaluates the economic consequences of the successful eradication of hookworm disease from the American South in the early twentieth century. I find that reducing hookworm infection in this region brought about large increases in human capital and earnings. I then place these results in the context of contemporaneous questions about the economic burden of tropical disease. The second essay (joint with Kevin Cowan) examines the role that partial dollarization of debt may have played in recent emerging-market financial crises. Much has been written recently about the problems that result from 'mismatches' between foreign-currency denominated liabilities and assets (or income flows) denominated in local currency. Specifically, it is supposed that the expansion in the 'peso' value of 'dollar' liabilities resulting from a devaluation could, via a net-worth effect, offset the expansionary competitiveness effect. Our results suggest that, for this sample of firms in these episodes, this net-worth channel is likely small in comparison with the more traditional competitiveness effect. The final essay (joint with Aimee Chin) considers the role that English-language skill plays in the economic performance of immigrants to the United States. This study exploits the fact that younger children tend to learn languages more easily than adolescents or adults to construct an instrumental variable for English proficiency. We find that low English proficiency significantly lowers earnings and educational attainment. Indeed, much of the effect of language skills on wages in our sample appears to be mediated by years of schooling.

This paper analyzes how markets for old-age care respond to the aging of populations. We consider how the biological forces, which govern the stocks of frail and healthy persons in a population, interact with economic forces which govern the demand and supply for labor-intensive care. Many economists have argued that aging will raise the market demand for long-term care, and hence price and quantity through classic market effects. We argue that the direct effect of aging is to lower the demand for market care by increasing the supply of home production. By influencing the length of frail lifetimes, aging may have a further indirect effect, which may reinforce or counteract the direct negative demand effect. By providing healthy spouses, the marriage market provides care-givers for home production of long-term care; therefore, growth in old-age longevity may lower the demand for market production. Growth of elderly males serves to contract the long term care market because it eases the scarcity of men in the old-age marriage market; growth of females serves to expand the market care because it worsens the scarcity of men. These predictions lend themselves to an interpretation of the rapid deccelration in output growth that has taken place in the US over the last two decades, despite a constant rate of longevity growth, and enormous growth in demand subsidies: since growth in elderly males has risen dramatically relative to growth in elderly females, the rate of long-term care growth has slowed significantly. We test our predictions empirically, using state- and county-level evidence on the US market for long-term care in nursing homes over the past three decades. The evidence provides support for, among other things, the predictions we offer concerning the response in output growth to aging and the contraction of output due to the aging of males.

This paper provides an overview of the mushrooming economics literature on how community attributes influence the level of civic engagement. Since 1997, at least fifteen empirical papers have investigated the consequences of heterogeneity for social capital. Social capital has been measured using indicators of group participation such as volunteer activity, organizational membership and activity, entertaining and visiting friends and relatives, and voting and indicators of the strength of network ties such as trust. These papers cover different nations, different social capital measures, and even different centuries. But a common theme emerges across these fifteen studies. More homogeneous communities foster greater levels of social capital production. We provide an overview of this literature and then focus on synthesizing our past work on volunteering and membership with new findings on trust and voting.

This paper looks at the fertility, mortality, and marriage experience of racial, ethnic, and nativity groups in the United States from the 19th to the late 20th centuries. The first part consist of a description and critique of the racial and ethnic categories used in the federal census and in the published vital statistics. The second part looks at these three dimensions of demographic behavior. There has been both absolute and relative convergence of fertility across groups. It has been of relatively recent origin and has been due, in large part, to stable, or even slightly increasing, birth rates for the majority white population combined with declining birth rates for blacks and the Asian-origin, Hispanic-origin, and Amerindian populations. This has not been true for mortality. The black population has experienced absolute convergence but relative deterioration in mortality (neonatal and infant mortality, maternal mortality, expectation of life at birth, and age-adjusted death rates), in contrast to the Amerindian and Asian-origin populations. The Asian-origin population actually now has age-adjusted death rates significantly lower than those for the white population. The disadvantaged condition of the black population and the deteriorating social safety net are the likely origins of this outcome. Finally, there was a trend toward earlier and more extensive marriage from about 1900 up to the 1960s. At this point, coincident with the end of the 'Baby Boom,' there has been a movement to later marriage for both males and females among whites, blacks, and the Hispanic-origin populations. This trend has been more extreme in the black population, especially among females. There has also been a significant rise in proportions never-married at ages 45-54 among blacks and, to a lesser extent, among Hispanics. So here too, there has been some divergence.

Many states in the US have considered adopting ‘English-only’ legislation since the early 1980s. This paper argues that this legislation has a variety of expected and unexpected outcomes. Using the 1980 and 1990 Integrated Public Use Microdata Series (IPUMS), this study finds that Asian immigrants in general and Hispanic immigrants who arrived to the US prior to 1970 acquired more English fluency on average during the 1980s when residing in states that passed English-only (EO) legislation compared to their similar peers. The passage of such legislation also appears to have distorted residence decisions as well as the relative quality of immigrants between EO and non-EO states.

Background : While the inheritance of eye colour is likely polygenic, blue eye colour is thought to follow an inheritance pattern similar to that of a recessive trait. Consequently, age-related differences in the prevalence of blue eye colour would be unanticipated. Aim : This study explores the finding and explanation for birth cohort differences in the prevalence of blue eye colour in the US white population. Subjects and methods : Data from the first (1971-1975) and third (1988-1994) US National Health and Nutrition Examination Surveys (NHANES-I and NHANES-III), nationally representative surveys of the US population, were analysed. Trends in eye colour prevalence by birth cohort were analysed together with mortality rates according to eye colour. US census data (1980) were examined to explore cohort differences in ancestry and assortative mating by ancestry. Results : The prevalence of blue eye colour among non-Hispanic whites in NHANES-III was 57.4% (95% CI: 50.1-64.7) for individuals born between 189...