Accuracy of regularized D-rule for binary classification

We consider a regularized D-classification rule for high dimensional binary classification, which adapts the linear shrinkage estimator of a covariance matrix as an alternative to the sample covariance matrix in the D-classification rule (D-rule in short). We find an asymptotic expression for misclassification rate of the regularized D-rule, when the sample size $n$ and the dimension $p$ both increase and their ratio $p∕n$ approaches a positive constant $\gamma$. In addition, we compare its misclassification rate to the standard D-rule under various settings via simulation.     

Free versus regulated migration: Comparing the wages of the New Zealand-born,other migrants and the Australia-born workers in Australia

New Zealanders can cross borders freely, work and live in Australia indefinitely thanks to the Trans-Tasman Travel Agreement. This paper uses a recently developed decomposition method to decompose the weekly wage gap at various quantiles on the wage distribution between New Zealand-born (NZ-born) and Australian-born workers, and between NZ-born workers, migrants from other English speaking countries (OESC), and from non-English speaking countries (NESC) to determine how free and regulated migration influences migrants’ performance in the Australian labour market. We found that NZ-born workers earned higher weekly wages than both Australian-born and NESC workers but earned lower wages than OESC migrants. Differences in endowment were primarily responsible for the wage gaps between NZ-born and Australian-born workers and between NZ-born and OESC migrants. However, differences in returns to worker and job characteristics are mainly responsible for the wage gap between NZ-born and NESC migrants.     

Global poverty estimates based on 2011 purchasing power parity: where should the new poverty line be drawn?

This paper aims to present a new global poverty line based on the 2011 Purchasing Power Parity (PPP). To calculate a new global poverty threshold based on 2011 PPP, this paper moves away from the World Bank’s method of anchoring a single global poverty line on the national poverty lines of the 15 poorest countries. It instead proposes an alternative method of using equivalent poverty lines. Each country is shown to have a different equivalent poverty line. This paper finds that there is no single international poverty line in 2011 PPP that is equivalent to $1.25 in 2005 PPP. Single poverty lines vary for each region because countries have experienced different inflation rates and have different PPP conversion rates between 2005 and 2011. To calculate a single poverty line in 2011 PPP, this paper measures the weighted average of equivalent poverty lines of 101 countries around the globe with weights proportional to their populations. Based on the new method, the corresponding poverty line is estimated at $1.93 in 2011 PPP. The World Bank has officially adopted the poverty line of $1.90 in 2011 PPP. This paper demonstrates that our proposed poverty line performs better than the World Bank’s in terms of preserving the real purchasing power of the previous line of $1.25 in 2005 PPP. Given the new poverty threshold of $1.93, the number of poor worldwide is reduced by 6.42 million, with the reduction largely occurring in South Asia and Sub-Saharan Africa.     

Measuring the impact of price changes on poverty

This paper develops a methodology to measure the impact of price changes on poverty measured by an entire class of additive separable poverty measures. This impact is captured by means of the price elasticity of poverty. The total effect of changes in price on poverty is explained in terms of two components, income and distribution effects. The income effect measures the change in poverty when all prices increase uniformly, whereas the distribution effect captures the change in poverty because of changes in relative prices. Using this decomposition, the paper derives an empirically operational index, which reveals whether the price changes have been pro-poor or anti-poor. The paper also derives a new price index for the poor. While this index can be computed for any poverty measures, our empirical analysis applied to Brazil is based on three poverty measures, the head-count ratio, the poverty gap ratio and the severity of poverty. The empirical analysis shows that price changes in Brazil during the 1999–2006 periods have been anti-poor. Nevertheless, during the last 2 to 3 years, the price changes have affected the poor less adversely than the non-poor.     

In Social Capital We Trust?

It has been widely postulated in the literature that social capital is positively related to or is the same as trust. The present study presents three theoretical perspectives regarding how individual and organizational social capital may be related to network trust and generalized trust: the compositional element (Putnam), functional equivalence (Fukuyama), and mutual independence (Lin). To each of these perspectives we allocated distinct measures of individual and organizational network mechanisms considering their operational definition of social capital. Using nationally representative data sets from the United States and China, we developed a comparative research design through which the three perspectives were put into an empirical test. In conclusion, we found that the compositional element perspective is most prone to cross-national contingencies, whereas the mutual independence perspective is freer from such contextual influences. In particular, the positive association between number of membership in voluntary associations (organizational social capital) and generalized trust exists only in the United States, whereas individual social capital based on network diversity and resources is unrelated to generalized trust in neither country. Lastly, the functional equivalence perspective does not get empirical support from the data, particularly with regard to the proposed negative relationship between network closure and generalized trust. These findings challenge the assumptions concerning the positive association or equivalence between social capital and trust in the literature.

     

Promoting older women's health and well-being through social leisure environments: what we have learned from the Red Hat Society

The purpose of this study was to describe the ways that participation in a leisure organization contributed to the health and wellbeing of middle-aged and older women. We analyzed 1,693 members' responses to a query about meaningful experiences garnered through participation in the Red Hat Society. Results suggested that older women's lives have been enriched and changed by their experiences, with the women in this study citing multiple psychosocial health benefits from their participation in the Red Hat Society. Main themes encapsulating these health benefits were creating happy moments, responding to transitions and negative events, and enhancing the self. These findings are related to research on positive psychology, social support and coping, transformative leisure processes, and social identity formation. We conclude by providing suggestions for applying these findings to leisure and health promotion programming to enhance women's health and well-being in later life.     

Cross-national comparison of social support structures between Taiwan and the United States

The study compares the effects of structural bases and functional elements of social support on mental health in Taiwan and the United States, using the study conducted in the United States by Lin, Ye, and Ensel (1999) as a reference. Based on a nationally representative sample of Taiwanese adults (n = 2,835), a fundamental similarity in social support structure and function between the two countries was observed. First, the structural bases of social support had a hierarchical order in their effects on depression: Binding (presence of an intimate relationship) was the strongest in reducing depression, whereas belonging (community participation) was the weakest, with bonding (social networks) in between. Regarding the functional elements, perceived social support was a better protector of mental health than actual social support, a finding in line with previous research. On the other hand, several notable differences in the structural bases and functional elements of social support between the two societies were observed, possibly due to the differential cultural and historical characteristics.     

GENETIC VARIATION IN VOLUNTEERISM

Research has shown that prosocial behaviors of various kinds are passed from generation to generation, but the role played by genetics in the transmission of volunteerism has been unexplored. Data from the Midlife in the United States (MIDUS) Twins and Siblings samples are used to estimate genetic heritability of hours volunteered per month. Although unique environmental factors of the kind sociologists have traditionally focused upon account for most of the variance, women do owe some of their disposition to perform volunteer work to their genes. There is no genetic effect for men.     

Bayesian Estimation of the Two-Parameter Gamma Distribution

In this article, we present and discuss an original price index being a special case of a general formula for price indices. We show that the discussed formula satisfies most postulates coming from the axiomatic price index theory. We compare this index to other known and popular price indices in a simulation study.     

A New Poverty Decomposition

This note proposes a new poverty decomposition that can be used to explain changes in poverty over time. The change in poverty is derived as the exact sum of four elements: (i) the overall growth effect, assuming inequality in the distribution does not change; (ii) the impact of differences in growth rates between the groups; (iii) the effect of the change in inequality within the different groups; (iv) the impact of changes in the population shares of the various groups.