This dissertation addresses statistical issues related to endogenous peer selection in the context of social networks, social interaction models and snowball sampling methods. The first chapter studies the peer effects of friends, studymates, and seatmates on academic performance. We use data on social networks, personality traits, and cognitive ability measures collected through a unique survey conducted in three schools in Hong Kong. We estimate a social interaction model which accounts for endogenous network formation and correlation between multiple networks with a Bayesian approach. Our results show that the cognitive ability of studymates and the conscientiousness of friends positively affect a student's mathematics exam score while the conscientiousness of studymates and the cognitive ability of friends have no effect. The second chapter proposes a novel identification strategy for social interactions in a model with endogenously formed social networks. The network endogeneity arises from the correlation between the links of the network and the unobservables that determine the outcome of interest. We show that the eigenvectors of the adjacency matrix that defines the social network are control variables for network endogeneity without imposing any parametric assumption. We propose an information criterion to select the number of eigenvectors to be included as control variables. We apply the proposed method to the same empirical application as the first chapter and compare the results. The third chapter studies a snowball sampling method for social networks with endogenous peer selection. Snowball sampling is a sampling design which preserves the dependence structure of the network. It sequentially collects the information of vertices linked to the vertices collected in the previous iteration. The snowball samples suffer from a sample selection problem because of the endogenous peer selection. We propose a new estimation method that uses the relationship between samples in different iterations to correct selection. We use the snowball samples collected from Facebook to estimate the proportion of users who support the Umbrella Movement in Hong Kong.

This thesis analyzes estimation methods and testing procedures for handling large data series. The first chapter introduces the use of the adaptive elastic net, and the penalized regression methods nested within it, for estimating sparse vector autoregressions. That chapter shows that under suitable conditions on the data generating process this estimation method satisfies an oracle property. Furthermore, it is shown that the bootstrap can be used to accurately conduct inference on the estimated parameters. These properties are used to show that structural VAR analysis can also be validly conducted, allowing for accurate measures of policy response. The strength of these estimation methods is demonstrated in a numerical study and on U.S. macroeconomic data. The second chapter continues in a similar vein, using the elastic net to estimate sparse vector autoregressions of realized variances to construct volatility forecasts. It is shown that the use of volatility spillovers estimated by the elastic net delivers substantial improvements in forecast ability, and can be used to indicate systemic risk among a group of assets. The model is estimated on realized variances of equities of U.S. financial institutions, where it is shown that the estimated parameters translate into two novel indicators of systemic risk. The third chapter discusses the use of the bootstrap as an alternative to asymptotic Wald-type tests. It is shown that the bootstrap is particularly useful in situations with many restrictions, such as tests of equal conditional predictive ability that make use of many orthogonal variables, or `test functions'. The testing procedure is analyzed in a Monte Carlo study and is used to test the relevance of real variables in forecasting U.S. inflation.

Este artigo visa investigar os determinantes econométricos da oferta de trabalho voluntário no Brasil. Nas últimas décadas, foram realizados estudos empíricos buscando explicitar os fatores que levam os indivíduos a trabalhar voluntariamente em alguns países, mas nenhuma averiguação desse caráter foi feita entre os brasileiros ainda. De fato, pouco se sabe a respeito dos trabalhadores voluntários no Brasil. Utilizando dados da PNAD/IBGE, esta pesquisa mostra que eles eram cerca de 208 mil em 2012, o equivalente a 0,17% da população de 15 a 64 anos de idade no país, e trabalhavam em média 21 horas por semana. Os resultados do pooled Probit e Tobit indicam que pessoas com menores rendimentos potenciais do trabalho (custo de oportunidade) têm maior probabilidade de serem voluntárias e dedicam mais tempo ao voluntariado. A oferta de trabalho voluntário também aumenta com a renda domiciliar dos demais moradores e a escolaridade, e sua relação com a idade tem formato de U. Entre os homens, a renda não proveniente do trabalho é outro fator que exerce uma influência positiva sobre a decisão de despender algumas horas por semana em atividades voluntárias. Já entre as mulheres, a disponibilidade para trabalhar voluntariamente diminui com a presença de crianças no domicílio. De maneira geral, a magnitude dos efeitos obtidos através do modelo Probit é mais relevante. / This paper aims to investigate the econometric determinants of volunteer labor supply in Brazil. In recent decades, empirical studies trying to identify the factors that lead individuals to volunteer have been conducted in some countries, but no investigation of such character has been made in Brazil so far. Indeed, little is known about the nation\'s volunteer workers. Using data from PNAD/IBGE, this research shows that they were about 208 thousand in 2012, or 0.17% of the Brazilian population aged 15-64 years, and worked on average 21 hours a week. The pooled Probit and Tobit results indicate that people with lower potential wages (opportunity cost) are more likely to volunteer and devote more time to voluntary activities. Volunteer labor supply also increases with the income of other household members and higher levels of education. Furthermore, there is a U-shaped relation between volunteering and age. Among men, income from other sources than wages is another factor that has a positive influence on the decision to spend some hours a week volunteering. Among women, having a child at home diminishes the willingness to volunteer. In general, the magnitude of the effects of the Probit model is more relevant.

We provide empirical evidence of the TED spread as a risk factor in the cross-section of stock returns. Portfolios with high sensitivities to the TED spread have high average risk-adjusted returns. The pricing of TED spread risk is especially strong among small caps. TED spread is a usual measure of funding difficulties in interbank markets and our results are consistent with the Margin-CAPM model of Garleanu and Pedersen (2011). / Esta dissertação apresenta evidência empírica da TED Spread como um fator de risco na cross-section dos retornos de ações. Portfólios com elevada sensibilidade à TED Spread possuem elevados retornos médios ajustados para outros fatores de risco. O apreçamento do risco de TED Spread é especialmente forte entre small caps. TED Spread é uma medida usual de dificuldades de financiamento em mercados interbancários e o resultado obtido é consistente com o modelo Margin-CAPM de Gârleanu and Pedersen (2011).

Vector autoregression models have become widely used research tools in the analysis of macroeconomic time series. Cointegrated techniques are an essential part of empirical macroeconomic research. They infer causal long-run relationships between nonstationary variables. In this study, six information criteria were reviewed and compared. The methods focused on determining the optimum information criteria for detecting the correct lag structure of a two-variable cointegrated process.

The resource curse thesis states that countries with an abundance of natural resources tend to experience lower economic growth rates. However, does this theory apply also to the regional level? The purpose of this thesis is to both test the natural resources curse theory at the regional level in Sweden, as well to examine if different types of natural resource dependencies appear to have the same effect on regional income growth in the country. The methodological approach builds on an econometric (OLS) analysis using two different panel datasets over time intervals, 2000-2017 and 2007-2017, respectively. The results from one of the datasets suggest that the dependency on natural resources does not appear to affect Swedish counties' income growth, nor is there any difference in the impact on income growth from different types of natural resources. However, the corresponding results from the other dataset suggest a positive relationship, although these results are not robust across various model specifications. / Teorin om naturresursernas förbannelse förutspår att länder med ett överflöd av naturresurser kommer att ha en lägre ekonomisk tillväxt. Stämmer även denna teori på regional nivå? Syftet med denna studie är att undersöka om resursförbannelsen existerar på regional nivå i Sverige samt att undersöka om olika typer av naturresurser har olika påverkan på den regionala inkomstutvecklingen i landet. Metodiken för att besvara detta bygger på en ekonometrisk (OLS) analys med två olika paneldataset med skilda tidsintervaller, 2000–2017 och 2007–2017. De resultat som baseras på det ena datasetet visar att naturresurser inte har en påverkan på inkomstutvecklingen på svensk länsnivå, samt att olika typer av naturresurser inte heller har en påverkan på den regionala inkomstutvecklingen. Det andra datasetet påvisar dock att det finns ett positivt förhållande mellan regional inkomstutveckling och naturresurser, men detta resultat är inte robust.

No description available.

No description available.

No description available.

No description available.