Submitted by Renata Lopes (renatasil82@gmail.com) on 2017-04-04T18:11:10Z No. of bitstreams: 1 diegodasilvarodrigues.pdf: 406421 bytes, checksum: aeaa7de595b7c22c154883eda85bdc86 (MD5) / Approved for entry into archive by Adriana Oliveira (adriana.oliveira@ufjf.edu.br) on 2017-04-06T10:43:39Z (GMT) No. of bitstreams: 1 diegodasilvarodrigues.pdf: 406421 bytes, checksum: aeaa7de595b7c22c154883eda85bdc86 (MD5) / Made available in DSpace on 2017-04-06T10:43:39Z (GMT). No. of bitstreams: 1 diegodasilvarodrigues.pdf: 406421 bytes, checksum: aeaa7de595b7c22c154883eda85bdc86 (MD5) Previous issue date: 2009-12-21 / Esse trabalho tem por finalidade estimar a probabilidade de migração interestadual dos trabalhadores que se destinam ao setor informal. Especificamente, busca-se analisar a probabilidade de migração conforme o nível de renda dos estados de destino. Esse objetivo é motivado pelo fato do Brasil apresentar intensa migração interna, o que leva à necessidade de compreender se as características dessa migração são as mesmas se considerarmos apenas o setor informal, que apresenta peculiaridades. Para isso, será elaborado um modelo probabilístico de migração com base em um banco de dados montado a partir da Ecinf (IBGE/2003). Inicialmente, é estimado um modelo probit simples, visando entender o impacto de características observadas pessoais, de trabalho e da região na decisão do indivíduo migrar. Depois, é estimado um modelo multinomial, buscando entender o impacto que essas características observadas têm quando o destino da migração é diferenciado pelo nível de renda dos estados. Os principais resultados obtidos indicam que, entre os informais, a migração segue características semelhantes às observadas na literatura, como ser mais propensa entre as mulheres, e apresentar renda maior entre os migrantes em comparação com os não-migrantes, sendo esta uma variável importante para a migração às regiões mais ricas. Por outro lado, os resultados também mostram que o aumento do nível de instrução tende a diminuir a probabilidade de um trabalhador informal migrar, indo de encontro ao que se observa noutros mercados de trabalho. / This paper aims to estimate the probability of interstate migration of informal workers. More specifically, it has the objective to analyze the probability of migration according to the income level of the destination states. This goal is motivated by the fact that Brazil has a strong internal migration, which leads to the need of understanding if the characteristics of the internal migration are the same if one considers only the informal sector, which presents peculiarities. This way, a probabilistic model of migration is made based on a database from the Ecinf (IBGE/2003). Initially, it is estimated a simple probit model, in order to understand the impact of observed personal, job and regional characteristics on the individual's decision to migrate. After that, it is estimated a multinomial model, trying to understand the impact that these observed characteristics have when the destination of the migration is differentiated by the income level of the states. The main results show that, among informal workers, the migration has characteristics similar to those observed in the literature and in the proposed model, as being more likely among women, and presenting higher incomes among the migrants when compared with non-migrants, and that being an important variable for migration to richer regions. On the other hand, the results also show that, among the informal workers, increasing of the educational level of the individuals tends to reduce the probability to migrate, against what is observed in other job markets.

Obesity is a growing public health concern that is being confronted by both developed and developing countries. South Africa is no exception, facing the highest burden of obesity amongst African countries. Using two waves of data from the National Income Dynamics Study, this study aims to investigate the relationship between obesity and employment status for working age individuals in the context of the South African labour market. This study contributes to existing literature on this subject by explicitly accounting for potential simultaneity and endogeneity between obesity and employment. Given the hypothesised two-way causal relationship between obesity and unemployment, two different models are used to assess whether this issue exists for the dataset; a bivariate probit model to assess if there is a bivariate relationship between obesity and employment, and a recursive bivariate probit model to assess if obesity is an endogenous regressor of employment. A change in state univariate probit model is then implemented across the two waves to better understand if fluctuations in weight status are a result of labour market state transitions. The results of the study show that obesity and employment are independent in the bivariate probit models and obesity is an exogenous regressor of employment status in the recursive bivariate probit models. Changes in labour market state do not have a significant impact on the probability of transitioning to obese compared to no changes in labour market state, bar transitioning from not economically active to employed, which increases the probability of becoming obese. The findings suggest that, in the South African labour market context, obesity and employment are not related, indicating that there are other underlying factors, such as nutritional intake and genetic composition, that may contribute to fluctuations in weight status. The results suggest that obesity is prolific in South Africa, and impacts individuals across the entire distribution for labour market status and income.

Master of Arts / Department of Economics / William F. Blankenau / This paper describes the socioeconomic determinants of school dropout for Vietnamese children aged 11-18. It seeks to answer the following two questions: 1) What is the magnitude of the dropout rate for children between the ages of 11-18 and how does this figure change over time? 2) What are the determinants that affect the decision to drop out of school? This paper is based largely on the two Vietnam Living Standard Surveys (VLSS) conducted in 2004 and 2006. A probit model is used to ascertain the major determinants affecting the probability to drop out of school. The decision to drop out of school is affected by key factors including the child’s demographic characteristics (such as age, gender, ethnicity, and regions where they are living) and the child’s household’s situation (such as income per capita, cost of schooling, household size, and parental education). In particular, age and household size have significantly positive effects on the dropout probability. The dropout rate is also shown to vary between girls and boys, but this gender gap has narrowed substantially. Minority girls face more obstacles in staying school than minority boys. The school dropout rate is also very sensitive to changes in the household’s income and cost of schooling. However, the cost of schooling has different impacts on families in different quintiles. Region is another determinant affecting child’s decision to drop out of school. Vietnam’s population is unequally distributed in 8 regions with different socio-economic conditions, and hence the dropout situation is also regionally specific. Moreover, the parents’ perception of the value of education may increase the child’s probability of school retention. Since it is difficult to measure parental attitude to schooling, the paper uses parents’ education level instead, assuming that parents who have more education will appreciate education more.

Feature selection is demanded in many modern scientific research problems that use high-dimensional data. A typical example is to find the genes that are most related to a certain disease (e.g., cancer) from high-dimensional gene expression profiles. There are tremendous difficulties in eliminating a large number of useless or redundant features. The expression levels of genes have structure; for example, a group of co-regulated genes that have similar biological functions tend to have similar mRNA expression levels. Many statistical methods have been proposed to take the grouping structure into consideration in feature selection and regression, including Group LASSO, Supervised Group LASSO, and regression on group representatives. In this thesis, we propose to use a sophisticated Markov chain Monte Carlo method (Hamiltonian Monte Carlo with restricted Gibbs sampling) to fit T-probit regression with heavy-tailed priors to make selection in the features with grouping structure. We will refer to this method as fully Bayesian T-probit. The main feature of fully Bayesian T-probit is that it can make feature selection within groups automatically without a pre-specification of the grouping structure and more efficiently discard noise features than LASSO (Least Absolute Shrinkage and Selection Operator). Therefore, the feature subsets selected by fully Bayesian T-probit are significantly more sparse than subsets selected by many other methods in the literature. Such succinct feature subsets are much easier to interpret or understand based on existing biological knowledge and further experimental investigations. In this thesis, we use simulated and real datasets to demonstrate that the predictive performances of the more sparse feature subsets selected by fully Bayesian T-probit are comparable with the much larger feature subsets selected by plain LASSO, Group LASSO, Supervised Group LASSO, random forest, penalized logistic regression and t-test. In addition, we demonstrate that the succinct feature subsets selected by fully Bayesian T-probit have significantly better predictive power than the feature subsets of the same size taken from the top features selected by the aforementioned methods.

Denna uppsats beräknar ett skuggpris för rädslan att gå ensam i mörkret i individens hemmakvarter samt undersöker skillnader mellan män och kvinnors skuggpriser för att vara mindre rädd. Tidigare studier som har använts sig av skuggpris för att beräkna rädsla har studerats och analyserats för att på bästa sätt kunna genomföra denna uppsats. Datamaterialet som använts i uppsatsen kommer från European Social Survey och består av 1791 observationer. Respondenterna fick i undersökningen svara på en skala mellan 1 till 4, där 1 var ’’mycket säker’’ och 4 var ’’mycket osäker’’ angående individens rädsla vid mörker. Skuggpriset motsvarar den summa av hushållets nettoinkomst som individen är villig att offra för att känna sig en enhet mindre rädd. Skuggpriset för att gå ensam i mörkret jämförs med män och kvinnors separata skuggpriser. Skuggpriset för att gå ensam i mörkret beräknades till 15 744 kronor. I uppsatsen skattades männens skuggpris till 12 881 kronor medan kvinnornas skuggpris blev 27 956 kronor. Därmed kan kvinnorna i Sverige tänka sig offra mer av sin månadsinkomst för att vara mindre rädd än männen i Sverige. Skillnaderna mellan män och kvinnors skuggpriser kan bero på att könen har separata psykologiska kostnader samt olika sociala roller i samhället. En kvinna tenderar att känna mer rädsla i vardagen än vad en man gör. Skillnaden mellan män och kvinnors skuggpriser skulle eventuellt kunna förklaras med att samhället kanske inte anser att en man ska känna rädsla och detta skulle kunna leda till att männen är villiga att betala mindre för att minska sin rädsla.

<p>Uppsatsen skapar en statistisk modell för beräkning av sannolikheter för utfall vid Allsvenska fotbollsmatcher. Modellens skattade sannolikheter för hemmavinst, oavgjort och bortavinst jämförs med utvalt spelbolag och visar på likvärdiga estimat. Utifrån de skattade sannolikheterna skapas odds som genererar lägre förluster än spelbolaget vid test utifrån ett spelbolags synvinkel. Sannolikheterna används även från en spelares perspektiv till att skapa ett spelsystem som genererar en positiv avkastning.</p>

Uppsatsen skapar en statistisk modell för beräkning av sannolikheter för utfall vid Allsvenska fotbollsmatcher. Modellens skattade sannolikheter för hemmavinst, oavgjort och bortavinst jämförs med utvalt spelbolag och visar på likvärdiga estimat. Utifrån de skattade sannolikheterna skapas odds som genererar lägre förluster än spelbolaget vid test utifrån ett spelbolags synvinkel. Sannolikheterna används även från en spelares perspektiv till att skapa ett spelsystem som genererar en positiv avkastning.

The membership of the Group of 20 was selected without any official criteria. This paper investigates whether group membership can be explained through the consideration of several different factors that coincide with the mission of the organization. I found strong evidence that membership in the Group of 20 was based on some combination of land mass and economic output. The results demonstrate that these factors are highly predictive of group membership.

This dissertation comprises two essays that attempt to determine, empirically, the fiscal response of governments’ to international assistance. The first essay examines whether an increasingly popular recommendation in international aid policy to switch from tied foreign assistance to untied foreign assistance affects investment in critical development expenditure sectors by developing countries. In the past, most international aid has been in the form of tied assistance as donors believed that tying aid will improve its effectiveness. It has been argued, that if tied aid is well designed and effectively managed then its overall effectiveness can be improved. On the contrary, it is also believed that tied aid acts as an impediment to donor cooperation and the building of partnership with developing countries. In addition, it is also argued that it removes the ‘feeling’ of ownership and responsibility of projects from partner countries in aid supported development. Two other more popular arguments used to challenge the effectiveness of foreign aid is that it is compromised when tied to the goods and services of the donor countries because almost 30 percent of its value is eliminated and also because it does not allow recipient countries to act on their priorities for public spending. These problems bring into question whether tied aid is truly the most effective way to help poor countries. A recommendation by the international community is that a switch to untied aid would be necessary. With untied aid, the recipient country is not obligated to buy the goods of the donor country neither is it compelled to pursue the public expenditure priorities of donors. Instead with untied aid they will have greater flexibility over spending decisions and can more easily pursue the priorities of their countries as they see fit. Hence, one could expect that a one dollar increase in untied aid will increase spending in the critical priority sectors by more than a one dollar increase in tied assistance. The question therefore is whether national domestic priorities coincide or not with what the international community has traditionally deemed should be priority. Empirically, we test this prediction using country-by-country data for 57 countries for the period 1973 to 2006. The results suggest that on average untied aid has a greater impact on pro-poor spending than do tied aid. In addition, the results also suggest that fungibility is still an issue even after accounting for the effects of untied aid. However, one could argue that fungibility may not be as bad as it appears since the switch to untied aid improves spending in the sectors that are essential for growth and development. The second essay explores the hypothesis that the expectations of debt forgiveness can discourage developing countries from attaining fiscal independence through an improvement of their tax effort. On the one hand, the international financial community typically advises poor countries to improve revenue mobilization but, on the other hand, the same international community routinely continues to bail-out poor countries that fail to meet their loan repayment obligations. The act of bailing-out these countries creates an expectation on the part of developing country governments that they will receive debt forgiveness time and again in the future. Therefore, the expectation of future bail outs creates a moral hazard that leads to endemic lower tax efforts. The key prediction of our simple theoretical model is that in the presence of debt forgiveness, tax ratios will decline and this decline will be stronger the higher the frequency and intensity of the bailouts. Empirically, we test this prediction using country-level data for 66 countries for the period 1989 to 2006. The results strongly suggest that debt forgiveness plays a significant role in the low tax effort observed in developing countries. Our empirical model allows for the endogeneity of tax effort and debt forgiveness. Interestingly we find that more debt forgiveness is actually provided to countries with lower tax effort. The results are robust to various specifications.

The Chinese insurance industry has been growing substantially, and this provides a motivation to examine the insurance market in China. This study used survey data to identify key determinants related to Chinese consumers’ ownership of life insurance, by using a probit model. The results revealed that several groups of variables influence Chinese consumers’ life insurance purchases, including knowledge and trust, consumer profile and investment preferences, importance of product attributes, and socio-demographics. Also, this study applied factor analysis to investigate factors that are important for Chinese consumers regarding life insurance. Factor analysis results indicated that four factors are identified including importance of product attributes, consumers’ financial strength, consumers’ attitude and trust toward the life insurance industry, and consumer attributes. Lastly, to better understand Chinese consumers regarding life insurance, consumers were segmented into three main groups through applying cluster analysis. Each cluster shows distinct differences in purchasing criteria and socio-demographic characteristics.