This thesis aims to examine the economic effect in terms of real GDP at constant 2017 national prices (in mil. 2017 US$) of being a member in the European Union in the case of Sweden. The method used to answer the research question is mainly by the Synthetic Control Method. Difference-in-difference estimations were also conducted as a complement to compare the results between the two methods. Previous literature in the subject has for the most part received results indicating that a membership in the European Union contributes to an increase in certain economic growth aspects for the member countries. The economic theories about common markets points at that a common market can be both advantageous but also disadvantageous depending on the setting. Theories more closely related to the European Union present rather ambiguous answers to certain economic effects of a membership but in some cases, it seems to be possible to assume some positive economic effect from being a European Union-member. The results in this paper seem to be robust in the way that the effect on the real GDP at constant 2017 national prices (in mil. 2017 US$) in Sweden from its membership in the European Union is negative in all estimations that were made. However, none of the results were statistically significant so there is not enough evidence to say that the Swedish real GDP at constant 2017 national prices (in mil. 2017 US$) would have been different to the factual outcome if Sweden did not join the European Union in 1995.

This dissertation is organized into three independent chapters. In Chapter 1, I consider the selection of weights for averaging a set of threshold models. Existing model averaging literature primarily focuses on averaging linear models, I consider threshold regression models. The theory I developed in that chapter demonstrates that the proposed jackknife model averaging estimator achieves asymptotic optimality when the set of candidate models are all misspecified threshold models. The simulations study demonstrates that the jackknife model averaging estimator achieves the lowest mean squared error when contrasted against other model selection and model averaging methods. In Chapter 2, I propose a model averaging framework for the synthetic control method of Abadie and Gardeazabal (2003) and Abadie et al. (2010). The proposed estimator serves a twofold purpose. First, it reduces the bias in estimating the weights each member of the donor pool receives. Secondly, it accounts for model uncertainty for the program evaluation estimation. I study two variations of the model, one where model weights are derived by solving a cross-validation quadratic program and another where each candidate model receives equal weights. Next, I show how to apply the placebo study and the conformal inference procedure for both versions of my estimator. With a simulation study, I reveal that the superior performance of the proposed procedure. In Chapter 3, which is co-authored with my advisor Professor Youngki Shin, we provide an exact computation algorithm for the maximum rank correlation estimator using the mixed integer programming (MIP) approach. We construct a new constrained optimization problem by transforming all indicator functions into binary parameters to be estimated and show that the transformation is equivalent to the original problem. Using a modern MIP solver, we apply the proposed method to an empirical example and Monte Carlo simulations. The results show that the proposed algorithm performs better than the existing alternatives. / Dissertation / Doctor of Philosophy (PhD)

Estimating the Euro effect with Synthetic Control Method for Eastern Europe Abstract This thesis estimates the effect of Euro adoption on newest Eurozone members using synthetic control method. The effect is estimated on income per capita and GDP growth. Estimates indicate overall indecisive effect for Slovakia and Malta, neutral effect for Estonia and negative effect for Slovenia and Cyprus. The cost of Euro for Cyprus is estimated to be as high as 1/3 of GDP per capita. In some cases the direction of the effect changed before and after the financial crisis. The quality of inference suffers from low number of observations. Methodological assumptions are discussed, concluding that quality of Eastern European time series likely causes substantial bias in the results.

THE ABSTRACT In the new global economy majority of the developed countries use imposition of sanctions in the case of some offences. I have applied the synthetic control methods on this particular case of political pressure to quantify the real costs of anti-Russian sanctions to the economy. Based on the results of this study it was identified, that real GDP growth rate of Russia was continuously reducing by about -1,19% per quarter on average. Reaching the highest value of sanction's effect of -2,8% in real GDP growth rate at the end of 2015. It was also revealed that sanctions had positive effect on current account balance of Russia, that indicator was increasing during the whole studied after sanction's period by about 3,15% per quarter on average. This thesis is presented as a research of interconnection the imposition of 2014 sanctions against Russia and indicators of economic development (GDP, inflation) using synthetic control methods. It was revealed that former economic and social mechanisms aren't capable to provide further development of economy of Russia, its self-sufficiency and safety. JEL Classification F12, F21, F23, H25, H71, H87 Keywords sanctions, synthetic control method, Russia, GDP growth rate, current account balance Author's e-mail 51375259@fsv.cuni.cz Supervisor's e-mail...

Cette thèse évalue l'impact économique des politiques de qualité liée à l'origine en mobilisant des méthodes d'évaluation que nous adaptons à la spécificité des démarches de certification, notamment à leur dimension spatiale. Dans un premier temps, nous analysons l'effet de l'adoption d’une Appellation d’Origine Protégée (AOP) sur le prix ainsi que sur les coûts de production des producteurs de lait AOP en Franche-Comté, en exploitant la discontinuité géographique à la frontière (Geographic Regression Discontinuity). Dans un second temps, nous utilisons les méthodes d'appariement statistique pour analyser les hétérogénéités régionales en termes de prix et de coûts de production entre les régions Auvergne, Rhône-Alpes et Franche-Comté avant de nous intéresser aux effets à une échelle nationale. Enfin, nous évaluons l'effet de l'adoption d'une Indication Géographique Protégée (IGP) sur les prix payés aux producteurs de Café de Colombie à l'aide des méthodes comparatives (synthetic control methods). En moyenne, les résultats montrent un effet positif de l'adoption de l'indication géographique sur la performance économique des exploitations. Cependant, cet effet est hétérogène entre les filières et les régions d'origine des produits. / This thesis evaluates the economic impact of quality-related-to-origin policies using evaluation methods adapted to the specificity of such certifications, in particular, the spatial dimension. First, we analyze the effect of a Protected Designation of Origin (PDO) adoption on price and production costs of PDO milk producers in Franche-Comté using a Geographic Regression Discontinuity (GRD) design. Secondly, we use propensity matching methods to analyze regional heterogeneity in Auvergne, Rhône-Alpes, and Franche-Comté before focusing on the effect of the PDO adoption at a national level. Finally, we evaluate the effect of the adoption of a Protected Geographical Indication (PGI) on prices paid to Colombian coffee producers using synthetic control methods. Overall, we find that PDO and/or PGI adoption is associated with positive effects on farmers' economic performances. However, these effects are not homogeneously distributed.

Submitted by Vitor Augusto Possebom (vitorapossebom@gmail.com) on 2016-03-08T00:06:59Z No. of bitstreams: 1 possebom_2016_masters-thesis.pdf: 905848 bytes, checksum: 1d5af42563617b7a8058b09baab1e040 (MD5) / Rejected by Letícia Monteiro de Souza (leticia.dsouza@fgv.br), reason: Prezado, Vítor, O seu trabalho foge totalmente das normas ABNT ou APA. Por gentileza, verificar trabalhos dos seus colegas postados na Biblioteca Digital para conhecimento. Qualquer dúvida, estou a disposição para falar ao telefone, onde fica mais fácil a comunicação. Atenciosamente, Letícia Monteiro 3799-3631 on 2016-03-08T11:53:56Z (GMT) / Submitted by Vitor Augusto Possebom (vitorapossebom@gmail.com) on 2016-03-08T20:40:10Z No. of bitstreams: 1 possebom_2016_masters-thesis.pdf: 926719 bytes, checksum: 6db5399b2d8e24a4b01f2bec748e4e95 (MD5) / Rejected by Letícia Monteiro de Souza (leticia.dsouza@fgv.br), reason: Prezado Vítor, Favor alterar o seu trabalho conforme as normas da ABNT. 1 - O Epigrafo deve constar na 5ª página, anteriormente a Dedicatória. 2 - Agradecimentos na 7ª página: Deve constar uma versão em português antes da versão em inglês. O título deve ser em caixa alta, centralizado e em negrito. 3 - Assim como em Agradecimentos, os títulos de: Resumo, Abstract e Sumário, devem ser em caixa alta, centralizado e em negrito. Estou a disposição para eventuais dúvidas. Atenciosamente, Letícia Monteiro 3799-3631 on 2016-03-09T12:15:54Z (GMT) / Submitted by Vitor Augusto Possebom (vitorapossebom@gmail.com) on 2016-03-10T00:21:36Z No. of bitstreams: 1 possebom_2016_masters-thesis.pdf: 933731 bytes, checksum: 69d467a1d6cb459ddd326d7fd593b4f9 (MD5) / Approved for entry into archive by Letícia Monteiro de Souza (leticia.dsouza@fgv.br) on 2016-03-10T11:59:00Z (GMT) No. of bitstreams: 1 possebom_2016_masters-thesis.pdf: 933731 bytes, checksum: 69d467a1d6cb459ddd326d7fd593b4f9 (MD5) / Made available in DSpace on 2016-03-10T12:48:54Z (GMT). No. of bitstreams: 1 possebom_2016_masters-thesis.pdf: 933731 bytes, checksum: 69d467a1d6cb459ddd326d7fd593b4f9 (MD5) Previous issue date: 2016-02-17 / This Master Thesis consists of one theoretical article and one empirical article on the field of Microeconometrics. The first chapter\footnote{We also thank useful suggestions by Marinho Bertanha, Gabriel Cepaluni, Brigham Frandsen, Dalia Ghanem, Ricardo Masini, Marcela Mello, Áureo de Paula, Cristine Pinto, Edson Severnini and seminar participants at São Paulo School of Economics, the California Econometrics Conference 2015 and the 37\textsuperscript{th} Brazilian Meeting of Econometrics.}, called \emph{Synthetic Control Estimator: A Generalized Inference Procedure and Confidence Sets}, contributes to the literature about inference techniques of the Synthetic Control Method. This methodology was proposed to answer questions involving counterfactuals when only one treated unit and a few control units are observed. Although this method was applied in many empirical works, the formal theory behind its inference procedure is still an open question. In order to fulfill this lacuna, we make clear the sufficient hypotheses that guarantee the adequacy of Fisher's Exact Hypothesis Testing Procedure for panel data, allowing us to test any \emph{sharp null hypothesis} and, consequently, to propose a new way to estimate Confidence Sets for the Synthetic Control Estimator by inverting a test statistic, the first confidence set when we have access only to finite sample, aggregate level data whose cross-sectional dimension may be larger than its time dimension. Moreover, we analyze the size and the power of the proposed test with a Monte Carlo experiment and find that test statistics that use the synthetic control method outperforms test statistics commonly used in the evaluation literature. We also extend our framework for the cases when we observe more than one outcome of interest (simultaneous hypothesis testing) or more than one treated unit (pooled intervention effect) and when heteroskedasticity is present. The second chapter, called \emph{Free Economic Area of Manaus: An Impact Evaluation using the Synthetic Control Method}, is an empirical article. We apply the synthetic control method for Brazilian city-level data during the 20\textsuperscript{th} Century in order to evaluate the economic impact of the Free Economic Area of Manaus (FEAM). We find that this enterprise zone had positive significant effects on Real GDP per capita and Services Total Production per capita, but it also had negative significant effects on Agriculture Total Production per capita. Our results suggest that this subsidy policy achieve its goal of promoting regional economic growth, even though it may have provoked mis-allocation of resources among economic sectors. / Esta dissertação de mestrado consiste em um artigo teórico e um artigo empírico no campo da Microeconometria. O primeiro capítulo contribui para a literatura sobre técnica de inferência do método de controle sintético. Essa metodologia foi proposta para responder a questões envolvendo contrafactuais quando apenas uma unidade tratada e poucas unidades controle são observadas. Apesar de esse método ter sido aplicado em muitos trabalhos empíricos, a teoria formal por trás de seu procedimento de inferência ainda é uma questão em aberto. Para preencher essa lacuna, nós deixamos claras hipóteses suficientes que garantem a validade do Procedimento Exato de Teste de Hipótese de Fisher para dados em painel, permitindo que nós testássemos qualquer hipótese nula do tipo \emph{sharp} e, consequentemente, que nós propuséssemos uma nova forma de estimar conjuntos de confiança para o Estimador de Controle Sintético por meio da inversão de uma estatística de teste, o primeiro conjunto de confiança quando temos acesso apenas a dados agregados cuja dimensão de \emph{cross-section} pode ser maior que a dimensão temporal. Ademais, nós analisamos o tamanho e o poder do teste proposto por meio de um experimento de Monte Carlo e encontramos que estatísticas de teste que usam o método de controle sintético apresentam uma performance superior àquela apresentada pelas estatísticas de teste comumente analisadas na literatura de avaliação de impacto. Nós também estendemos nosso procedimento para abarcar os casos em que observamos mais de uma variável de interesse (teste simultâneo de hipótese) ou mais de uma unidade tratada (efeito agregado da intervenção) e quando heterocedasticidade está presente. O segundo capítulo é um artigo empírico. Nós aplicamos o método de controle sintético a dados municipais brasileiros durante o século 20 com o intuito de avaliar o impacto econômico da Zona Franca de Manaus (ZFM). Nós encontramos que essa zona de empreendimento teve efeitos positivos significantes sobre o PIB Real per capita e sobre a Produção Total per capita do setor de Serviços, mas também teve um efeito negativo e significante sobre a Produção total per capita do setor Agrícola. Nossos resultados sugerem que essa política de subsídio alcançou seu objetivo de promover crescimento econômico regional, apesar de possivelmente ter provocado falhas de alocação de recursos entre setores econômicos.

Submitted by Sammara Soares (sammaracavalcanti@hotmail.com) on 2017-07-26T16:16:32Z No. of bitstreams: 1 PAPERS_TESE FINAL_v2.pdf: 1617261 bytes, checksum: 0140c13084348a38cf38ac5ce12badb9 (MD5) / Approved for entry into archive by Suzinei Teles Garcia Garcia (suzinei.garcia@fgv.br) on 2017-07-28T17:22:43Z (GMT) No. of bitstreams: 1 PAPERS_TESE FINAL_v2.pdf: 1617261 bytes, checksum: 0140c13084348a38cf38ac5ce12badb9 (MD5) / Made available in DSpace on 2017-07-28T17:26:36Z (GMT). No. of bitstreams: 1 PAPERS_TESE FINAL_v2.pdf: 1617261 bytes, checksum: 0140c13084348a38cf38ac5ce12badb9 (MD5) Previous issue date: 2017-07-29 / O ensino profissionalizante vem sendo apontado cada vez mais como uma solução alternativa no combate ao desemprego juvenil, especialmente àqueles que saem da escola direto para o mercado de trabalho. As evidências empíricas, no entanto, são amplamente divergentes quanto à eficácia de programas de ensino profissionalizante em todo o mundo. Além disso, os potenciais impactos não pecuniários, como fecundidade, foram raramente explorados. Neste contexto, a presente tese agrega importantes evidências ao explorar três diferentes modalidades de educação profissionalizante no Brasil. Explora-se uma variedade de potenciais resultados, desde indicadores de mercado de trabalho e de escolaridade, até fecundidade e formação de família. O primeiro artigo traz uma avaliação da política do Jovem Aprendiz no Brasil. O programa, formalmente criado no ano de 2000, tem como objetivo suavizar a transição da escola para o mercado de trabalho dos jovens. Com base no banco de dados longitudinal da RAIS, exploramos a adesão desigual do jovem aprendiz por entre os municípios e ao longo do tempo, como fonte de variação à exposição ao programa a nível municipal. Desta forma, aplicou-se o método de controle sintético para estimar seu efeito sobre resultados no mercado de trabalho, escolaridade e fecundidade. Em geral, os resultados sugerem que, apesar de o jovem aprendiz aumentar a empregabilidade aos 17 anos, não há efeito sobre os salários nem no médio nem no longo prazo. Por outro lado, temos evidências sugestivas de que existe um impacto positivo na empregabilidade a médio prazo. Finalmente, não encontramos impacto sobre fecundidade nem sobre variáveis relacionadas à escolaridade, como taxa de abandono ou reprovação. O segundo artigo, por sua vez, busca estimar o impacto de um programa de educação profissional sobre a formação familiar e indicadores de fecundidade dos jovens. Explora-se os dados experimentais de Santa Catarina, Brasil, onde cursos com excesso de demanda, financiados pelo PRONATEC, tiveram as suas vagas sorteadas. Desta forma, por meio de intention to treat effect e local average tratmente effect, estimamos o efeito de ter frequentado um curso técnico de nível secundário por meio da bolsa formação, sobre uma série de indicadores de fecundidade e formação familiar para homens e mulheres com até 35 anos de idade. Os resultados mostram que não há efeito para homens e nem para mulheres com relação à decisão de casar ou tornar-se chefe de família. No entanto, os homens tratados apresentaram uma maior probabilidade de ter filhos, especialmente aqueles na modalidade subsequente. Por outro lado, apesar das estimativas pontuais negativas, as subamostras femininas não evidenciaram qualquer diferença de média entre o grupo de tratamento e controle. Finalmente, o terceiro artigo teve o objetivo de avaliar o efeito de um credencial de educação profissionalizante de curto prazo sobre as chances de uma mulher, jovem, com ensino médio completo, ser convidada para uma entrevista de trabalho formal. Recorremos ao teste de correspondência, um experimento de campo que cria pares de currículos fictícios em que a única diferença é um credencial de educação profissionalizante no currículo tratado. Os currículos são enviados para vagas de emprego reais, de forma que explora-se o comportamento dos empregadores no ambiente real do mercado de trabalho. O resultado mostra que existe um efeito positivo modesto (3 pontos percentuais com um p-value a 11%) do curso profissionalizante de curto prazo sobre as chances de ser convidada para uma entrevista de emprego. / Vocational education has been increasingly perceived as an alternative solution to tackle youth unemployment, specially those not college bound. Empirical evidence, however, is widely divergent when it comes to the effectiveness of vocational programs all over the world. Additionally, potential non-pecuniary impacts, as fertility, were barely explored so far. In this context, the current thesis aggregate important and novel evidence regarding three different vocational education modalities in Brazil. It also encompass, a variety of dimensions, ranging from labor market and schooling indicators to fertility and family formation outcomes. The first paper provides an evaluation of the apprenticeship’s policy in Brazil. The program, formally created in 2000, intends to smooth the school-to-work transition of youngsters. Using a longitudinal database (RAIS), we explore the staggered pattern of the program’s placement across the municipalities and over time, as a source of variation in the apprenticeship exposure by municipality level. Therefore, we apply the synthetic control method to estimate the effect of the program’s exposure on a range of labor market, schooling and fertility outcomes. Overall, the results suggest that, despite the program increased the employability at 17 years old, it does not further reflect on superior wages neither in medium, nor in long terms. On the other hand, we have suggestive evidence that there exists a positive impact on medium term employability. Finally, we find no evidence regarding fertility neither schooling outcomes. The second paper, in turn, provides evidence of the impact of a vocational education program on family formation and fertility outcomes of youngsters. We take advantage of experimental data from Santa Catarina, Brazil, where oversubscribed courses, sponsored by PRONATEC, randomly assigned scholarships to individuals enrolled in classes that presented a number of registered candidates higher than its available vacancies. Therefore, applying intention-to-treat and local average treatment effects, we estimate the effect of having attended a secondary-level vocational course, on fertility and family formation outcomes for both men and women up to 35 years old. The results show that there is no effect for both men and women, regarding marriage or the probability to become head of family. However, the treated men present a higher probability of becoming parents than their counterfactual, specially from the subsequent modality. On the other hand, despite of negative point estimates, the women subsamples evidence no mean difference between treated and control groups. Finally, the third paper has the objective of assessing the effect of a short-term vocational education credential on the chances of a young high school graduated women being invited to a formal job interview. We resort to the correspondence test, a field experiment that creates pairs of fictitious resumes in which the only difference is an additional education credential in one of them. Therefore, they are sent to actual job vacancies, exploring the behaviour of the employers 2 in real labor market environment. The result evidences that, there exists a modest positive effect (3 percentage points with a p-value at 11% ) of the short-term vocational course credential on the chances of being invited to a job interview.

Cette thèse regroupe trois travaux d'économétrie liés par l'application du machine learning et de la statistique en grande dimension à l'évaluation de politiques publiques. La première partie propose une alternative paramétrique au contrôle synthétique (Abadie and Gardeazabal, 2003; Abadie et al., 2010) sous la forme d'un estimateur reposant sur une première étape de type Lasso, dont on montre qu'il est doublement robuste, asymptotiquement Normal et ``immunisé'' contre les erreurs de première étape. La seconde partie étudie une version pénalisée du contrôle synthétique en présence de données de nature micro-économique. La pénalisation permet d'obtenir une unité synthétique qui réalise un arbitrage entre reproduire fidèlement l'unité traitée durant la période pré-traitement et n'utiliser que des unités non-traitées suffisamment semblables à l'unité traitée. Nous étudions les propriétés de cet estimateur, proposons deux procédures de type ``validation croisée'' afin de choisir la pénalisation et discutons des procédures d'inférence par permutation. La dernière partie porte sur l'application du Generic Machine Learning (Chernozhukov et al., 2018) afin d'étudier l'hétérogénéité des effets d'une expérience aléatoire visant à comparer la fourniture publique et privée d'aide à la recherche d'emploi. D'un point de vue méthodologique, ce projet discute l'extension du Generic Machine Learning à des expériences avec compliance imparfaite. / This dissertation is comprised of three essays that apply machine learning and high-dimensional statistics to causal inference. The first essay proposes a parametric alternative to the synthetic control method (Abadie and Gardeazabal, 2003; Abadie et al., 2010) that relies on a Lasso-type first-step. We show that the resulting estimator is doubly robust, asymptotically Gaussian and ``immunized'' against first-step selection mistakes. The second essay studies a penalized version of the synthetic control method especially useful in the presence of micro-economic data. The penalization parameter trades off pairwise matching discrepancies with respect to the characteristics of each unit in the synthetic control against matching discrepancies with respect to the characteristics of the synthetic control unit as a whole. We study the properties of the resulting estimator, propose data-driven choices of the penalization parameter and discuss randomization-based inference procedures. The last essay applies the Generic Machine Learning framework (Chernozhukov et al., 2018) to study heterogeneity of the treatment in a randomized experiment designed to compare public and private provision of job counselling. From a methodological perspective, we discuss the extension of the Generic Machine Learning framework to experiments with imperfect compliance.

The doctoral thesis consists of three essays that address the application of synthetic control method to various economic policy intervention. In the first essay I evaluate the quantitative effects of the Czech National Bank's commitment to keep the Koruna from appreciating that were put in place in 2013. I find that the commitment helped decrease unemployment substantially. The effect on overall output is also strongly positive, almost 2 percentage points for growth in 2015. The effect of the commitment on inflation is positive but not statistically significant at standard levels. In the second essay I focus on the impact of joining the EU on the Czech agriculture. The results show that the Czech Republic would have had a higher food index if it had not entered the CAM and CAP. Moreover, I show that the CAP and CAM had different impacts on farms in the Czech Republic and Bulgaria, which have the most comparable agriculture according to the results of the synthetic control method. In the final essay I estimate how the UK financial markets would have evolved if the Remain camp had won the referendum. The results suggest that there would not have been any significant change in the development of the FTSE 100 Index in the medium to long term if there had not been a referendum. On the other hand, I...

The objective of this thesis is to analyze the impact of the electronic records of sales (ERS) implemented on December 1, 2016, in the Czech Republic. The ERS orders given entrepreneurs to use special cash boxes and to send data about single sales to the Financial Authority immediately after each transaction. Experience shows that the ERS reduced tax evasion and the shadow economy. This thesis fills a gap in literature as no study has yet examined the impact of ERS on macroeconomic variables - gross domestic product (GDP) per capita, unemployment rate and harmonized consumer price index (HCPI). To analyze it, I apply SCM by Abadie and Gardeazabal (2003), Abadie et al. (2010) and Abadie et al. (2015) that proposes the way how to evaluate the impact of political interventions using a quantitative approach. Using data from the Eurostat and OECD databases from Q1 2004-Q3 2019, the method allows me to construct a counterfactual outcome for GDP per capita, unemployment rate and HCPI in the absence of the implementation of the ERS. I also apply the extension of original SCM - generalize synthetic control method by Xu (2017) and augmented synthetic control method by Ben-Michael et al. (2018) as robustness checks. Moreover, the interference procedures include the modified confidence interval by Firpo and Possebom...