In this dissertation auction formats are developed and discussed that focus on three specific economic environments. Regarding the impossibility results from mechanism design, the main task for the implementation of auction designs is to balance allocative efficiency and incentive compatibility – the main characteristics a mechanism should provide. Therefore, the dissertation investigates the limits of conceivable relaxations of allocative efficiency and incentive compatibility for complex settings such as double auctions, interdependent-valuation environments and electricity market designs. The overall aim is to carefully weigh up the advantages and disadvantages for either relaxing allocative efficiency or respectively incentive compatibility.

O objetivo desta tese é discutir, sob a ótica da multimodalidade, os impactos na economia brasileira decorrentes da redução de custos de transportes. Para atingir este objetivo, um modelo inter-regional de equilíbrio geral computável denominado BR-Transport foi implementado, inspirado na teoria da escola australiana. Os dados calibrados para o modelo foram desagregados conforme quatro modais de transporte: rodoviário, ferroviário, hidroviário e dutoviário. Para a desagregação da atividade de transporte e estimação das elasticidades de substituição preço cruzado, usou-se a base de dados dos Planos Nacionais de Logística e Transportes do Brasil, a qual dispõe de informações sobre os fluxos de transporte entre 551 pontos distribuídos pelas 27 unidades federativas do país e o exterior, incluindo custos e volume de transporte por tipo de modal, discriminados em 29 grupos de produtos. Na estrutura teórica do modelo BR-Transport, introduziram-se mecanismos que se baseiam na racionalidade dos agentes, que minimizam custos, e contemplam as elasticidades de substituição preço cruzado entre os modais de transporte, as quais foram estimadas econometricamente. Foram admitidas duas abordagens para a estimação da função de demanda: a translog do tipo Constant Elasticity of Substitution, que corresponde à especificação teórica das funções de produção implementadas no modelo de equilíbrio geral computável; e a Multinomial Conditional Logit, tipicamente usada em modelos de escolha de modos de transporte. Ambas as estimações contaram com controles adicionais, como atributos do modal, efeitos fixos e variáveis instrumentais. As elasticidades obtidas a partir das diferentes formas funcionais permitiram realizar uma análise de sensibilidade dos resultados do modelo de equilíbrio geral, com base no uso dos diferentes vetores de elasticidades. Os resultados indicaram que choques generalizados de redução de custos no sistema de transporte brasileiro geram um aumento das exportações e importações inter-regionais e beneficiam principalmente os produtos regionais brutos dos estados das regiões Norte e Nordeste do país. Tais resultados estão diretamente associados à melhoria de acesso desses estados aos principais mercados consumidores e fornecedores, localizados nas regiões Sudeste e Sul, e revelam o potencial de redistribuição regional de renda das políticas de reduções de custos de transportes. Em uma análise dos impactos específicos dessas reduções de custos em cada modal, notou-se a preponderância do transporte rodoviário na economia nacional, em todos os estados e setores. Simulando uma redução de custos generalizada no modal ferroviário, o estado mais beneficiado seria o Pará, sobretudo porque ligações específicas por este modal entre os estados do Pará, de São Paulo e Minas Gerais estão entre as vinte mais importantes em termos de benefícios ao Produto Interno Bruto e às exportações nacionais. Melhorias no modal hidroviário, por sua vez, teriam impactos relevantes nos estados da região Norte do país. Já o modal dutoviário ainda mostra pequena expressão em relação a impactos regionais e setoriais / The objective of this thesis is to discuss, from the perspective of multimodality, the impacts on the Brazilian economy resulting from the retrenchment of transportation. To achieve this goal, an interregional computable general equilibrium model named BR-Transport was implemented, inspired by the theory of the Australian school. The calibrated data used in the model were disaggregated according to four modes of transport: roads, rails, waterways and pipelines. For the disaggregation of the transport activity and the estimation of the cross-price elasticities of substitution, the Brazilian National Plan for Logistics and Transport database was used; it contains information about traffic flows between 551 points over the 27 Brazilian states and the rest of the world, including costs and volume of transport for each mode of transportation, discriminated in 29 different product groups. In the theoretical structure of the BR-Transport model, mechanisms were introduced based on the rationality of agents, which minimize costs; they include the cross-price elasticities of substitution between modes of transport, which were econometrically estimated. Thus, two distinct functional forms were admitted: the first one was the Constant Elasticity of Substitution, a translog which corresponds to the theoretical specification of the production functions implemented in the computable general equilibrium model; and the second one was the Multinomial Conditional Logit, which is typically used in choice models of modes of transportation. Both estimations were made using additional controls, such as modal attributes, fixed effects and instrumental variables. The elasticities obtained from the different functional forms allowed the implementation of a sensitivity analysis of the general equilibrium model results, based on the use of different elasticities vectors. Results show that general shocks of retrenchment in the Brazilian transportation system increase inter-regional exports and imports and especially benefit gross regional products in Brazil\'s North and Northeast regions. These results are directly related to the improvement of the accessibility of these states to big markets and suppliers, located in the Southeast and South regions, and they demonstrate that politics of retrenchment of transport improve regional income redistribution. An analysis of the specific impacts of retrenchment in each mode indicates a preponderance of road transport in the national economy, in all states and sectors. A simulation of a general retrenchment of the railway mode shows that the state of Pará would have more benefits due to the fact that specific railway links between this state and the states of São Paulo and Minas Gerais are among the 20 most important ones in terms of benefits to the national Gross Domestic Product and the national exports. On the other hand, improvements in waterways would have significant impacts on the states of the North region. Finally, pipelines still show little expression with respect to regional and sectoral impacts.

O objetivo desta tese é discutir, sob a ótica da multimodalidade, os impactos na economia brasileira decorrentes da redução de custos de transportes. Para atingir este objetivo, um modelo inter-regional de equilíbrio geral computável denominado BR-Transport foi implementado, inspirado na teoria da escola australiana. Os dados calibrados para o modelo foram desagregados conforme quatro modais de transporte: rodoviário, ferroviário, hidroviário e dutoviário. Para a desagregação da atividade de transporte e estimação das elasticidades de substituição preço cruzado, usou-se a base de dados dos Planos Nacionais de Logística e Transportes do Brasil, a qual dispõe de informações sobre os fluxos de transporte entre 551 pontos distribuídos pelas 27 unidades federativas do país e o exterior, incluindo custos e volume de transporte por tipo de modal, discriminados em 29 grupos de produtos. Na estrutura teórica do modelo BR-Transport, introduziram-se mecanismos que se baseiam na racionalidade dos agentes, que minimizam custos, e contemplam as elasticidades de substituição preço cruzado entre os modais de transporte, as quais foram estimadas econometricamente. Foram admitidas duas abordagens para a estimação da função de demanda: a translog do tipo Constant Elasticity of Substitution, que corresponde à especificação teórica das funções de produção implementadas no modelo de equilíbrio geral computável; e a Multinomial Conditional Logit, tipicamente usada em modelos de escolha de modos de transporte. Ambas as estimações contaram com controles adicionais, como atributos do modal, efeitos fixos e variáveis instrumentais. As elasticidades obtidas a partir das diferentes formas funcionais permitiram realizar uma análise de sensibilidade dos resultados do modelo de equilíbrio geral, com base no uso dos diferentes vetores de elasticidades. Os resultados indicaram que choques generalizados de redução de custos no sistema de transporte brasileiro geram um aumento das exportações e importações inter-regionais e beneficiam principalmente os produtos regionais brutos dos estados das regiões Norte e Nordeste do país. Tais resultados estão diretamente associados à melhoria de acesso desses estados aos principais mercados consumidores e fornecedores, localizados nas regiões Sudeste e Sul, e revelam o potencial de redistribuição regional de renda das políticas de reduções de custos de transportes. Em uma análise dos impactos específicos dessas reduções de custos em cada modal, notou-se a preponderância do transporte rodoviário na economia nacional, em todos os estados e setores. Simulando uma redução de custos generalizada no modal ferroviário, o estado mais beneficiado seria o Pará, sobretudo porque ligações específicas por este modal entre os estados do Pará, de São Paulo e Minas Gerais estão entre as vinte mais importantes em termos de benefícios ao Produto Interno Bruto e às exportações nacionais. Melhorias no modal hidroviário, por sua vez, teriam impactos relevantes nos estados da região Norte do país. Já o modal dutoviário ainda mostra pequena expressão em relação a impactos regionais e setoriais / The objective of this thesis is to discuss, from the perspective of multimodality, the impacts on the Brazilian economy resulting from the retrenchment of transportation. To achieve this goal, an interregional computable general equilibrium model named BR-Transport was implemented, inspired by the theory of the Australian school. The calibrated data used in the model were disaggregated according to four modes of transport: roads, rails, waterways and pipelines. For the disaggregation of the transport activity and the estimation of the cross-price elasticities of substitution, the Brazilian National Plan for Logistics and Transport database was used; it contains information about traffic flows between 551 points over the 27 Brazilian states and the rest of the world, including costs and volume of transport for each mode of transportation, discriminated in 29 different product groups. In the theoretical structure of the BR-Transport model, mechanisms were introduced based on the rationality of agents, which minimize costs; they include the cross-price elasticities of substitution between modes of transport, which were econometrically estimated. Thus, two distinct functional forms were admitted: the first one was the Constant Elasticity of Substitution, a translog which corresponds to the theoretical specification of the production functions implemented in the computable general equilibrium model; and the second one was the Multinomial Conditional Logit, which is typically used in choice models of modes of transportation. Both estimations were made using additional controls, such as modal attributes, fixed effects and instrumental variables. The elasticities obtained from the different functional forms allowed the implementation of a sensitivity analysis of the general equilibrium model results, based on the use of different elasticities vectors. Results show that general shocks of retrenchment in the Brazilian transportation system increase inter-regional exports and imports and especially benefit gross regional products in Brazil\'s North and Northeast regions. These results are directly related to the improvement of the accessibility of these states to big markets and suppliers, located in the Southeast and South regions, and they demonstrate that politics of retrenchment of transport improve regional income redistribution. An analysis of the specific impacts of retrenchment in each mode indicates a preponderance of road transport in the national economy, in all states and sectors. A simulation of a general retrenchment of the railway mode shows that the state of Pará would have more benefits due to the fact that specific railway links between this state and the states of São Paulo and Minas Gerais are among the 20 most important ones in terms of benefits to the national Gross Domestic Product and the national exports. On the other hand, improvements in waterways would have significant impacts on the states of the North region. Finally, pipelines still show little expression with respect to regional and sectoral impacts.

Utilisant des données sur les expéditions de robots industriels, ce document constate que les secteurs volatils sont automatisés de façon disproportionnée dans les pays avec des règles strictes sur le licenciement. L'idée derrière le résultat empirique est que contrairement aux travailleurs humains, les entreprises peuvent disposer librement des robots, tout comme avec tout autre bien de capital. Puisque les droits des robots ne sont pas protégés par la loi, ils fournissent des services de main d’œuvre à volonté. L'incitation à l'automatisation apportée par la réglementation est plus forte dans les secteurs volatils, où l'incertitude sur les conditions d'affaires augmente les exigences de flexibilité des entreprises. Mes estimations prédisent que dans les secteurs incertains, le pays le plus réglementé (l'Italie) devrait être deux fois plus automatisé que le plus réglementé (États-Unis). Les données montrent que le nombre de robots par employé en Italie est 90% plus élevé qu'aux États-Unis. La stratégie d'identification consiste à exploiter le calendrier différent dans les réformes du travail entre les pays. L'identification provient de l'effet avant-après sur l'investissement sectoriel dans les robots dans les pays réformés (le "groupe de traitement"), vis-à-vis de l'effet avant-après dans les pays où l'EPL n'a pas changé (le "groupe témoin"). Ce document explique pourquoi les entreprises investissent dans l'automatisation. Contrairement à la sagesse conventionnelle, les robots n'augmentent pas la productivité parce qu'ils sont meilleurs ou plus rapides à faire les choses, mais plutôt parce qu'ils augmentent l'efficacité de la répartition. Le lien entre réglementation et automatisation implique que la politique du marché du travail peut être utilisée pour atténuer l'effet perturbateur de la technologie. / Using data on shipments of industrial robots, this paper finds that volatile sectors are disproportionally automated in countries with strict rules on employment dismissal. The idea behind the empirical result is that unlike for human workers, firms can freely dispose of robots, just as with any other capital good. Since robot’s rights are not protected by law, they deliver labor services at will. The incentive to automate induced by regulation is stronger in volatile sectors, where uncertainty about business conditions increase the flexibility requirements of firms. My estimates predict that in uncertain sectors, the most regulated country (Italy) should be twice as automated as the least regulated one (the United States). Data show that the number of robots per employee in Italy is 90% higher than in the United States. The identification strategy consists in exploiting different timing in labor re-forms across countries. Identification comes from the before-after effect on sectorial investment in robots in reformed countries (the \treatment group"), vis-a-vis the before-after effect in countries where EPL did not change (the \control group").This paper sheds light on why firms invest in automation. In contrast to conventional wisdom, robots do not increase productivity because they are better or faster at doing things, but rather because they increase allocative efficiency. The link between regulation and automation implies that labor market policy can be used to mitigate the disruptive effect of technology.

There have been a number of support initiatives rendered to the small-scale sugarcane farmers in Mpumalanga but the sugarcane production there has remained low. This raised some questions on the efficient use of available resources and the fruitfulness of the Producer Development Initiatives (PDIs). Justification for further assistance to the small-scale farmers requires empirical evidence of efficient resource use. This study employed the stochastic frontier production function to calculate the technical, allocative and cost efficiency. This study provides insight into three issues: the levels of technical, economic and allocative efficiencies of small-scale sugarcane farmers; the relationship between efficiency level and various farm/farmer specific factors; and implications of policy and strategies for improving small-scale sugarcane production. The technical, allocative and economic efficiencies of small-scale sugarcane farming were estimated in order to identify the potential increase in production without incurring additional costs. The study used data obtained from a field survey covering 231 small-scale sugarcane farmers in the Nkomazi region for the 2009/2010. sugarcane production season. According to the stochastic frontier production function using the Cobb-Douglas model, labour, herbicides and fertilizer showed significant positive effects on sugarcane production. The results also indicated that the small-scale sugarcane farmers suffer from considerable lack technical, allocative and cost efficiency. The mean technical, allocative and cost efficiency estimates are 68.5%, 61.5% and 41.8% respectively. A Tobit regression was used to analyse the impact of the farm/farmer characteristics on efficiency. The impact analysis revealed that age, level of education and gender are significant determinants of technical efficiency. On the other hand, level of education, off-farm income, land size and experience are significant determinants of allocative efficiency. In so far as cost efficiency is concerned, the significant determinants are level of education, land size and experience in sugarcane farming. The findings of the study justify the need for improved agricultural partnerships between the sugar mills and the sugarcane farmers. An example of such collaboration would be if millers were to not only give credit to the farmers, but also give technical guidance to small producers in return for the delivery of a specific quantity and quality of cane at a stipulated time. The collective efforts of these farmers and millers, once harmoniously co-ordinated, can enhance production efficiency and economic prosperity. Also, appropriate policy formulation and implementation is an effective instrument to improvement in farm efficiency and productivity which promotes overall growth of the economy. Therefore, there is need for all stakeholders (both private and public sector) to make combined efforts to remove the bottlenecks that have constrained effective policy implementation in, and its accrued benefits to, South African agriculture. / Dissertation MSc (Agric)--University of Pretoria, 2013. / gm2014 / Agricultural Economics, Extension and Rural Development / unrestricted

This paper surveys the theoretical and empirical research on the efficient assignment of policy tasks to different levels of government and applies the results on the delimitation of competences within the European Union. The main results are: (i) A precise derivation of an optimal degree of decentralisation is not possible because of mixed theoretical suggestions. The adequate degree of decentralisation has to be detected case-by-case. (ii) Systematic evidence on direct relationships between economic performance and fiscal decentralisation is ambiguous and scarce. (iii) Comparing the de facto delimitation of EU-competences with the normative recommendations, remarkable discrepancies arise in the fields of agriculture and defence. (iv) The establishment of a flexible assignment-scheme by the European Convention is an undeniable necessity in order to guarantee reversibility and to cope efficiently with changing general conditions. (author's abstract) / Series: EI Working Papers / Europainstitut

Seit 2003 hat die Effizienzmessung im dynamischen Kontext erheblich an Aufmerksamkeit gewonnen. Die dynamische Effizienzanalyse berücksichtigt sowohl die zeitliche Interdependenz der Produktionsentscheidungen als auch Anpassungskosten. Zudem wird zwischen variablen und quasi-fixen Produktionsfaktoren unterschieden. Allerdings haben strukturelle dynamische Effizienzmodelle bisher Unsicherheit vernachlässigt, was zu irreführenden Effizienzwerten führen kann. Unsicherheit beeinflusst die optimale Anpassung von Produktionsentscheidungen; dies ist besonders relevant für die optimale Anpassung der quasi-fixen Faktoren im Zeitablauf. Deshalb ist es das Ziel dieser Doktorarbeit, diese Lücke zu schließen und ein theoretisches Modell für die dynamische Effizienzmessung unter Unsicherheit basierend auf einer Kostenminimierung zu entwickeln. Um ein solches Modell herzuleiten, verwendet die Autorin zwei Komponenten: den statischen Schattenkostenansatz und ein stochastisches duales Investitionsmodel unter Unsicherheit. Während der Schattenkostenansatz die ökonomische Effizienz in eine technische und eine allokative Komponente zerlegt, erlaubt das stochastische intertemporale Dualitätsmodell, Unsicherheit und Anpassungskosten zu berücksichtigen. Die resultierenden empirischen stochastischen Nachfragegleichungen dienen als Grundlage für die ökonometrische Schätzung der technischen und allokativen Effizienz. Die theoretischen Erkenntnisse des hergeleiteten Modells wurden anschließend mit Hilfe einer Simulation überprüft, mit dem Ziel, einerseits die Höhe der Verzerrung der geschätzten Koeffizienten durch ausgelassene Variablen zu ermitteln, wenn Unsicherheit bei der optimalen Faktoranpassung vernachlässigt wird, und andererseits den Einfluss der Unsicherheit auf die Faktornachfragegleichungen zu analysieren. Die Simulationsergebnisse zeigen, dass eine Vernachlässigung der Unsicherheit zur Verzerrung der geschätzten Modellparameter führt. / Since 2003, measuring efficiency in dynamic contexts has received considerable attention. Dynamic efficiency analysis accounts for both the interdependency of production decisions over time, as well as adjustment costs, and also distinguishes between variable and quasi-fixed inputs in the production process. However, structural models of dynamic efficiency have thus far ignored uncertainty; this may lead to misleading measures of efficiency. Uncertainty affects the optimal allocation of input decisions and it is particularly true for the optimal adjustment of quasi-fixed factors over time. Hence, to fill this gap, this thesis aims to develop a theoretical model of dynamic efficiency under uncertainty based on the cost-minimization problem. To derive such a model, the author uses two components, namely the static shadow cost approach and a stochastic dual model of investments under uncertainty. The shadow cost approach allows one to disentangle economic inefficiency into technical and allocative inefficiency, while the stochastic intertemporal duality model enables one to consider uncertainty and adjustment costs. Formulating an empirical model requires one to specify the functional form of the respective value function. Here, the specified value function properties facilitate output and price uncertainty to influence optimal factor demand equations. The resulting empirical stochastic factor demand equations then serve as a starting point for the econometric estimation of technical and allocative inefficiency measures. Theoretical findings from the derived model were subsequently tested using a simulation, to determine how large the omitted variable bias is on the estimates of the coefficients if uncertainty is ignored in optimal factor allocations, and to analyze the influence of uncertainty on factor demand equations. The simulation results reveal that disregarding uncertainty in optimal factor allocations leads to biased estimates of model parameters.

A maioria dos trabalhos sobre o impacto macroeconômico da abertura da conta de capital não encontra nenhum efeito da liberalização sobre as variáveis reais. No entanto, uma leitura cuidadosa desta literatura revela que a maioria destes estudos não trata realmente da teoria que se propõe a testar. Aqueles que defendem um impacto positivo da liberalização financeira sobre o crescimento econômico aceitam as previsões do modelo de crescimento neoclássico de redução permanente no custo do capital e aumento temporário no investimento nos mercados emergentes, quando estes liberalizam suas contas de capital. A maior parte dos artigos que não encontram efeitos da liberalização sobre as variáveis reais não testa estas previsões. Uma ramificação pequena, mas crescente, desta literatura sobre a relação entre liberalização da conta de capital e crescimento econômico, que leva em conta a natureza temporal das previsões do modelo neoclássico (os artigos que adotam o chamado enfoque do experimento de política), encontra evidências de que a abertura da conta de capital em um país emergente gera efeitos significativos sobre o investimento e crescimento econômico. A desagregação dos dados, ou seja, a aplicação do enfoque do experimento de política a dados de firmas, fornece uma ligação mais forte com a teoria e graus de liberdade suficientes para se adotar uma nova estratégia de identificação que permite testar a eficiência na alocação dos recursos dentro do país em desenvolvimento que abriu sua conta de capital. O objetivo desta dissertação é tratar das teorias e evidências do impacto da liberalização da conta de capital nos países emergentes sobre o crescimento de suas economias, analisando-se as metodologias empíricas existentes aplicadas no teste desta relação, enfatizando as teorias que dão suporte a seus testes empíricos, bem como suas principais descobertas. Os efeitos da abertura da conta de capital no Brasil, na década de 90, serão analisados, utilizando-se uma estratégia de identificação recente a partir de dados de firmas, para se checar a validade empírica das previsões do modelo de crescimento neoclássico nesta situação. / Most writings on the macroeconomic impact of capital account openness find no effects of liberalization on real variables. However, a critical reading of this literature reveals that most of these studies do not really address the theory they set out to test. Those who defend a positive impact of financial liberalization on economic growth accept the neoclassical growth model’s predictions of permanent reduction on the cost of capital and temporary increase in investment in emerging markets when they liberalize their capital account. The lion’s share of papers that find no effect of liberalization on real variables do not test these predictions. A small but growing branch of this literature on the relationship between capital account liberalization and economic growth that takes the time series nature of the neoclassical model’s predictions seriously (papers that adopt the policy-experiment approach) find that opening the capital account within an emerging country generates significant effects on investment and economic growth. Disaggregating the data, that is to say, applying the policy-experiment approach to firm-level data, provides a tighter link to the theory than aggregate data and enough degrees of freedom to adopt a new identification strategy that allows of the test on the efficiency of capital allocation within the developing country that opened its capital account. The objective of this dissertation is to address the theories and evidences of the impact of capital account liberalization in emerging countries on their economic growth, analyzing the existing empirical methodologies applied to test this relationship, emphasizing the theories that support their empirical tests as well as their main findings. The effects of capital account opening in Brazil, in the 90’s, will be analyzed, using the recent firm-level data identification strategy, to check the empirical validity of the neoclassical growth model’s predictions in this situation.

Die große Bedeutung grenzüberschreitender Kapitalbewegungen und die zunehmende Mobilität des Faktors Arbeit verstärken immer mehr die Relevanz der steuerlichen Behandlung grenzüberschreitend erzielter Einkünfte. Die vorliegende Arbeit beschäftigt sich mit den Auswirkungen des derzeitigen Systems der internationalen Besteuerung auf grenzüberschreitende Faktorallokationen. Im Idealfall sollte die Freizügigkeit der Faktorströme zu effizienten Produktionsstrukturen führen, indem Unternehmen, Kapitalanleger und Arbeiter ihren Investitions-, Kapitalanlage- bzw. Arbeitsort dort wählen, wo ihre Produktivität am höchsten ist; die Besteuerung sollte dabei ohne Einfluß sein. Zur Untersuchung, ob dieser Idealfall auch in der Praxis gegeben ist, werden zunächst die internationalen Besteuerungsprinzipien vorgestellt. Die darauffolgenden Kapitel befassen sich sodann mit den Auswirkungen der internationalen Besteuerung auf die räumliche Allokation der Faktoren Kapital und Arbeit im Rahmen einer theoretischen Analyse und der Frage, welchen Anforderungen ein internationales Steuersystem genügen muß, um eine global optimale Allokation zu gewährleisten. Vor dem Hintergrund eines solchen idealtypischen internationalen Steuersystems erfolgt eine Analyse der bestehenden Besteuerungsvorschriften für grenzüberschreitend erzielte Einkünfte zwischen Deutschland und Frankreich bzw. Deutschland und den USA verbunden mit einer Einführung in das nationale und internationale Steuerrecht der betrachteten Länder. Aufbauend auf diesen Erkenntnissen wird schließlich ein Reformsteuersystem mit Beispielen für die jeweiligen Länder und Einkunftsarten erarbeitet, das eine allokationsneutrale internationale Einkommenserzielung gestattet. Dieser Reformvorschlag zeichnet sich durch seine pragmatische Ausgestaltung aus, die zur Sicherung einer Realisierungschance so wenig wie möglich in die Besteuerungsautonomien der einzelnen Länder eingreift und administrativ relativ einfach umzusetzen ist. / The importance of international capital flows and the increasing mobility of human capital give more and more weight to the question of taxation of the revenues thereof. This study analyses the effects of the actual system of international taxation in this context. Ideally, free factor flows should lead to efficient production structures with companies, capital owners and workers choosing their location of input where the productivity is highest; taxation should have no influence on this process. The analysis starts with a description of the existing international taxation principles. It then continues with the effects of the international taxation on the geographical allocation of the production factors capital and work and the conditions a system of international taxation has to fulfil for optimal global allocation. The existing international tax law for cross border revenues between Germany and France, and Germany and the United States, respectively, is compared with such an ideal system joined with an introduction to the national and international tax law of the countries considered. Given this information, a reform proposal for international taxation guaranteeing efficient factor allocations is presented with examples for the three countries and the different revenues. This reform proposal is characterised by its pragmatic concept: It disturbs the tax autonomies of the countries the least it has to and it is relatively easy to introduce.

A maioria dos trabalhos sobre o impacto macroeconômico da abertura da conta de capital não encontra nenhum efeito da liberalização sobre as variáveis reais. No entanto, uma leitura cuidadosa desta literatura revela que a maioria destes estudos não trata realmente da teoria que se propõe a testar. Aqueles que defendem um impacto positivo da liberalização financeira sobre o crescimento econômico aceitam as previsões do modelo de crescimento neoclássico de redução permanente no custo do capital e aumento temporário no investimento nos mercados emergentes, quando estes liberalizam suas contas de capital. A maior parte dos artigos que não encontram efeitos da liberalização sobre as variáveis reais não testa estas previsões. Uma ramificação pequena, mas crescente, desta literatura sobre a relação entre liberalização da conta de capital e crescimento econômico, que leva em conta a natureza temporal das previsões do modelo neoclássico (os artigos que adotam o chamado enfoque do experimento de política), encontra evidências de que a abertura da conta de capital em um país emergente gera efeitos significativos sobre o investimento e crescimento econômico. A desagregação dos dados, ou seja, a aplicação do enfoque do experimento de política a dados de firmas, fornece uma ligação mais forte com a teoria e graus de liberdade suficientes para se adotar uma nova estratégia de identificação que permite testar a eficiência na alocação dos recursos dentro do país em desenvolvimento que abriu sua conta de capital. O objetivo desta dissertação é tratar das teorias e evidências do impacto da liberalização da conta de capital nos países emergentes sobre o crescimento de suas economias, analisando-se as metodologias empíricas existentes aplicadas no teste desta relação, enfatizando as teorias que dão suporte a seus testes empíricos, bem como suas principais descobertas. Os efeitos da abertura da conta de capital no Brasil, na década de 90, serão analisados, utilizando-se uma estratégia de identificação recente a partir de dados de firmas, para se checar a validade empírica das previsões do modelo de crescimento neoclássico nesta situação. / Most writings on the macroeconomic impact of capital account openness find no effects of liberalization on real variables. However, a critical reading of this literature reveals that most of these studies do not really address the theory they set out to test. Those who defend a positive impact of financial liberalization on economic growth accept the neoclassical growth model’s predictions of permanent reduction on the cost of capital and temporary increase in investment in emerging markets when they liberalize their capital account. The lion’s share of papers that find no effect of liberalization on real variables do not test these predictions. A small but growing branch of this literature on the relationship between capital account liberalization and economic growth that takes the time series nature of the neoclassical model’s predictions seriously (papers that adopt the policy-experiment approach) find that opening the capital account within an emerging country generates significant effects on investment and economic growth. Disaggregating the data, that is to say, applying the policy-experiment approach to firm-level data, provides a tighter link to the theory than aggregate data and enough degrees of freedom to adopt a new identification strategy that allows of the test on the efficiency of capital allocation within the developing country that opened its capital account. The objective of this dissertation is to address the theories and evidences of the impact of capital account liberalization in emerging countries on their economic growth, analyzing the existing empirical methodologies applied to test this relationship, emphasizing the theories that support their empirical tests as well as their main findings. The effects of capital account opening in Brazil, in the 90’s, will be analyzed, using the recent firm-level data identification strategy, to check the empirical validity of the neoclassical growth model’s predictions in this situation.