Orientador : Profª. Drª. Betina Treiger Grupenmacher / Dissertação (mestrado) - Universidade Federal do Paraná, Setor de Ciências Jurídicas, Programa de Pós-Graduação em Direito. Defesa: Curitiba, 29/03/2016 / Inclui referências : f. 145-150 / Área de concentração: Direito do estado / Linha de pesquisa: Perspectivas da dogmática crítica / Resumo: Esta dissertação tem por objetivo examinar se o "custo dos direitos" representa uma limitação implícita ao exercício da competência tributária. Partindo das premissas de que o Direito não se resume a um "modelo de regras", mas se constitui em uma "prática argumentativa"; de que os "textos normativos" não se confundem com as "normas jurídicas", já que estas são construídas, através da interpretação, a partir daqueles; e de que a interpretação do direito pressupõe o uso de argumentos interpretativos, tais como os argumentos linguísticos, sistêmicos e teleológicos/avaliativos, investigamos a compatibilidade com a atual Ordem Constitucional da instituição e/ou majoração de tributos cuja expectativa de receita supere o "custo" estimado dos direitos, custos estes avaliados na perspectiva do "Estado eficiente". Para tanto, confrontamos a "tributação excessiva" com o princípio da justiça tributária; com as justificações racionais autorizativas e legitimadoras do poder de tributar; e com o Princípio da Eficiência do Estado, para demonstrar que a "competência tributária" não confere ao Estado o poder de instituir ou majorar tributos cuja expectativa de arrecadação seja superior ao efetivo e eficiente "custo dos direitos". Por fim, demonstramos que o exercício da competência tributária, para que seja compatível com a observância do princípio da eficiência, com as justificações racionais autorizativas do poder de tributar e com o princípio da justiça tributária, deverá, necessariamente, observar o "postulado da proporcionalidade", no que se refere aos exames de adequação, necessidade e "proporcionalidade em sentido estrito" do meio escolhido para a realização do fim. Palavras-chave: Limite. Competência tributária. Justiça. Justificação Racional. Princípio da Eficiência. Postulado da Proporcionalidade. / Abstract: This work aims to examine whether the "cost of rights" represents an implied limitation on the exercise of the taxing power. Starting from the premise that the law is not limited to a "model of rules", but constitutes an "argumentative practice"; that the "legal texts" should not be confused with the "legal norms", since they are built, through interpretation, from those; and that interpretation of the law requires the use of interpretative arguments, such as the linguistic arguments, systemic and teleological / evaluative, we investigated the compatibility with the current constitutional order of the institution and/or increase of taxes which expected revenue exceeds the estimated cost of rights, these costs evaluated from the perspective of "Efficient State". To this end, we confront the "excessive taxation" with the principle of tax justice; with the authorizing and legitimating rational justifications of the power to tax; and with the State's Principle of Efficiency, to demonstrate that "taxing authority" does not give the state the power to impose or jack taxes which the expected revenue is higher than the effective and efficient "cost of rights". Lastly, we demonstrate that the exercise of the taxing power, to be compatible with the principle of efficiency, with the authorizing rational justifications of the power to tax and the principle of tax justice, must necessarily observe the "principle of proportionality" regarding the suitability tests, necessity and "proportionality in the strict sense" of the means chosen to perform the purpose. Keywords: Limit. taxing power. Justice. Rational justification. Efficiency Principle. Principle of Proportionality.

The first chapter studies the impact of variance risk in the Treasury market on both term premia and the shape of the yield curve. Under minimal assumptions shared by standard structural and reduced-form asset pricing models, I show that an observable proxy of variance risk in the Treasury market can be constructed via a portfolio of Treasury options. The observable variance risk has the ability to explain the time variation in term premia, but is largely unrelated to the shape of the yield curve. Using the observable variance risk, I also propose a new representation of no-arbitrage term structure models. All the pricing factors in the model are observable, tradable, and hence economically interpretable. The representation can also accommodate both unspanned macro risks and unspanned stochastic volatility in the term structure literature. The second chapter shows that it is beneficial to incorporate a particular zero-cost trading strategy into approaches that extract a stochastic discount factor from asset prices in a model-free manner (e.g. the Hansen-Jagannathan minimum variance stochastic discount factor). The strategy mimics the Radon-Nikodym derivative between two pricing measures with alternative investment horizons, and is hence characterized by the term structure of the SDF (or the dynamics of the SDF). Incorporating the strategy into the Euler equation significantly enhances the ability of the extracted stochastic discount factor to explain crosssectional variation of expected asset returns. Furthermore, the strategy remarkably tightens various lower bounds for the stochastic discount factor, hence setting a more stringent hurdle for equilibrium asset pricing models. The third chapter studies variance risk premiums in the Treasury market. We first develop a theory to price variance swaps and show that the realized variance can be perfectly replicated by a static position in Treasury futures options and a dynamic position in the underlying. Pricing and hedging is robust even if the underlying jumps. Using a large options panel data set on Treasury futures with different tenors, we report the following findings: First, the term-structure of implied variances is downward sloping across maturities and increases in tenors. Moreover, the slope of the term structure is strongly linked to economic activity. Second, returns to the Treasury variance swap are negative and economically large. Shorting a variance swap produces an annualized Sharpe ratio of almost two and the associated returns cannot be explained by standard risk factors. Moreover, the returns remain highly statistically significant even when accounting for transaction costs and margin requirements.

Unlike tax accountants and advisors within direct tax, Customs law in many countries requires importers to employ licensed Customs agents. This study extends the tax literature by examining the role of Customs agents in import tax compliance. In Malaysia, as an example of a country where Customs are responsible for about one-third (MYR30 billion on average between 2005 to 2014) of total government revenue collections, the function of Customs agents is to: assist importers in meeting their import tax liabilities; prepare and submit all necessary import documentation to Customs; as well as collect and pay all revenue to the Customs administration. Customs agents are bound by the Customs Act 1967 and are required to pass a public exam before becoming formally qualified and licensed Customs agents. Exploratory interviews with senior Customs officers at The Royal Malaysian Customs Department (RMCD) suggest that a significant amount of tax revenue is lost because Customs agents do not pay the full amount of import duty and tax due. Most interviewed officers felt that tougher penalties and sanctions are required to improve compliance and root-out fraud; though some indicated that other measures may need to be developed in order to improve compliance practice. Drawing on the tax compliance literature within the direct tax domain, the Theory of Planned Behaviour (TPB) has been identified as one of the most robust social cognitive theories to explain compliance decision making. A key output from this research is a compliance behaviour model (based on the theory of planned behaviour) that depicts various economic and non-economic variables to predict compliance behaviour. Building on the model, a large scale survey of Customs agents across Malaysia was conducted. Overall, the response rate was 42% (n=650), representing 12.8% of the total Customs agents population located at Malaysia’s primary ports of entry. The results indicate that psychological, sociological, structural / institutional factors, which consist of attitude, ethical beliefs, social norms, law, enforcement, complexity of procedure and quality of tax assessment service, are significant in explaining Customs agents’ behavioural intention to comply with import tax law. However, they also suggest inconsistencies in the relationship between behavioural intention and behaviour, and the need to incorporate other factors and moderating variables. In particular, the findings identified the influence of two referent groups (subjective norms): (i) the importers who influenced Customs agents’ import tax compliance directly through instructions, as well as indirectly by sharing their ethical beliefs; and (ii) other Customs agents (their peers) who influence Customs agents’ ethical beliefs. Overall this study highlights the importance of incorporating behavioural elements and facilitating elements (such as better quality of tax assessment service and less complex procedures) together with economic variables to achieve an optimum compliance level. The findings indicate that simply applying sanctions to improve Customs agents’ compliance, as is Royal Malaysian Customs current enforcement strategy does not optimise revenue yield. Appropriate reforms that go beyond sanctions and enforcement are recommended. It also identifies another essential but largely neglected strategy for improving compliance which is to work on improving the ethics of Customs agents, possibly by offering access to trade facilitation measures, or through coercion (e.g. public naming and shaming) and sanctions (e.g. withholding access to trade facilitation measures). Finally, this study also demonstrates the wide applicability of the TPB, including its application in tax compliance research and specifically in the context of import tax. The method (exploratory sequential mixed method) used in this study could also be used to replicate further studies to generate a more holistic compliance behaviour model.

Chapter 2, titled "Hotelling Tax Competition" shows how competition among governments for mobile firms can bring about excessive differentiation in levels of taxation and public good provision. Hotelling's Principle of Minimum Differentiation is applied in the context of tax competition and shown to be invalid. Instead, when an equilibrium exists, differentiation of public good provision is maximized. Non-existence of equilibrium, which is possible, is a metaphor for intense tax competition. The chapter also shows that, to some extent, perfect tax discrimination presents a solution to the existence problem created by Hotelling tax competition, but that the efficiency problem of Hotelling tax competition is exacerbated. Chapter 3 shows how the institutional rules imposed on its signatories by the GATT created a strategic incentive for countries to liberalize gradually. Ree trade can never be achieved when punishment for deviation from a trade agreement is limited to a 'withdrawal of equivalent concessions' , the most severe form of punishment allowed (Article XXVIII). Retaliation is not allowed to entail higher tariffs than those set by the initial deviant. If, in addition, tariff bindings (Article 11) limit an initial deviation from an agreement in a similar way, then efficient self-enforcing tariff reductions must proceed in a series of steps or 'rounds'. Chapter 4 provides an answer to the question "Why are trade agreements regional? " It argues that free trade agreements (FTAs) are regional because, in their absence, optimal tariffs are higher against (close) regional partners than (distant) countries outside the region. Optimal tariffs shift rents from foreign firms to domestic citizens. Lower transport costs imply higher rents and therefore higher tariffs. So regional FTAs have a higher payoff than non-regional FTAs. Therefore, only regional FTAs may yield positive gains when sponsoring an FTA is costly. To analyze equilibrium, standard theory of non-cooperative networks is extended to allow for asymmetric players. Naive best response dynamics show that 'trade blocks can be stepping blocks' for free trade.

This study developed and empirically examined a model to help understand how key individuals in businesses decide on whether or not to buy and utilise technology in the context of managing their obligations in relation to business taxes. In restricting the frame of reference to a taxation context, it enabled a link to be made to individual decision making, as it is ordinarily the case that one lead buyer is evident in the context of tax and technology in organisations. The model was developed from a review of extant literature in the areas of technology acceptance, behavioural intention, and consumer and business buying models. The overall model was built on a framework that has at its core the Augmented Technology Acceptance Model (Taylor & Todd, 1995a). A correspondence between attitude to use of the technology and product quality is theorised, allowing a connection to a wider model of purchase intention. The initial model was developed with thirteen hypotheses, ultimately leading to an examination of intention to buy tax technology. After an initial pilot study, in the main study a questionnaire was designed to capture empirical data for measurements related to the model. Data collected from 125 informants (i.e. senior tax staff in large organisations) about tax technology buying decisions they were currently considering was used to empirically test the model, using Structural Equation Modelling. The low sample size caused a need to simplify the original model to retain statistical power. This had the result of reducing the number of hypotheses to ten. The analysis was performed testing the measurement model and the model fit and thereby investigating its underlying hypotheses. The results supported the key hypotheses and the overall explanatory power of the model in examining intention to buy tax technology was strong. The use of technology acceptance principles as core to helping explain buying intention for tax technology was strongly supported. Only one hypothesis was not supported, relating to a proposed positive relationship between Relationship Quality and Intention to Buy constructs. Potential explanations for this finding with regard to relationship quality were introduced. The general research contributions and implications of the study were also discussed.

This thesis explores the way economic behaviour responds to taxation both theoretically and empirically. Chapter 1 studies the impact of transation taxes on the housing market, using UK administrative data and quasi-experimental variation created by notches, tax reforms, and stimulus. Transaction taxes have large effects on house prices and purchases, and adjustments to tax changes are fast. A temporary elimination of transaction taxes stimulated housing market activity by 20% in the short run (timing and extensive responses) followed by a smaller slump in activity after the policy was withdrawn (timing response). The success of this stimulus program stems from the large distortions created by the tax in the first place. Chapter 2 presents evidence on three ways in which firms affect workers’ earnings responses in Pakistan. First, third-party reporting of salaries by employers reduces evasion. Second, firms’ equilibrium salary-hours offers are tailored to aggregate worker preferences in response to adjustment costs in the labour market. Third, workers learn about the tax schedule from firms and become more responsive to taxation both contemporaneously (by 130%) and in subsequent years (by 100%). Third-party reporting does not eliminate misreporting: 19% of workers underreport their salaries, creating a loss of 5% of tax revenue, and indicating high returns to investments in improving enforcement. Chapter 3 develops a theory of optimal income taxation allowing for career effects of current work effort on future wages. Such effects are empirically important, but have been ignored by the optimal tax literature. We provide analytical characterizations that depend on estimable entities, including the elasticity of future wages to current work effort. We explore the magnitude of this “career elasticity” in a meta-analysis of the empirical literature on the returns to work experience and tenure, and provide numerical simulations calibrated to US micro-data. Our results show that career effects have important qualitative and quantitative implications for optimal tax design.

This thesis explores the argument that the need for governments to raise tax revenue, as opposed to relying on resource rents or other sources of non-tax revenue, may increase the likelihood that they will be responsive and accountable to their citizens. It employs a combination of quantitative and qualitative methods, first testing the relationship between tax reliance and accountability econometrically using cross-country data and then turning to detailed case studies from Ghana, Kenya and Ethiopia. The econometric results conclude that while existing data is consistent with the argument that tax reliance contributes to greater responsiveness and accountability, it is not possible to establish causality due to a combination of data limitations and the complexity of the underlying causal processes. This ambiguous finding provides motivation for the detailed case studies that follow. The causal model developed here proposes that the need for governments to rely on taxation may strengthen taxpayer demands for responsiveness and accountability, owing to the possibility of tax resistance and the role of taxation as a catalyst for collective action. Consistent with this model, the case study chapters present detailed historical narratives that capture significant examples from each of the three countries in which the need for taxation has contributed significantly to the expansion of responsiveness and accountability. As importantly, the case study evidence provides a nuanced understanding of the nature of the connections between taxation, responsiveness and accountability, highlighting three distinct types of causal processes at work, as well as the most significant social, political and economic contextual factors that shape the potential for tax bargaining. These lessons point toward important policy implications for foreign aid and tax reform more broadly.

The objective of the thesis is to provide a theory that explains the stylized facts regarding the trend of taxation policies, public spending and sovereign debt in advanced economies for the past couple of decades. The thesis focuses on distinguishing two types of public spending - productive investment and welfare payment and develops two different frameworks to examine the importance of the composition of these two types of public spending for economic growth and welfare. Chapter 2 presents a dynamic political-economy model in which voters decide tax rates and the proportion of public goods expenditure devoted to non-productive (but utility-enhancing) public goods. This non-productive public goods expenditure gives rise to a habit effect - it has to be at least as large as a fraction of last period value to provide utility. The median voter theorem applies. Starting from a steady state without the habit effect, its introduction leads to transitional dynamics that mimic several stylized facts: in particular, countries with higher income tend to have larger government and spend more on welfare programme. Chapter 3 studies the impact of public deficit on long-run economic growth by distinguishing the different types of government spending: investment and welfare payment. The model in this chapter predicts a non-monotonic or threshold effect in the relationship between public deficit and steady state growth rate. The composition of the public spending (the ratio between productive and non-productive) dictates the "threshold" in the national debt level. Countries which spend more on providing productive public goods could maintain a higher level of national debt that promotes growth.

This thesis attempts to resolve the deadlock to achieve an equitable division of taxes, and thereby internation equity. As the present tax laws were not negotiated, it is not considered here as fair. In this thesis it is proposed that an equitable division could be achieved through a division based on the levels of human development (combining Rawlsian schema and Sen’s capability approach). Therefore, it is argued that such a division would be equitable; because it would be based on entitlements (territorial claims), it would generate cooperation and thereby lead to greater efficiency. Importantly, this thesis establishes that the present tax treaties neither generate cooperation nor cohere with global welfare. Similarly, it is also found that the other institutions (OECD, IMF, WB and Dispute Settlement) related the tax regime presently do not promote development based on human capabilities. This could help developing countries to pursue a division favouring development (laws favouring development) and understand the institutions better suited to pursue such goals.

Quoting James Andreoni, `the problem of tax compliance is as old as taxes themselves'. The sources of missing tax revenues have traditionally concerned tax administrations and particularly now in times when public finances are striving. In the quest for analysing the revenue that is foregone, tax administrations have started to produce a report of their tax gap, understood as the difference between the theoretical tax liability and the actual collection, to obtain a measure of the extent of non-compliance. Due to the complexity of the non-compliance behaviour and the lack of visibility of certain types of income, different methods are usually put in place in order to offer a plausible range for the estimates. This dissertation dedicates its two first chapters to providing an alternative method for estimating the income-gap (de fined to be one minus the proportion of reported to actual income) for two populations: the self-employed and the employees. The underlying data used for both cases is publicly available survey data on expenditures and income that is generated on a timely manner. This carries substantial advantages. First, relying on a general purpose survey dataset means that the estimation can be updated more frequently than if it was to rely solely on either the timing of administrative data or on survey data that is speci fically targeted to measure non-compliance. Second, it provides an alternative estimation using an independent source of data which allows for the triangulation of the estimate obtained using administrative sources. Third, it allows tax administrations which do not have readily available administrative data to perform estimations using a type of survey widespread available in most countries. The third chapter of this thesis explores the role of the extrinsic and intrinsic incentives in explaining engagement in the hidden economy defined as undeclared work practices. This chapter contributes firstly to the literature on shadow economy and to the debate of whether crowding effects are found between extrinsic and intrinsic motivations in a tax environment.