The rapid adoption of machine learning in critical domains has intensified concerns about algorithmic bias and the fairness-performance trade-off. While existing literature often focuses on mitigation techniques, model evaluation still tends to assess performance and fairness in isolation. This study proposes a unified evaluation framework centered on a Fairness Score (FS) on a common [0, 1] scale. The framework builds on the Joint Fairness-Performance Index (JFPI), which combines predictive performance (F1-score) and a fairness composite derived from Demographic Parity Ratio, Equalized Odds Ratio, and Predictive Rate Parity. The framework then applies structural adjustments that penalize extreme imbalance and enforce a minimum justice requirement only when a model violates it. We treat the minimum justice threshold as the exogenous policy input and illustrate it with the Four-Fifths Rule. Given this requirement, the framework calibrates the weighting parameter a endogenously from the evaluated model set, rather than treating it as a discretionary preference. We evaluate the framework through controlled Monte Carlo experiments on synthetic credit-risk datasets and assess external validity on the German Credit benchmark. The empirical analysis shows that high-capacity models, especially neural networks, dominate the Pareto frontier in high-bias regimes and outperform linear baselines. Among mitigation strategies, pre-processing interventions provide the most robust results by improving fairness while preserving model stability and predictive integrity. The proposed framework provides a practical tool for responsible and reliable automated decision-making under explicit institutional fairness requirements.

This dissertation investigates the effects of unconventional monetary policy on the industrial
sector of the U.S. economy in the context of the COVID-19 crisis. The main objective is to
assess whether the large-scale quantitative easing programs implemented by the Federal
Reserve were able to mitigate the recessionary impact on industrial production and prevent a
deeper economic collapse. To this end, a Bayesian Mixed-Frequency VAR (MF-VAR) model
is employed, combining weekly and monthly variables from January 2017 to December 2024,
following the methodological framework of Schorfheide and Song (2015) and Feldkircher,
Huber and Pfarrhofer (2021). The results indicate that the expansionary shock associated with
increases in the monetary aggregate M2 did not generate persistent inflationary pressures nor a
robust expansion of aggregate output. Inflation responses remained moderate, while GDP
effects were transitory and statistically limited. In contrast, industrial production displayed
greater sensitivity to monetary expansion, suggesting the predominance of the financial
transmission channel, particularly through risk premium compression and asset price
revaluation.The empirical findings reinforce the post-Keynesian interpretation that there is no
automatic monetary multiplier capable of transforming reserve expansion into sustained output
growth or inflation acceleration. Investment decisions fundamentally depend on effective
demand expectations, and monetary policy alone has limited capacity to induce structural
productive expansion. The evidence suggests that quantitative easing primarily functioned as a
financial stabilization instrument, necessary to mitigate systemic risks but insufficient to
generate a self-sustained cycle of industrial growth without coordination with expansionary
fiscal policies.

This thesis examines the role of universities within innovation systems in developing countries, with particular emphasis on the Brazilian context and the space sector. It argues that universities should not be understood merely as providers of skilled labor, but as structural actors capable of shaping productive, technological, and territorial trajectories through their institutional choices, especially regarding the design of academic programs, curricula, and training strategies. In innovation systems characterized by strong state involvement and a relatively weak private sector in high-complexity technological fields, universities occupy a strategic position, although their influence is constrained by institutional limitations and fragmented public policies.

The thesis is structured around four interconnected articles. The first establishes the theoretical framework by critically revisiting the literature on National Innovation Systems and proposing a hierarchical and vertically organized systemic approach better suited to Global South contexts, in which the state plays a coordinating role in close interaction with universities and, potentially, with industry. The second article analyzes the historical evolution of Brazilian higher education, demonstrating how processes of marketization, curricular flexibilization, and institutional precarization have affected universities’ capacity to perform their educational, scientific, and innovative functions. The third article provides an empirical analysis of the Brazilian space innovation system based on data from the National Council for Scientific and Technological Development (CNPq), highlighting the central role of public universities, the strong dependence on state actors, and the high degree of institutional and regional concentration. The fourth article examines the emergence of a space cluster in Brasília, showing that it resulted from contingent institutional arrangements linked to the creation of an aerospace engineering program and the trajectories of its graduates, rather than from a deliberate and coordinated public policy strategy.

The findings indicate that, while universities can induce innovative dynamics and foster the emergence of new productive arrangements, the absence of integrated and long-term policy coordination often leads to fragile and potentially short-lived outcomes. The thesis concludes that stronger articulation between higher education policies, innovation policies, and territorial development strategies is essential to transform opportunistic initiatives into sustainable trajectories and to strengthen innovation capacities in strategic sectors such as the space industry.

This work aims to investigate the effects of macroprudential policy on credit concentration and the behavior of financial agents in a complex setting where they interact and learn about the environment. Using the bottom-up approach of agent-based models, we simulate a situation where banks, depositors, a central bank, firms, and a clearing house compose an artificial financial system under different scenarios regarding macroprudential policy instances. Information asymmetry is introduced in the lending market. We find that the prudential policy increases concentration, measured by the Herfindahl-Hirschman Index (HHI) and Concentration Ratio (CR), in the lending market, reducing the supply of loans to small firms. At the same time, the increased concentration does not appear to exert a negative impact on financial stability.

Argentina is a country that has experienced economic crises over the last two decades. To find out why, the role of institutions in the economy and their influence on a country's economic performance, as well as in its democracy, is analyzed. Specifically, it examines Argentina’s fiscal federalism is studied and its dysfunctionality. Based on the article by Gervasoni (2010), an index of subnational democracy is developed for each Argentine province and the city of Buenos Aires, which is the dependent variable in a random effects model. Using this model, we observe the influence of the transfer of federal resources to the provinces on their democracy, as well as the influence of the national leader with the results obtained, we conclude that both federal transfers and the national leader have an impact on subnational democracy

The idiosyncratic risk puzzle is a source of significant divergence between rationalists and proponents of behavioral finance theory. This debate began with the work of Malkiel and Xu (2002), in which the authors found that idiosyncratic risk was a statistically significant variable in explaining the cross-section of returns for U.S. stocks, supposedly providing evidence against the rationality of economic agents. Conversely, Pastor and Veronesi (2009) argued that there was a measurement issue in these models, as they did not disaggregate idiosyncratic risk from the risk arising from innovations and their uncertainties. Pastor and Veronesi (2009), therefore, treated the problem as an omitted variable bias and argued that if the model were properly measured, idiosyncratic risk would no longer be statistically significant. Given this debate, the present study opted to construct a portfolio of stocks representing a technological revolution in Brazil and analyzed the behavior of idiosyncratic risk, controlled for behavioral factors through the addition of an investor sentiment index. The results indicated that idiosyncratic risk was not significant in explaining the cross-section of returns for this portfolio, both in the absence and presence of the sentiment variable. When compared to a portfolio representative of stocks that had not undergone any technological revolution, the Chow test indicated no structural differences between the model's coefficients, challenging the central thesis of Pastor and Veronesi (2009).

This article evaluates the impact of the Bolsa Família Program on the consumption of goods that cause addiction of beneficiary households. To do so, the following variables are analyzed: (i) Proportion of spending on goods that cause addiction in relation to household income; (ii) Proportion of spending on goods that cause addiction in relation to spending on other goods; (iii) Proportion of spending on goods that cause addiction in relation to total household spending; (iv) Dummy variable that indicates whether there was consumption of a good that causes addiction. Microdata from the Pesquisa de Orçamentos Familiares (POF) for the years 2017 and 2018 are analyzed using the propensity score matching method. As robustness tests, Rosenbaum Sensitivity Analysis, Heterogeneous Effects by regions and schooling ranges are implemented.

Exploring a dataset spaning from 2005 to 2021, comprised of 38 countries, between advanced economies and emerging and developing markets, with and without a formal, single country inflation-targeting regime, across three different borrowing sectors, this research empirically investigates the complex relationship between macroprudential policy, inflation-targeting and credit growth. Using a macroprudential policy index composed of 17 different policy tools, it is possible to capture the net tightening/loosening stance and to measure it's effects on quarterly credit stock growth. The results points towards the negative effects of macroprudential policy tightening on credit growth, however, with some new important insights: inflation-targeting, especially for EMDE, is associated with a overall lower baseline credit growth. Inflation-targeting countries, both AE and EMDE, show a relatively faster response of macroprudential policy action effect, when compared to their non-targeting counter parts. On the other hand, Government borrowing seems to be less affected by macroprudential policy action in inflationtargeting countries.

This study examines the expansion of Distance Education (DE) in Brazil and the distinct student profiles associated with this modality, drawing on a novel administrative dataset. The empirical framework integrates four major data sources: (i) the School Census, which annually tracks K–12 students; (ii) the ENEM exam, Brazil’s primary gateway to public universities and basis for scholarships and student loans; (iii) the Higher Education Census, available since 2007, covering enrollment details, funding schemes, and learning modalities; and (iv) the Annual Social Information Report (RAIS), which records all formal labor contracts in the country. By leveraging masked social security numbers (CPF), we can link students’ academic and employment trajectories over time. Our sample comprises successive cohorts of 9th-grade students from 2007 to 2013, each stratified by state, gender, school sector, and race—amounting to about 50,000 students per year. Additionally, Prova Brasil assessments offer baseline proficiency measures in Portuguese and Mathematics for most of the 2011 and 2013 cohorts, allowing us to account for prior academic achievement. Findings reveal that DE particularly attracts older students, women, and graduates from public schools or otherwise disadvantaged educational backgrounds, underscoring the modality’s inclusive potential to broaden higher education access.

This dissertation aims to analyze the functioning of parliamentary amendments in the National Congress of Brazil, their interactions with the Executive Branch, and the implications of recent legislative changes, especially with the enactment of Complementary Law No. 210 of 2024. The study seeks to understand the dynamics of the allocation of amendments, focusing on individual amendments, committee amendments, and state delegation amendments, as well as the repercussions of these practices on the legislative process, fiscal management, and the balance of public accounts in the country. The research uses descriptive statistical methods to assess the distribution of resources and the effects of amendments to the federal budget. In the Brazilian context, parliamentary amendments have been a crucial mechanism in political negotiations between lawmakers and the Executive Branch, being used as a tool to secure political support and fund projects in the electoral bases of congress members. The allocation of these amendments, particularly committee amendments and state delegation amendments, has sparked intense debates about transparency, efficiency, and the fiscal impact of this practice on the country. The allocation of these amendments often reflects a logic of political compensation, which may influence the prioritization of regional projects at the expense of other more urgent budgetary needs. Complementary Law No. 210 of 2024, prompted by a decision of the Federal Supreme Court, brings significant changes to the regulation of these amendments, aiming to increase control over their execution and reinforce the principles of public spending planning and rationalization. The legislation establishes new rules for the distribution of resources and the definition of priorities, with the goal of reducing practices like the "secret budget" and increasing transparency, as well as fostering greater fiscal responsibility from both lawmakers and the Executive. The descriptive statistical analysis conducted in the study allows for an assessment of the evolution of amendments over time, the regions most benefited, and the impact of these allocations on public finances. In addition to the legislative transformations, the fiscal impact of parliamentary amendments is one of the main points of discussion in this dissertation. The practice of allocating large amounts of resources to local projects, while legitimate from a political perspective, generates a considerable impact on public finances, especially when resource allocation does not follow a logic of economic efficiency or national necessity. The study examines how committee amendments, which are directed to specific areas of public policy, and state delegation amendments, which support regional projects, contribute to the increase of the government’s discretionary spending, affecting fiscal stability and the execution of other essential public policies. Throughout the dissertation, the technical and political aspects involved in the formulation and execution of parliamentary amendments are discussed, aiming to provide a critical view of the advances and challenges of the Brazilian budgetary system. The research contributes to the debate on the efficiency of the fiscal system and best public management practices in the country, highlighting the need for improvements in the control and transparency of the allocation of parliamentary amendments, in order to mitigate negative impacts on public finances and promote greater fiscal fairness.

This study investigates the Laffer Curve through a real-world effort
experiment with 292 participants, conducted in classrooms. The main objective is to
observe a relationship between tax rates and revenue collection, indicating a Laffer

Curve, and to test the impact of information about the use of collected resources
(donations to a charity) on the behavior of participants. The effort activity consisted
of drawing pictures for eight minutes, with payment given in raffle tickets proportional
to the effort, and different tax rates (20%, 40%, 60%, and 80%). The control group
performed the task without any additional information, while the treatment group was
informed of the social destination of the collected resources. The results indicate that
the average revenue collection increased with the increase in the tax rate, with no
evidence of a classic optimum point of the Laffer Curve within the tested range.
Furthermore, the treatment group showed greater average effort and revenue
collection, although without statistical significance at the 95% confidence level. It is
observed that, although the experiment does not confirm a traditional Laffer Curve, it
suggests that normative and social factors, such as the perception of the use of
taxes, can influence the behavior of participants.

This study investigates the motherhood penalty in the Brazilian labor market, which refers to the wage and occupational disadvantages faced by women due to the presence of children, even after controlling for observable characteristics such as education, experience, and type of occupation. Using data from the 2022 and 2023 Continuous National Household Sample Survey (Continuous PNAD), with particular emphasis on the educational supplementation from the second quarter of 2023 and the fifth wave of 2022, the analysis examines wage differences between men and women, as well asthe disadvantages faced by mothers compared to childless women, both in terms of earnings and labor market participation. The study considers variables such as the presence of children in specific age groups, access to daycare, and the domestic workload. Based on literature and empirical data, the main hypothesis is that women earn lower wages due to the greater time devoted to caregiving and household tasks, which increases their reservation wage and reduces their participation and advancement in the labor market. Thus, this research contributes to a deeper understanding of the mechanisms that sustain gender inequality and the motherhood penalty within the Brazilian context.

This master’s thesis studies the relation between local governments’ public spending and the reelection probability of it’s mayors in Brazil. Under the political budget cycles framework with a panel probit model, its investigated which budget functions yield a higher electoral reward when the incumbent focuses the spending on the second half of the term. Aditionally, the voters’ characteristics are interacted with the budget functions to identify if those rewards change with the populations heterogeneity. It was used the municipalities spending data by functional classification from the National Treasury Secretariat (STN), the results of the municipal elections from 2008 to 2020, social security data (Cadastro Único), and the electorate’s statistics from the Superior Electoral Court (TSE). The results show that there’s a positive reward to spending at the end of the term on the functions of Culture, Sports & Recreation and Health, and negative to Administration. Also, there’s a significative bonus when the mayor is of the same party as the governor, if it has a degree, and if is a man. Concerning the electorate’s characteristics, it was found that the age plays an important role on the budgetary preferences: the marginal effect of pre-electoral spending with Health and Culture grows the higher the proportion of elders is, an inverse of what is found with Sports & Recreation and Education, functions more rewarded by younger voters. About education level, the marginal effect of Health was higher among the municipalities with the higher percentages of people with secondary education

This study examines digital financialization as an extension of the financialized regime, analyzing its impact on Brazil's economic system and its global implications. Drawing on a heterodox perspective influenced by Marxist and Post-Keynesian approaches, the research investigates how advancements in digital technologies have intensified financialization processes, heightening challenges and exacerbating their economic and social effects. The study begins by exploring financialization as a historically situated process, characterized by a shift from a production-driven regime to one dominated by finance, where fictitious capital and speculation play central roles in accumulation. It then considers digital financialization as an evolution of this regime, highlighting the role of technological innovations—such as automated trading, crypto-assets, and economic platformization—in reshaping financial markets and expanding financialization’s reach. Particular attention is given to the role of the State as a key promoter of financial digitalization and the capture of public policy by financialized agendas, especially in developing economies. In the Brazilian context, initiatives such as PIX, the Digital Real, and Open Finance are analyzed, focusing on their effects on financial inclusion, data economies, and market reorganization. The findings reveal that while digital financialization introduces opportunities for innovation, it also deepens inequalities, destabilizes local economies, and exacerbates systemic instability risks. The study concludes that regulatory and public policy efforts must prioritize social justice and safeguard the real economy against the pressures and impacts of financialized logics

This work consists of three papers on Natural Language Processing (NLP) applied to economics and finance. The first paper surveys the main methods, requisites and applications of NLP in the field. Structured to help shorten the path economic researchers have to trail to get introduced to these methods, our work covers the following topics: (i) useful NLP tasks to economics; (ii) NLP models; (iii) economic and financial textual data for NLP; and (iv) NLP applications in economics and finance. We further contribute by resorting to bibliometric tools to help us visualize the literature map of this field, also providing valuable insights to our survey and the remainder of our work. We finally indicate that there is much room to apply natural language processing to economic issues, but alert that, more than ever, researchers must be careful not to stray away from questions motivated by hypotheses closely tied to economic theories. The second paper provides a forward-looking measure of how central bankers implicitly coordinate their actions, as measured from public manifestations in the form of their speeches’ transcripts. In order to do that, we resort to the the central bankers’ speeches database made available by the Bank for International Settlements and build a network of similarities that connects central banks, adapting for this context the method proposed by Cajueiro et al. (2021). Our results show that our network successfully captures the long-term global importance of central banks overseeing the G10 currencies, with word-level point to evidence of their orthodox approach to monetary policy. We further explore this framework on a dynamic setting, with findings that indicate that coordination tends to increase in times of economic stress, as in the years of the Great Financial Crisis and the period after the Covid-19 pandemic. An evolution analysis on word occurrences then shows that our proposed measure is driven by mentions to policy instruments and economic views proffered by policymakers. The third paper proposes a framework for estimating expectation-embedded multi-dimensional sentiment from monetary policy communication, combining economic fundamentals and state-of-the-art deep learning neural networks. The economic basis of our approach is set by the Litterman and Scheinkman (1991) yield curve decomposition model, with its level, slope and curvature factors accounting for the three dimensions of our sentiment gauge. For text modeling we incorporate in our framework the Bidirectional Encoder Representations from Transformers, BERT (Devlin et al., 2019). In an application to policy communication by the Brazilian Central Bank, our results reinforce the need for more than one sentiment dimension to comprehensively capture the relevant nuances of communication for monetary policy assessment. These results indicate that the added dimensions complement the usual hawk/dove gauge of policy sentiment, and can at times be more relevant in setting the overall tone of policy communication, eventually even preceding shifts in monetary policy stance.

Humanity's main challenges today are reversing the progressive loss of biodiversity and ecosystem services. Now on the decision-maker's agenda, this challenge leads to a necessary connection between conserving biodiversity and promoting human well-being. Moreover, in this way, it emphasizes the responsibility and the need for action by governments for its promotion. It then becomes necessary to assess the actual size of the contribution of governments to face this challenge. Therefore, this work aimed to evaluate the contribution of the federal government of Brazil to the conservation of biodiversity between the years 2000 and 2019 and to evaluate the development of funding for conservation as a field of research. Four essays were developed to achieve these goals: The first is a systematic review of the literature on Conservation Economics and Conservation Financing. Then, three studies were developed based on the stages of the Finance for Biodiversity Initiative – BIOFIN developed by the United Nations. The second essay is a political and institutional review study related to biodiversity within the scope of the Brazilian federal government. The third essay is a Review of Spending on Biodiversity carried out by the federal government through a survey, analysis, and description of financial resources allocated to conservation. The third essay is a funding gap analysis. The results showed that the number of studies on funding for conservation has been growing. Nevertheless, it is still an area with many knowledge gaps and an underdeveloped theoretical foundation and methodological framework. Regarding the results of applying the BIOFIN approach to the case of the federal government of Brazil in the period evaluated, there were at least 21 norms and about 195 instruments provided for in federal legislation to comply with the National Strategy and Action Plan for Biodiversity – EPANB. However, most budgetary actions carried out by the management bodies of biodiversity policies at the federal level distort the guidelines provided for in these listed norms. The results also showed a difference between direct spending on biodiversity conservation of around R$176 million/year (or 0.006% of GDP and 0.029% of the federal budget) and indirect spending of around R$1.5 billion/year (or 0.029% of GDP and 0.146% of the federal budget). We estimated a funding gap of R$411 billion. When adjusted by BIOFIN factors, this gap is BRL 155 billion. Finally, the results showed a dimension of more than R$ 50 billion annually between existing and potential resources in ten listed financial solutions that can be applied to conservation financing. These results demonstrate, on the one hand, the downward trend in expenditures observed in the evaluated period, placing Brazil in a preoccupying situation. Since spending on biodiversity is below established international targets, on the other hand, it is necessary to consider the difficulties presented concerning the accounting methods for these expenses. Since there is still much imprecision in defining what is spent on biodiversity and in assessing its effectiveness, it is necessary to consider the results found here as a first effort to indicate a more accurate diagnosis of the situation. 

This thesis examines public debt issuance strategies, focusing on the frequency (biweekly or weekly) of auctions, bid selection criteria (uniform or discriminatory pricing), competitiveness and formats (hybrid or pure auctions) used by the Brazilian National Treasury between 2014 and 2023. First, we assess the impact of increasing the frequency of NTN-F auctions to weekly, observing reduced fluctuations in yields before and after auctions, as well as stable secondary market traded volumes. Data suggest rate volatility around auctions ("V" effect in prices) is associated with limited risk absorption capacity of bidders. Reduced model analysis of bid selection criteria presents evidence of lower premiums in uniform-price auctions. Also, contrary to theoretical expectations, there were significant higher premiums in hybrid auctions relative to traditional ones. In addition we examined networks and competition in auctions. Our simulations and tests point to absence in strategic bid coordination. Analyzing groups of bidders, we found that dealers bid more aggressively, probably due to obligations and privileges of dealership.

This work comprises two studies examining outsourcing and labor market trends. The first study analyzes the impacts of Brazil’s 2017 labor reform and outsourcing laws on employment patterns, highlighting a shift from traditional formal employment to more flexible, independent work structures, such as Individual Microentrepreneurs (MEI).  The findings reveal that labor reforms and deregulation accelerated the growth of individual businesses, with heterogeneous effects across industries and regions. The second study aims to uncover the economic drivers behind labor outsourcing and worker registration as individual entrepreneurs, as well as the financial implications of this transition for different worker profiles. The second study aims to investigate the incentives driving labor outsourcing and the formalization of workers as individual entrepreneurs, as well as the financial implications of this transition for different worker profiles. The results indicate an initial improvement in financial activity indicators immediately after transitioning to entrepreneurship, evidenced by increased credit balances. However, these gains are followed by significant and persistent financial losses after the first year of business operation, underscoring the sustainability challenges faced by new businesses. These findings underscore the need for targeted policies, including financial education, improved credit conditions, and the implementation of accessible and well-structured retirement programs, to provide adequate support for this growing contingent of workers.

This Thesis examines the interaction between monetary and fiscal policies in Brazil, focusing on the implications of bounded rationality in economic modeling. It seeks to compare the impacts of rational and behavioral approaches on macroeconomic variables in response to policy shocks, aiming to provide a deeper understanding of Brazil’s economic dynamics. The study utilizes a new Keynesian DSGE model with Bayesian estimation, analyzing quarterly data spanning from 2000Q1 to 2023Q4. The model evaluates the differential responses of macroeconomic variables — GDP, primary surplus and deficit, private consumption, inflation, public debt and interest rate — under rational and behavioral approaches. The analysis reveals that, while the rational approach is theoretically consistent, its assumptions of instantaneous adjustments and fully rational agents render it less applicable to Brazil’s economic context. In contrast, the behavioral approach better captures the gradual adjustments and adaptive responses of economic agents, accounting for cognitive limitations and biases. This research contributes to the field of behavioral macroeconomics by applying the concept of bounded rationality to Brazil, an economy marked by structural volatility and adaptive policy dynamics. Finally, by integrating recent scenarios, such as the effects of the COVID-19 pandemic, this manuscript broadens the understanding of how economic shocks impact selected macroeconomic variables. As a result, it not only advances theoretical economic modeling but also serves as a practical tool to guide the formulation of more effective public policies tailored to the complexities of the Brazilian economy.

This doctoral thesis comprises three articles that use novel data on Brazilian twins to investigate key issues in labor economics.
The first article introduces TwinsBR, the largest twin dataset ever constructed, based on Brazilian administrative records.
The second article estimates the gender wage gap by comparing opposite-sex twins, controlling for unobservable factors such as family background and latent traits.
The third article examines the motherhood wage penalty by comparing same-sex female twins with and without children, finding a smaller effect than previously reported in the literature. By exploring the analytical potential of twin data in the Brazilian context, this thesis provides new empirical evidence on gender disparities. The findings highlight the importance of accounting for unobservable factors in economic analysis. The work also opens space for new methodological approaches in applied research.

ndustrial policy has re-emerged as a central pillar of development strategy, renewing longstanding debates about its design, implementation, and effectiveness. This dissertation contributes to that debate by advancing the Economic Complexity and Relatedness framework to better inform structural transformation and policy targeting. The first paper, Complexity Traps in the Product Space: Why Some Countries Get Stuck in Local Maxima, investigates how economies can become structurally constrained when certain products offer returns that are disproportionately high relative to their neighbors in the product space. These localized incentives make it more attractive to exploit products that function as local maxima—but with low complexity—than to explore nearby opportunities that would promote capability accumulation. While appealing in the short term, this pattern ultimately constrains diversification, limiting the range of activities a country can develop over time. Over the long run, such constraints give rise to persistent regions of structural stasis - complexity traps - where countries struggle to transition into more complex activities. These findings highlight the need for industrial strategies that deliberately counteract the distorting effects of local optima and promote broader diversification. The second paper, Less is More: How Relatedness Filtering Enhances Productive Upgrading Predictions, demonstrates that statistical noise in the product space can hinder accurate identification of viable diversification paths. By filtering weak and spurious connections, we significantly improve the ability to predict future productive transitions—especially for less diversified economies—offering a more precise empirical foundation for strategic industrial policy. The third paper, From Capabilities to Economic Convergence: A Structural Growth Framework Linking Economic Complexity, Institutions, and Human Capital, proposes an integrated model that explains both current income levels and future growth using a multidimensional view of capabilities. It introduces a new complexity measure based on input–output data, which captures the sophistication of production networks beyond trade flows. The results show that multidimensional complexity, institutional quality, and human capital jointly shape development trajectories, and that countries with unexpectedly high complexity relative to their income tend to grow faster. Together, the three studies offer both diagnostic and prescriptive contributions to the Economic Complexity literature, helping to identify structural bottlenecks, improve targeting of policy tools, and reframe long-run development strategies.

This thesis analyzes the role of municipal public administration as an agent that fosters local economic development, focusing on the collection of the Service Tax (ISS) and the influence of tax enforcement — measured by the police power fee — on the economic performance of municipalities in the state of Goiás, Brazil, between 2015 and 2021. The study is based on the premise that fiscal autonomy is one of the pillars of federalism and that strengthening administrative and enforcement capacity is essential for consolidating local governance and promoting regional development. The research adopts a quantitative and empirical approach, applying panel data models and spatial econometric models, particularly the Spatial Durbin Model (SDM), to capture both the direct and indirect effects of tax collection on municipal GDP per capita. Spatial econometrics enabled the identification of interdependencies among municipalities, showing that fiscal and economic performance are not isolated phenomena but rather part of spatial and institutional patterns that shape regional development in Goiás. The results confirm the central hypothesis that stronger fiscal enforcement positively affects ISS revenue and, consequently, municipal economic development. Municipalities with greater administrative and fiscal capacity show better economic performance, underscoring that management efficiency is a key determinant of local development. Furthermore, the presence of spatial spillover effects reveals that a municipality’s fiscal performance can influence that of its neighbors, highlighting the need for cooperative and regionally integrated public policies. From a theoretical and institutional perspective, the thesis demonstrates that local development depends not only on productive expansion but also on sound fiscal structures and efficient public management practices. The consolidation of ISS revenue, together with modern and transparent fiscal enforcement, contributes to reducing regional inequalities, expanding financial autonomy, and strengthening public investment capacity. Therefore, municipal tax administration, when guided by efficiency and fiscal justice, stands as a strategic vector for sustainable economic and social development at the local and regional levels.

This thesis is in the field of Economics of Pensions (ECP) and addresses two primary gaps in the literature: (i) the lack of a comprehensive and systematic mapping of existing scientific research in ECP; and (ii) the absence of a multidimensional ex-ante assessment of the expected impacts of a structural pension reform, introducing a nonfinancial defined contribution system (NDC or notional accounts) into the Brazilian national pension scheme for private-sector workers (RGPS). Chapter 1 – Economics of Pensions: a Scientometric Mapping – aimed to present a structured overview of ECP research, having been motivated by the observed thematic and methodological fragmentation in prior studies. The methodology employed a broad set of scientometric techniques, combining performance analysis (production and impact) and mapping (theoretical and thematic relationships), based on 7,230 Web of Science documents published between 1956 and 2023. The results revealed that past research had a theoretical and methodological structure marked by a micro–macroeconomic duality as its foundation, analytical rigor, pluralistic and interdisciplinary approaches, empirical sophistication, and a concentration in high-impact journals. The social and institutional configuration of the field indicates growth, maturation, and institutionalization, with research and citations concentrated in a few key sources, alongside increasing scientific collaboration and international networking. The study identified that past research had diverse research fronts, connected to macroand microeconomic agendas, consolidated lines on supplementary pensions, insurance, annuities, and finance, and emerging frontiers related to health, well-being, behavioral finance, pension planning, and financial literacy. It concludes that ECP is a mature, dynamic, and interdisciplinary field whose research agenda remains closely aligned with public policy and pension reform debates. Chapter 2 – An Evaluation of a Non-Financial Defined Contribution Pension System for the Brazilian RGPS – conducted an ex-ante evaluation of the impacts of a structural RGPS reform introducing an NDC scheme for new contributors. The reform was motivated by its potential to enhance sustainability and transparency, while incentivizing contributions and delayed retirement. The methodology utilized a dynamic actuarial microsimulation model, using representative groups of contributors constructed from microdata from the Continuous Brazilian National Household Sample Survey (PNADC/IBGE) and the Brazilian Social Security Statistical Yearbook (AEPS). The analysis compared indicators of adequacy, sustainability, and equity between the current system (Constitutional Amendment 103/2019) and the proposed NDC design under different scenarios of contribution density and risk contribution rates. The use of synthetic indicators – likely unprecedented in the Brazilian literature – enabled a clearer understanding of the trade-offs inherent in pension policy evaluation. Overall, the results suggest that adopting the NDC model could improve RGPS sustainability by reducing actuarial liabilities through its quasi-automatic adjustment to demographic changes. However, it would result in less adequate benefits and heterogeneous effects on equity, reducing the progressivity of pension policy and disproportionately impacting women and low-density contributors. The findings highlight that an NDC reform would make the tradeoff between adequacy and sustainability more explicit, thereby enhancing transparency and resilience. In conclusion, the thesis establishes a direct connection between the ECP research agenda and public policy needs. The first study consolidates the conceptual and methodological foundations of the field, while the second applies these foundations to a reform consistent with international research concerns. The results demonstrate that balancing sustainability and adequacy lies at the core of contemporary pension reforms, reaffirming the relevance of ECP as a vital tool for designing and evaluating pension policy.

Introducing a law establishing a minimum wage for workers aims to provide a better quality of life for those who receive it, particularly for the poorest individuals. However, implementing a minimum wage may increase the probability of workers who receive it becoming unemployed. By exploring this trade-off between higher minimum wages and a greater likelihood of job loss, I intend to measure the impact of minimum wage increases on unemployment in Brazil using panel data from the Monthly Employment Survey (PME) and the empirical strategy of differences in differences.

This dissertation investigates the use of advanced models for forecasting ICMS revenue in Minas Gerais, employing Long Short-Term Memory Recurrent Neural Networks (LSTMs). The study proposes an innovative approach in tax data analysis, highlighting the ability of LSTMs to capture and learn complex temporal patterns. The dataset includes historical revenue information, and other relevant economic variables. The methodology involves training and validating the models based on this data, resulting in an enhancement of tax forecasting techniques. The results demonstrate the effectiveness of LSTMs in anticipating ICMS revenue, providing valuable insights for more efficient fiscal management.

This work consists of an impact assessment of the adoption of the Social Health Organizations (OSS) management model in hospital units. This management model, established by Law No. 9,637/98, was conceived as a reflection of the paradigm shift brought about by New Public Management and is presented as an alternative to the rigidity of the State in the provision of public policies, seeking greater efficiency by establishing partnerships between the Public Sector and the Third Sector in the provision of goods and services that are not exclusive to the State. An econometric model of Differences-in-Differences with two-way fixed effects, at the level of hospital 

unit and year, was developed to estimate the effect of the implementation of OSS management on the mortality rate from Primary Care-Sensitive Hospitalizations (PCSH). Additionally, the Staggered Differences-in-Differences model was applied as a robustness test to validate the results and ensure that the observed effects were consistent over time, considering the different adoption dates of the treatment. The estimated results indicate that OSS management significantly contributed to the reduction of the mortality rate from Primary Care-Sensitive Hospitalizations, both in the model without control variables and in the model with these variables. This result supports the hypothesis that OSS management is more efficient in managerial aspects, which are reflected in health indicators, as this group of primary diagnoses indicates illnesses that can be clinically treated when various procedures within Primary Care are followed, such as preventive measures, correct diagnosis, and appropriate treatment. Thus, the study constitutes a contribution to the understanding of the impacts of this management model, corroborating the hypothesis that by ceasing to be a direct provider of public policies and assuming a regulatory role, the State can achieve more efficient results in the provision of public policies.

The post-keynesian theory, particularly Minsky's financial instability hypothesis, suggests that variables related to agents' expectations may have a greater ability to anticipate the behavior of credit granting to companies by private banks than variables related to the economic real situation. Based on this concept, this dissertation aims to identify, among the time series provided by Bacen, FIPE, FGV, B3, and IBGE, which ones are most closely aligned with the behavior of credit concession to companies by private banks through working capital loans in Brazil between 2010 and 2016. To this end, a graphical comparative analysis of the series is performed.

This thesis presents three essays in applied microeconomics. The first chapter investigates the impact of tourism on educational outcomes. To estimate causal effects, I use a instrumental variables approach. The main findings are consistent with the hypothesis that tourism increases the GDP level and raises the opportunity cost for children and young individuals to study. The second chapter explores the effects of tourism on public finances. Once again, the instrumental variables approach is utilized. The results indicate that tourism enhances the fiscal capacity of municipalities, which is associated with improved infrastructure and the selection of better politicians. Finally, the third chapter examines the influence of party alignment during the period of military dictatorship on political variables by exploring the existence of two parties throughout the Brazilian military regime — one party aligned with the military regime and the other party opposing the regime. The results suggest no evidence that municipal-level party alignment is related to electoral manipulation, public finances, and the provision of public goods.

This work comprises three studies on economics in big data contexts. The rst analyzes the impact of ESG (Environmental, Social, and Governance) news on the stock returns of leading Brazilian companies, using an unprecedented Dictionary of ESG Terms specically developed for this study to select and classify news according to the standards of the Sustainability Accounting Standards Board (SASB). The research indicates that only news with content that is nancially material to investors inuences stock returns. In other words, investors do not react for reputational or non-pecuniary reasons. The second study explores the high-frequency predictability of the Brazilian exchange rate (at the 1, 5, and 15-minute frequencies), employing both machine learning techniques and traditional linear regression for forecasting. Two types of exercises are conducted: one with contemporary predictors and another using out-of-sample data. We show that it is possible to beat the benchmark, the Random Walk, over a horizon of up to four minutes at a frequency of 1 minute. We also show that the most important predictors are those that carry local information, as well as the exchange rates of the BRICS or countries with economies similar to Brazil’s. When the rates from B3’s foreign exchange futures contracts are considered as predictors, we can beat the Random Walk over a horizon of up to 6 minutes. The third study measures consumption inequality at the municipal level using data from electronic payment methods, specically data from credit card and Pix payments. Furthermore, as an application, we examine the relationship between inequality and economic complexity. We demonstrate that greater economic complexity is associated with lower consumption inequality, marking the rst assessment of this kind for Brazilian municipalities.

This study examines the impact of asymmetric remedies imposed on service providers with Significant Market Power in Brazilian uncompetitive municipalities within the wholesale markets of dedicated lines, high-capacity data transport, and fixed access network infrastructure for data transmission. Employing a combination of Differencein-Difference analysis and Propensity Score Matching techniques, this inquiry explores how wholesale price regulation influences market concentration, service density, small providers' market share, fiber investments, user grievances, and perceived quality, particularly with respect to pricing and service excellence. The findings unveil a spectrum of effects across diverse remedies and indicators, leading to a proposition advocating the complete deregulation of fixed access network infrastructure for data transmission, the partial deregulation of dedicated lines, and a heightened focus on the pivotal high-capacity data transport wholesale market. Notably, the study underscores the need for robust analysis and careful consideration of the implications of asymmetric remedies on market dynamics and regulatory policy

This study aims to model optimum incentive structures in monetary sanctioning processes led by the Regulatory Telecommunications Authority in Brazil. The imposition of fines has historically been a tool used by the regulator to enforce the regulations it has established. Over time, a scenario has been observed with tens of thousands of fines imposed and billions of reais in unpaid fines. Since 2012, regulated entities have the prerogative of a 25% discount on the imposed penalty, provided they do not contest it and pay in the first instance. This research aims to investigate the implicit behaviours of the regulator and regulated entities in decisions regarding compliance with norms and fine payments, as well as to examine, in theoretical terms, conditions for establishing ideal incentive structures for the quick and efficient resolution of regulatory infractions. Therefore, utilizing game theory modelling, this work seeks to support any restructuring of the regulator's incentive policy in telecommunications.

This study aims to analyse how technological change brought by digitalization may impact economic life in several separate but correlated features. The rapid evolution of digital technology, accelerated by the impact of the COVID-19 pandemic, is fundamentally reshaping economies and altering growth patterns. Digital technologies, particularly those associated with Digital Transformation, have profound economic and social implications, transforming human relations and institutional frameworks. Despite its benefits, embracing digital transformation presents challenges, triggering societal upheavals with winners and losers. This research focus on five main economic features of this digital technological transformation: the macroeconomic impacts of digitalization, including the i) productivity paradox; and the ii) reshaping of development strategies due to the augmenting economic relevance of the service sector; iii) widening social inequalities due to the digital divide; iv) increased market concentration led by the ascent of information technologies; and v) how new technologies such as artificial intelligence may impact the future of labour markets. The analysis is done by a literature review for each of those economic features, complemented by policy implications derived from the research.

This study examines the impact of reelection on the fiscal behavior of incumbents in terms of public spending, receipt of intergovernmental transfers, and strategies for deferring payments through prior-year expenses (DEA) and unsettled (RPNP) and settled (RPP) carryforwards. It compares expenditures, transfer revenues, and payment deferral mechanisms of mayors in their first and second terms during close elections from 2005 to 2020, utilizing a regression discontinuity approach. The findings suggest an increasing trend in the seasonality of political budget cycles over the period, with heightened concentration of revenues and expenditures in election years in municipalities with first-term mayors facing competitive elections since 2013, particularly in intergovernmental transfers and agreements, personnel expenses, other current expenses, and investments focused on education, sports, and leisure. In this context, the strategy of deferring payments is only advantageous for reelection when a mayor cannot maximize transfer revenues in election years, as was the case between 2005 and 2012. Therefore, this study underscores the need for (i) refinement of the Fiscal Responsibility Law and the Electoral Law to restore their original capacity to control public spending during electoral periods, (ii) addressing solutions for the current context of federal imbalances, and (iii) incorporating automatic triggers in the design of fiscal rules, providing flexibility for managers in times of constraint, to prevent "creative accounting."

This thesis consists of two papers that investigate the impact of FIES, the primary higher education nancing policy in Brazil, on access to college and on the behavior of higher education institutions. The identication strategy relies on a reform that implemented regional quotas for loans granted. These quotas depend on regional Human Development Index (HDI) values in a discontinuous way, allowing a quasi-experimental design that leverages both the policy change and the discontinuities introduced by the new allocation rule. In the rst paper, we nd that each additional loan leads to approximately 0.2 additional college graduates in six years. However, eects are quite heterogeneous, and concentrated mostly in for-prot higher education institutions (HEIs) and, notably, in evening programs. We nd that the impact of loans on initial enrollment is higher for more advantaged students, specically those coming from private high schools. Impact on graduation, on the other hand, is higher for students coming from public high schools, who are more likely to be nancially constrained. Further analysis suggests that less advantaged students increase their participation in the selection exam following an expansion in loan availability. However, they usually present lower scores, limiting their capacity to access loans in more competitive majors. Thus, despite highlighting the signicance of nancial constraints, our results also indicate that poor academic preparation seems to be an equality important obstacle to accessing higher education for students from less advantaged backgrounds. In the second paper, we investigate how for-prot higher education institutions (HEIs) respond to the availability of government funding. We nd that increased loan availability results in a signicant boost in revenue for such institutions, of $0.73-$0.78 for each additional $1 in loan disbursements by the federal government. Each $1 increase in revenue results in approximately a $0.4 increase in institutional prots, with the remainder resulting in increased expenses, especially in labor-related costs. Markup estimates are quite high on average (64%), and are higher for higher quality HEIs, less selective HEIs and HEIs that do not face competition of public universities. Nevertheless, we also nd indicative evidence that institutions take advantage of this revenue shock to improve quality standards by hiring permanent faculty with better credentials (master and doctoral degrees). Oversight of higher education programs seems to have a role in this behavior, since spending increases in areas included in annual quality assessments, resulting in higher quality scores for the institution

In the early 1930s, Nicholas Kaldor could be classified as an Austrian economist. In the theory of capital wars at that time, for instance, Kaldor defended the Austrian theory of capital against the offensive of Frank H. Knight. We reconstruct the intertwined paths of Kaldor and Friedrich A. Hayek to disequilibrium economics through the theoretical deficiencies exposed by the Austrian theory of capital and its consequences on equilibrium analysis. In particular, the integration of capital theory into a business cycle theory by the Austrians and its shortcomings - e.g., criticized by Piero Sraffa and Gunnar Myrdal - called attention to the limitation of the theoretical apparatus of equilibrium analysis in dynamic contexts. This was a central element to Kaldor’s emancipation in 1934 and his subsequent conversion to John Maynard Keynes’ General Theory (1936). In addition, it was pivotal to Hayek’s reformulation of equilibrium as a social coordination problem in “Economics and Knowledge” (1937). It also had implications for Kaldor’s mature developments, such as the construction of the post-Keynesian models of growth and distribution, the Cambridge capital controversy, and his critique of neoclassical equilibrium economics.

 This thesis is composed of three independent chapters, albeit correlated, which analyze the adoption of pandemic control measures by Brazilian mayors and the impacts of the pandemic on the electoral performance of mayors in the 2020 elections and of Jair Bolsonaro in the 2022 presidential elections. The first issue addressed concerns how the reelection incentive influenced the adoption of pandemic management policies by Brazilian mayors in 2020. The analysis, conducted through a regression discontinuity approach comparing the adoption and duration of pandemic control policies in municipalities that had elections in 2016 decided by a small margin of votes, indicates that mayors eligible for reelection were timider in adopting non-pharmacological measures to combat the pandemic. This result is observed especially in municipalities without local journalism presence and governed by mayors affiliated with right-wing parties, suggesting that the presence of local media and the mayor's party spectrum influence electoral incentives. The second issue analyzed evaluates the impacts of the pandemic on the electoral performance of mayors running for reelection in 2020. Using beta regression methodology to model the incumbent's percentage of votes and the probit model to outline the probability of election, it was identified that the number of Covid-19 cases was positively correlated with the mayor's votes and their probability of reelection, while deaths were negatively correlated. This result is consistent with an electorate that rewarded mayors who invested in identifying suspected cases of the disease and reinforced the municipal healthcare system, thus reducing its fatality rate. However, this effect was not uniform across all municipalities, but notably observed in municipalities governed by left-wing party-affiliated mayors. Finally, using the same econometric strategy, it is analyzed the impact of the pandemic on the performance of former President, Jair Bolsonaro, in the second round of the 2022 elections. The results indicate that the number of cases, municipal mobility, and Auxílio Emergencial (Emergency Aid) expenditures were positively correlated with Bolsonaro's votes, while deaths were negatively correlated. This result is consistent with a better performance of the then President as the lethality of the disease observed in the municipality decreased, and coherent with his own speech of downplaying the severity of the pandemic and disregarding the need for adopting social distancing measures.

The rise of fascism at the current time has been the subject of research in several areas of knowledge. In political sociology and political economy, it is necessary and urgent to revisit the works of Karl Polanyi and Francisco de Oliveira to investigate the nexus between market autonomy and contemporary fascism. This work aims to contribute to the development of new theoretical instruments for understanding this difficult nexus, in its specificity in Brazil at the end of the 20th century and beginning of the 21st century. To this end, we developed our research based on the investigation of Polanyi's works on fascism and the work produced by Oliveira in the period from 1984 to 2007. Polanyi, researching the 19th century market society, identified that in this form of organization that gave broad autonomy to the self-regulated market also is the reason for the disintegration of society at the beginning of the 20th century. In his interpretation, such disruption was the reason for the rise of fascism in Europe. Oliveira, observing the transformations in Brazil from the 1980s to 2000s, recognized a new movement towards market autonomy and dedicated himself to understanding its effects on a peripheral society. His findings led him to a conclusion that was congruent with Polanyi's and went further: in Brazil, neoliberalism has become a form of totalitarianism.

The debate about the relevance of public expenditure for economic growth is extremely relevant, as it directs public policies and consequently affects the lives of the population. It has always generated great controversy, because besides economic growth the subject involves the solvency of public sector and the stability of public debt, as well as other macroeconomic topics. The importance of the subject is present at all levels of government, and it gains even greater prominence at the state level in Brazil, since it is a country characterized by a high degree of fiscal decentralization, distributing great power to the subnational governments, furthermore in recent years it has become necessary to create institutional instruments with the aim of putting Brazilian states back on a sustainable fiscal path. In this context, it is expected, along with a reduction in expenditures, a better allocation for public expenditure that are in fact effective in contributing to economic growth. Therefore, the aim of this study is to analyze under the framework of the model proposed by Devarajan, Swaroop and Zou (1996) which expenses are productive - contributed to long-term economic growth - and which are unproductive. I use a panel data of Brazilian states between 2013 and 2016 in which the regressors consist of the proportions of each category and function of spending and the regressand is a moving average of growth rates between t+1 and t+4. The results for the division of expenditures in current and capital expenditures confirm those found by Devarajan et al., i.e., there’s a positive relationship between the fraction of current expenditures and economic growth, and also a negative relationship with capital expenditures. For the functional division of public expenses, the results for the linear model indicate that spending on Public Security has a positive and significant effect. As an extension to the original model proposed by Devarajan et al., a possible non-linear relationship between public expenditures and economic growth was examined, as proposed by Rocha and Giuberti (2007), and the results do not found evidence that there is a non-linear relationship between the public spending and economic growth.

This dissertation aims to analyze and discuss what were the changes that occurred in the dynamics of the manufacturing industry in Brazil between the years 2010 and 2020. As a key sector of the economy, we sought to understand within the authors of Brazilian deindustrialization if Brazil fits into this process and whether it is homogeneous across sectors, and which would be the best indicators to be used. It is understood that traditional analysis indicators, such as the VTI/VPBI ratio, may be biased and not express the total movement of the sector. Therefore, in this work, an attempt was made to contribute to the debate with alternative forms to the theme of deindustrialization, contributing to the construction of indicators at sectoral levels. Thus, as a measure of performance analysis, a new industrial performance indicator relative to the sector will be used, built with a machine learning method of Principal Components Analysis (PCA), verifying the changes in the productive structure by the different levels of technological intensity and identifying whether there is a relative loss of any specific sector within the manufacturing industry. In addition to this tool, the instrument of the input-output Matrix product model (MIP) will be read, allowing a more complete analysis of the dynamics of the manufacturing industry from the creation of the Rasmussen-Hirschman indices of forward and backward chaining and the import coefficients of tradable inputs and final demand.

This study investigates, through a Structural Decomposition of Employment and Total Output, with estimated and disaggregated in 67 sectors Input-Output Matrices for the brazilian economy, the process of tertiarization in the 2010s. The study concludes that there are signs which point out to the the fact that jobs created in the tertiary sector are linked to low productivity of labour. Furthermore, the study concludes that "Informational Services" gained importance in the productive structure.

This study examines the continuity of dishonesty between the university academic environment and that conducive to corruption. The main hypothesis to be tested is the greater propensity for corruption of academically dishonest individuals. The database has 1213 responses, which sought to measure “corruptibility” and academic dishonesty through the presentation of scenarios and statements, and the use of an agreement scale in relation to the latter. Using linear and non-linear models, the results obtained point to a positive, direct and statistically significant relationship between academic dishonesty and the level of “corruptibility” of the individual and his probability of performing a corrupt act.

This paper analyses the effect of the adoption of the Fiscal Administration Modernization on brazilian states (PROFISCO I), financed by the Inter-American Development Bank (IDB), on the Tax on the Circulation of Goods and Services (ICMS) collection. Implemented in stages in the federative units since 2009, PROFISCO I reached 23 of the 27 federal unities, promoting the scanning of tax documents and improving the fiscal data processing and analysis by tax administrations. Using an adapted version of the Synthetic Difference-in-Differences (SDID) estimator (ARKHANGELSKY et al., 2021), the impacts of each adoption of the program were estimated, as well as the way in which these impacts were distributed over time. The tests were performed considering monthly and yearly data aggregation. In general, the estimation results did not have a specific pattern. In addition, very few of these results had statistical significance. The exception is the insertion of exogenous variables in the SDID models in the annual frequency data, a situation in which the effects are positive and significant. It is not evident, therefore, that adoption of the program does necessarily imply an improvement in collection at any time.

The debate on the best way to conduct monetary policy is of great relevance in Latin America, where high inflation rates have been a recurring problem in several countries. This article aims to understand the main views of this debate, with an emphasis on the role that the Central Bank must fulfill according to different economic schools of thought. Initially, we will provide a brief review of the origins and key points of the rules versus discretion debate in monetary policy. Subsequently, we will succinctly present the different economic theories of inflation. Furthermore, we will analyze the role played by monetary autorities in building credibility, discussing fixed exchange rate regimes and escape clauses, as well as the concept of Central Bank independence, inflation targeting regimes, and the importance of transparency and effective communication. Finally, we will analyze the performance of Argentina and Brazil in relation to the different monetary policies adopted between 1990 and 2018. This study will allow us to evaluate how monetary policy choices have impacted inflation control and the economic outcomes of these countries over time. In doing so, we aim to contribute to a deeper understanding of the monetary policy debate in Latin America, providing insights on best practices and lessons learned from past experiences.

In the last two decades, business Dynamics, entrepreneurship, and fluidity in the labor market have substantially decreased in developed countries. This phenomenon has profound consequences for productivity growth, wage growth, and job creation. However, there is a lack of studies analyzing these factors for developing countries. Using combined data from companies and employees provided by RAIS (Annual Social Information Report), this article explores the evolution of business dynamism in Brazil. The results show that, similar to developed countries, Brazil has also seen a decline in business dynamism, with a reduction in the creation of new companies, participation and employment share of young companies, and job reallocation. On the other hand, unlike what is observed in developed countries, there is a decrease in labor market concentration.

The order of transactions within blocks in Ethereum affect the results of smart contracts such as decentralized exchanges, thereby providing opportunities for malicious actors to profit at the users’ expense. These attacks contribute to increased costs, reduced legitimate transaction throughput, and potentially threaten consensus-layer security. For those reasons, numerous strategies have been proposed to counteract them, including on-chain commit & reveal mechanisms. These approaches separate transaction submission and finalization into distinct blocks while concealing crucial transaction details during the initial submission, making attacks more difficult. Expanding upon previous agent-based studies of frontrunning in decentralized finance, this dissertation aims to quantify the impact of delays inherent in commit & reveal mechanisms on price accuracy and the potential profit loss for traders due to reordering attacks that remain feasible despite these countermeasures.

This study analyzes the role of the State and its responsibility on the issue of food insecurity, both in regular periods and in critical situations. Thus, the general objective of this thesis is to understand how food insecurity policies are conditioned by State policies and regulation in Latin American countries. We will especially analyze the Brazilian case, advancing the analysis to the period of the pandemic. For this purpose, this case study will delimit, within the macroeconomic field, how the main lines of thought understand the role of the State and the market; how employment and income are determined in an Economy and how the problem of food insecurity is faced.

The thesis deals with Social Organizations (SOs) in the provision of public services and aims to make, in the light of the theory of games, an adequate design for the efficient provision of quality services by SOs, analyzing the existing incentives in this partnership model in Brazil. Considering the relevance of the role of SOs in the provision of public services in the country, and that, more than twenty years after the institution of the OSs model, in the provision of public services in the country, there is no crystallized pattern of implementation of this form of partnership between the State and non-profit organizations, it is necessary to undertake a review of the current incentives included in this relationship in favor of a model that favors the quality of services and the achievement of results. Thus, the present study proposes to analyze the existing incentives in this service provision system and to develop and present an incentive design that describes a successful partnership in the achievement of public services, considering the efficiency of expenditure and the quality of services. For this purpose, partnerships via state and municipal OSs were analyzed, according to fortunate and unfortunate experiences and, based on the information learned, a model was built based on the Theory of Information and Incentives with a view to theoretically specifying the strengths and vulnerabilities present in this format of partnership. A statistical study was also carried out, using the Newcomb-Benford Law, in order to test the reasonableness of the theoretical model developed. As a result of the work, regarding the possibilities of changes in the design of the mechanism that regulates this relationship, with a view to better taking advantage of the advantages and reducing the vulnerabilities of the partnership model via OSs, the following strategies stand out: changes that promote contracting with altruistic organizations; redefinition of granting subsidies and exemptions for non-profit institutions, linking them to the results delivered by the organization; institution of incentives to obtain resources via donations; changes that promote contracting with organizations that have expertise in carrying out the activity that is the object of the partnership; institution of awards granted by the government or by private institutions in partnership with the government, in recognition of the provision of excellent services; improvement of transparency rules; institution of an official and public website to record complaints of corruption and poor service provision, as well as to record praise for the work of Social Organizations that partner the State; constitution of central groups in governments, within the scope of each federal entity, specific to manage the State's partner OSs, which act as disseminators of good practices and ideas capable of generating social benefit; and inclusion of a device in the management contract that provides for a form of recognition, compensation or reimbursement of expenses incurred by OS, with own resources applied in activities directed to the object of the contract. 

Brazilian fiscal sustainability has been questioned mainly since 2013 and after several recent shocks, such as the 2014-2016 recession and COVID-19 in 2020. This is despite the fiscal framework in force in that period, including the spending cap. Fiscal sustainability can affect economy’s growth and stability, mainly due to economic agents’ conventions and the currencies’ hierarchy. Therefore, the first two chapters seek to verify the sustainability of Brazil’s public debt. In the first chapter, the debt dynamics of Brazil until 2040 is made based on its conditioning factors, including outlining scenarios based on different hypotheses. The components of debt variation “r-g” and the primary result are quite relevant in this dynamic, but the stock-flow adjustments have proved to be crucial in the last 15 years, often being the first or second most important determinant of debt variation. It cannot be concluded that debt is on an explosive trajectory, but, in more pessimistic scenarios, this could happen. In the second chapter, after reviewing the literature on different methodologies for analyzing debt sustainability, three of them were used. The first is through the stationarity analysis of the public debt. The second is through a cointegration analysis between revenues and expenses to verify whether these series move together. The third is based on estimating the Brazilian fiscal reaction function. The debt measures showed stationarity. The revenue and expenditure series were not shown to be cointegrated. The fiscal reaction function showed a significant response of the primary result to debt increases. According to these approaches, except for the second which has certain limitations, the public debt appears to be sustainable from a general perspective. However, the situation is not comfortable, also because Brazil's debt is higher than that of other emerging countries. Debt could be reduced with an increase in economic growth, with an increase in revenues, with a reduction in expenses, or with a combination of these options, and the reduction of interest rates can help in the process. In this regard, a robust fiscal framework could favor economic agents' expectations about fiscal sustainability, which would tend to reduce interest rates. Furthermore, this framework can be more growth-friendly, also benefiting debt sustainability. In the third chapter, an analysis of the literature on fiscal rules is carried out, investigating international practice. It appears that the fiscal framework in Brazil has inconsistencies. The primary result rule often produces a pro-cyclical fiscal policy that penalizes more qualified spending, harming economic growth and fiscal sustainability itself. The spending ceiling proves to be unfeasible, unless the role of the State in accordance with the 1988 Constitution is modified. Furthermore, the ceiling induces practices such as tax waivers to encourage the economy, postponement of expenses and increasingly out of cap expenses. The fiscal framework is also ineffective when we estimate the country's fiscal reaction function considering a variable on the strength of fiscal rules in Brazil. The new framework proposed in 2023 by the government makes progress, but has some problems. The adoption of a cycle-adjusted primary result rule is suggested, with the abandonment of the golden rule and the expenditure cap. The rule would resolve the pro-cyclical character of fiscal policy and would take into account the commodity cycles that affect the Brazilian economy. Another relevant point in the discussion of fiscal sustainability is the composition of fiscal adjustments and their consequences on economic growth. Fiscal multipliers could form part of a strategy to optimize fiscal policy, favoring growth and, therefore, fiscal sustainability itself. Thus, the fourth chapter, after reviewing the literature on fiscal multipliers in Brazil and in the world, seeks to estimate the multiplier of public investment by the Central Government for the period between 2008 and 2022. The method used is local projections, which has a series of advantages compared to other alternatives, and has no application for investments in Brazil. The results point to high multipliers, particularly in the first 10 to 18 months after the public investment shock. This implies the need to preserve and increase public investment in the country, which can have positive effects on economic growth and on the sustainable trajectory of debt as a proportion of GDP in the country.

The present Thesis is divided into four chapters, each of which are independent, but have common objectives; they are: to analyze how the Sraffian Supermultiplier (SSM) was formed and how it behaves considering exports as an alternative proposal to autonomous consumption and including the labor market. The first chapter seeks to present the historical context of the formation of the SSM and how the national and international debate on the subject has been treated. For this, it was exposed that the original theory was based on the work of Sraffa and Garegnani; however, it was only formalized in the mid-1990s. The model published at the time proposes that economic growth must necessarily consider that the autonomous consumption of capitalists is responsible for determining the progress of economies. However, this proposal did not satisfy all critical thinkers of post-Keynesianism who, from Lavoie's work in 2015, have generated an extensive debate. The arguments and replies aim to verify if, in fact, the theory is efficient in considering this autonomous component as a guarantor of sustainable growth and the convergence of the utilization capacity to its normal level. The second chapter proposes a new approach considering two different autonomous components, exports and consumption, generating three different cases: (I) when the growth of exports (𝑔𝑋) is equal to the growth of autonomous consumption (𝑔𝑋); (II) 𝑔𝑋 > 𝑔𝑍; and (III) 𝑔𝑋 < 𝑔𝑍. Through this new model, the stability conditions and the existence of a Hopf bifurcation were analyzed for the three cases; however, only in the last two is it possible to guarantee stability and bifurcation. Furthermore, the robustness of the model proposed in this chapter is proved through numerical simulations and, with the use of the computational approach, the results of the endogenous cycle can be guaranteed. The third chapter seeks to contribute to the literature by proposing that labor supply, in the long-run, is not necessarily perfectly elastic, so that, in a special case, it may be perfectly inelastic. This hypothesis generates a significant difference with the traditional postKeynesian approach that, primarily, does not consider that the labor supply can be limited to values lower than the full capacity of using the factors in the long term. With this, it is possible to verify how the behavior occurs on both sides, productivity and demand, under the hypothesis of restricted work. As a result, this modification showed that, in its original form, SSM cannot sustain growth without generating excess demand and increasing labor shortages. The fourth chapter makes the labor market hypothesis in the original version of the model more flexible, showing how the elasticity of labor supply (and thus, not considering any of the extreme cases above), affects productivity and aggregate demand, imposing whatever dynamics of labor supply is restrictive to the model. Thus, as can be seen, vii despite being independent, all four chapters deal with the same theme and have a common objective.

The present thesis is divided into three chapters that have the common objective of studying the evolution of productivity and markups over time for selected industries, as well as discussing the procedures for the estimation of production functions. The initial chapter provides a literature review regarding the methodologies for estimating production functions and markups at the firm level. The second chapter investigates the evolution of productivity and markups of US passenger airlines from 1995 to 2019. The results indicate that the entry of Low-Cost carriers initially increased the level of competition in the sector, resulting in a reduction in markups during the period 1996-2008. The increase in fuel prices during this period also contributed to this reduction in markup, as airlines faced difficulties in passing cost increases to ticket prices. However, the sector later underwent several mergers and acquisitions, which apparently reduced the level of competition and led to an increase in markups. Lastly, the third chapter addresses the Brazilian steel industry during the period 1990-1998, a period marked by intense sector restructuring. During this period, noteworthy events include the privatization of major steel companies, administrative rationalization, subsequent corporate consolidation, and the movement towards trade liberalization in the first half of the decade. Despite these transformations, there was no significant improvement in productive efficiency in the Brazilian steel industry during this period. On the other hand, there was a substantial reduction in the dispersion of both productivity and markups, indicating an improvement in allocative efficiency within the sector. Thus, despite the significant increase in market concentration during the period, market power decreased.

In this thesis, we try to understand how inequalities manifest themselves in the territory in terms of economic complexity. In addition to the focus on regional productive structure, we seek to understand to what extent economic complexity and its spatial spill-overs influence GDP per capita growth and socio-economic development in Brazilian microregions. The research is conducted using the theoretical framework of structuralism and elements from the literature on economic complexity, emphasizing points of convergence that contribute to understanding regional development. By using formal labor market data, the Economic Complexity Index (ECI) proposed by Hidalgo and Hausmann (2009) is adapted to better capture the internal dynamics of the Brazilian economy. Thus, Regional Economic Complexity Indices (ECI-r) are proposed and calculated, and they are used to measure the productive structure of subnational entities. The temporal scope encompasses the years 2007 to 2020, and the empirical investigations cover Brazilian states and microregions. In addition to computing ECI-r, we innovate in this thesis innovates by testing the empirical validity of this indicator for Brazilian microregions. Furthermore, it contributes to the literature on economic complexity by considering regional and spatial dynamics through empirical analysis using spatial econometric panel data models. The results indicate that economic complexity in Brazilian states and microregions does not change significantly in the period between 2007 and 2020, supporting the understanding that structural changes occur in the long term. On the other hand, the geographical dimension points to the existence of inequalities between Brazilian states and microregions, which are generally not mitigated by gains in economic complexity. Spatial analysis suggests evidence of positive spatial autocorrelation for economic complexity, meaning that the dynamics of ECI-r in a specific microregion should not be understood in isolation. Finally, spatial econometric analysis shows a positive association between economic complexity and GDP per capita growth and socio-economic development indicators.

There is great interest in improving inflation forecasts for better planning and decision making by households, the private sector, and policy makers. However, even outperforming univariate models can be a difficult task. We use machine learning methods and a large data set to forecast industrial goods inflation on Brazilian IPCA for horizons up to t+12, considering the time span between January 2007 and August 2021. We assess the forecasts of four regularized linear methods and two nonlinear tree based methods, with random walk and AR models as benchmarks, in a pseudo out-of-sample framework. We also assess the results without unemployment data as regressors, considering the discussions around the relevance of unemployment data on inflation forecasting. The nonlinear methods outperform the regularized linear methods and the benchmarks. We also find evidence that the variable selection mechanisms of random forest and gradient tree boosting perform better than on linear regularized models to forecast industrial goods inflation. Random forest stands out in terms of forecasting error and as the method that better controls the bias-variance trade-off. It also displays a more uniform performance than gradient tree boosting across the forecasting horizons.

This work analyzes the link between wage expenditure among Premier League football clubs and their performance in the league. The main purpose is to check if higher wages imply on better performance on the pitch and to quantify this impact. I test three different econometric specifications for the two performance metrics chosen (team’s ranking and points percentage). The regarded specifications are Fixed Effects, System GMM and Exogenous Instrumental Variable System GMM. In the end I am able to observe a positive relation between wages and performance and also check an interesting behavior for the roster variable. The models adopted in this research present a smaller group of instruments in comparison to ”Aggregate wages of players and performance in Italian Serie A”, which is the academic paper that motivated this research. The smaller group of instruments allow me to have a better control over instrument proliferation and overfitting.

Mining activities have been present in Brazil since the colonization period. It has influenced social and economic transformations. Indigenous people have inhabited the Amazon and other regions since before Brazil's discovery. Their territories are known to be mineral sources and have inevitably attracted miners. Does indigenous land tenure impact mining? In this paper, we follow the Institutional and Organizational Analysis framework to build our hypothesis, and use the PPTAL program to run a propensity score matching and estimate the impact of homologation on mining. We find that the indigenous territories that were treated (homologated) have fewer mining requests compared to the control (not-homologated) group after the program ended

The benefits of trade liberalization have been shown in economic literature for decades. Nonetheless, there is a substantial lack of data, especially on a firm level, to support a deeper analysis on the impacts of these trade policies, but the development of production function methods that rely on fewer data have made this analysis possible. Therefore, studies have show up some unexpected results of trade liberalization processes on firms. This study replicates the methodology used to analyze indian trade liberalization, making use of the estimation of production functions to estimate the behavior of markups on Brazilian steel industries during the trade liberalization that happened over the 90s. The Brazilian steel industry undergone an aggressive trade reform and privatization, and yet no changes on relative prices can be seen. Also, some report on the period have show a rise on profits and the privatization data show that the sector suffered a concentration, caused by the privatization process where some companies bought many different industries. The production function and markups for the sector were estimated for the years between 1990 and 1998. Despite the alert point mentioned, the estimated markup have been show to decrease on the period, indicating that the trade reform had results as expected from the economic literature.

: In 2015 Brazil formalized the commitment among the members of the United Nations to universalize access to water and sewage by 2030 and, for that, it will have partial management assistance from the private sector. Through the National Sanitation Information System (SNIS) in the period 2006 to 2019, we will examine the incentives for municipalities to switch to private management. Using Synthetic Differences in Differences estimation, we assess sanitation quality indicators to measure the difference between private and public management. The results indicate an increase in providers' productivity and investments after the privatization processes; we did not find a statistically significant change in other indicators.

ABSTRACT

The Economics of Restoration is an interdisciplinary field that aims to integrate concepts and tools from the fields of ecologic and environmental economics and restoration ecology. Studies that focus on how to make better use of economic concepts and practices in restoration programs and projects are in great demand given a large number of areas in need of restoration, and the scarcity of resources. The contribution of this thesis is to investigate the advances and understand the challenges faced by the Economics of Restoration field. Five chapters, written as individual papers, were developed to conduct the investigation. A bibliometric analysis opens the thesis by mapping the literature and identifying the main topics of research, what are the basis of the field and the challenges to date. From the bibliometric analysis, we were able to identify the state of the art in how costs and benefits of restoration are being measured, incorporated, and communicated in the literature. This is presented in chapter 2. In chapter 3 we turn to the main topics that need to be further addressed to strengthen the decision-making processes when it comes to restoration actions. And, in the final two chapters, we present two case studies to investigate how economic and ecologic objectives can be integrated into restoration projects to account for different aspects of restoration areas. Of special interest was if economic activities could be used to fund restoration practices to enhance ecological resilience. In the thirty years of development of the Economics of Restoration field, much progress has been made regarding tools and instruments of restoration. The current main challenge is to improve communication and sharing of information between policymakers, society, researchers, and other stakeholders involved in restoration. The benefits of restoration are not yet fully shared with or understood by society, especially by the local communities. Regarding projects and programs, balancing ecological and economic outcomes is a challenge that needs to be addressed early in the planning of restoration. The economic viability of projects is tenuous given the complexity and specificity of restoration, and there is little room to maximize only ecological outcomes since economic activities must be incorporated into projects to balance the costs of restoration. This thesis shows how modelling the priorities of interest is crucial and we believe that this work is a useful tool and information source for policymakers and researchers in the field.

In this thesis, I provide a critical assessment of the current hegemonic order, highlights its contradictions and structural flaws, and most importantly opens the door to a debate analyzing authoritarian neoliberalism, one of the most disturbing contemporary developments. All this from the perspective of political economy without separating the current issues from their historical roots. The literature on authoritarian neoliberalism has increased recently, but it often tends to focus on a specific case and pay limited attention to authoritarian connections to the neoliberal path as a whole. For this reason, I sought to build a deep discussion that addresses the recent shift towards authoritarian neoliberalism without neglecting the connection of this shift with the historical course of neoliberalism. Thus, the thesis follows a chronological path that starts from a point in the past and leads us towards the recent crisis that struck neoliberalism with the outbreak of the Covid-19 pandemic. Although authoritarian neoliberalism is a current issue crystallized in the wake of the global financial crisis, I argue that its explanation depends on our awareness of past neoliberal transformations and the contradictions inherent in the core of today's dominant thought. Taking this into account allows emphasizing the critical role of the state in building and consolidating the neoliberal order, thus avoiding the fall into the mistake of inaccurate reductionisms that show neoliberalism as an anti-state ideology. I was keen to include experiences from different countries and reviewed the various manifestations of authoritarian liberalism revealed by the experiences of countries such as the United States, Turkey, Hungary and others. It also provided an assessment of the Brazilian experience during the COVID-19 pandemic, thus combining his critical discussion with contemporary analysis of a practical case that has recently sparked interest.

Due to the deficit in social security system in many economies, structural reforms in the current pay-as-you-go system are commonly discussed and the establishment of a fully funded system is a central topic. In this case, it is necessary to analyse if both systems are seen as substitutes ou complements for each other. In the case the systems are seem as substitutes, then there is an oppotunity for future pension reforms to implement a systems’ transition. Even more, in the case a fully funded system is stablished, a return rate needs to be set to capitalize the contributions, and the eventual surplus of the capitalized social security needs to be directed. Therefore, assuming that the national government is the only one who can set the return rate of the fully funded system, that the eventual surplus generated in this system should be distributed to the pay-as-you-go system, that the government’s goal is to maximize the social welfare and the newly workers will be able to choose to which system to contribute, we propose a model to calculate the optimal return rate that need to be settled by government and we estimated the proportion of the population that will choose to contribute to each social security system.