The U.S. social security system faces funding pressure due to the aging of the population. This dissertation examines the welfare cost of social security reform and social security policy uncertainty under rational expectations and under learning. I provide an overview of the U.S. social security system in Chapter I. In Chapter II, I construct an analytically tractable two-period OLG model with capital, social security, and endogenous government debt. I demonstrate the existence of steady states depends on social security parameters. I demonstrate a saddle-node bifurcation of steady states numerically, and demonstrate a transcritical bifurcation analytically. I show that if a proposed social security reform is large enough, or if the probability of reform is high enough, the economy will converge to a steady state. In Chapter III, I develop a three-period lifecycle model. The model is inherently forward looking, which allows for more interesting policy analysis. With three periods, the young worker's saving-consumption decision depends on her expectation of future capital. This forward looking allows analysis of multi-period uncertainty. Analysis in the three-period model suggests that policy uncertainty may have lasting consequences, even after reform is enacted. In Chapter IV, I develop two theories of bounded rationality called life-cycle horizon learning and finite horizon life-cycle learning. In both models, agents use adaptive expectations to forecast future aggregates, such as wages and interest rates. This adaptive learning feature introduces cyclical dynamics along a transition path, which magnify the welfare cost of changes in policy and policy uncertainty. I model policy uncertainty as a stochastic process in which reform takes place in one of two periods as either a benefit cut or a tax increase. I find the welfare cost of this policy uncertainty is less than 0.25% of period consumption in a standard, rational expectations framework. The welfare cost of policy uncertainty is larger in the learning models; the worst-off cohort in the life-cycle horizon learning model would be willing to give up 1.98% of period consumption to avoid policy uncertainty.

xiii, 230 p. : ill. (some col.) / I develop a model in which a representative consumer selects an affordable consumption bundle, not as a single choice, but as the end result of a series of smaller, incremental purchase decisions. If the array of such incremental choices facing the consumer is sufficiently complex relative to the consumer's computational abilities, then the consumer may choose to employ a simplifying heuristic or rule-of-thumb to guide her behavior. I demonstrate the existence of a simple and well-defined example of such a strategy, based upon a satisficing decision rule. I further show that in the strategic setting defined by the interaction between consumers and firms that compete in prices, this satisficing strategy can form part of a Nash equilibrium, despite being ex ante only boundedly rational. The use of this satisficing demand strategy fundamentally alters the nature of price competition between firms (relative to the standard Bertrand model), changing the shape of the firm best response functions. The use of a satisficing strategy alters the incentives of firms, and these altered firm incentives lead to pricing behavior which has the effect of rationalizing the satisficing consumption strategy, so that a truly novel class of Nash equilibria in price-competing markets can be shown to exist under certain conditions. We explore the nature of this new class of equilibria, and find that equilibrium prices may be higher than those which would be obtained in the standard Bertrand case. In general, demand curves for each distinct good will have a kinked shape, similar to those found in 1939 papers by both Sweezy and Hall & Hitch. The Nash equilibrium profile will involve the kink in each demand curve coinciding with the equilibrium price for the corresponding good. The equilibrium price vector will therefore be robust to "small" fluctuations in cost (since marginal revenue is discontinuous at the equilibrium price), and under certain conditions, we find that prices may be upwardly flexible but downwardly rigid. We make an argument that the main results of the paper generalize from a representative agent setting to one with a population of heterogeneous consumers. / Committee in charge: Dr. Van Kolpin， Chairperson； Dr. Christopher J. Ellis， Member； Dr. Jeremy Piger， Member； Dr. Renee Irvin， Outside Member

This dissertation studies the impact embedding boundedly rational agents in real business cycle-type news-shock models may have on a variety of model predictions, from simulated moments to structural parameter estimates. In particular, I analyze the qualitative and quantitative effects of assuming agents are boundedly rational in a class of DSGE models which attempt to explain the observed volatility and comovements in key aggregate measures of U.S. economic performance as the result of endogenous responses to information in the form of ``news shocks''. The first chapter explores the theoretical feasibility of relaxing the rational expectations hypothesis in a three-sector real business cycle (RBC) model which generates boom-bust cycles as a result of periods of optimism and pessimism on the part of households. The second chapter determines whether agents forming linear forecasts of shadow prices in a nonlinear framework can lead to behavior approximately consistent with fully informed individuals in a one-sector real business cycle model. The third chapter analyzes whether empirical estimates of the relative importance of anticipated shocks may be biased by assuming rational expectations. By merging the two hitherto separate but complementary strands of literature related to bounded rationality and news shocks I am able to conduct in-depth analysis of the importance of both the information agents have and what they choose to do with it. At its core, the study of news in macroeconomics is a study of the specific role alternative information sets play in generating macroeconomic volatility. Adaptive learning on the other hand is concerned with the behavior of agents given an information set. Taken together, these fields jointly describe the input and the ``black box'' which produce model predictions from DSGE models. While previous research has been conducted on the effects of bounded rationality or news shocks in isolation, this dissertation marks the first set of research explicitly focused on the interaction of these two model features.

O mercado de capitais brasileiro tem se desenvolvido ao longo dos anos, e com o fim do longo período inflacionário, houve a possibilidade das pessoas fazerem planejamentos de longo prazo, sem se preocupar apenas com a perda do valor do dinheiro no curto prazo. Alguns fatores levaram à entrada de investidores no mercado de capitais, que tem sido crescente nos últimos anos. Para que se atraia cada vez mais investidores para esse mercado, e de forma sustentável, instruindo-os para que possam ter mais consciência na hora de investir, é essencial conhecer vieses que influenciam suas decisões, pois, diferentemente do que apontam as Teorias Tradicionais e Modernas de Finanças, os investidores (e as pessoas em geral) não agem de forma completamente racional quando fazem escolhas, podendo ser influenciados, de forma mais ou menos intensa, por vieses, como excesso de confiança, falácia de custos irrecuperáveis, aversão à perda, entre outros, que poderão afetar essas escolhas, e por fim, o mercado em geral. Dessa forma, o objetivo deste estudo foi de identificar a intensidade em que vieses estão presentes em decisões de investidores individuais do mercado de capitais brasileiro, e verificar se essa intensidade está relacionada ao sexo e ao tempo como investidor do mercado, fornecendo subsídio para que sejam desenvolvidos programas, focados inicialmente naqueles vieses que se mostram mais presentes nas decisões, para instruir investidores e possíveis investidores sobre essa influência, ajudando-os a identificar padrões em suas escolhas que possam ser prejudiciais a eles. Para isso, o instrumento de pesquisa utilizado foi um questionário com questões múltipla escolha, disponibilizado em provedor de serviços de pesquisas eletrônicas, por meio do qual foi efetuada a coleta de dados. O link do questionário foi enviado a instituições relacionadas ao mercado de capitais para divulgação a investidores e a grupos de investidores por meio de redes sociais. No total, 178 pessoas responderam à pesquisa, sendo que 80 são investidores no mercado de capitais, cujas respostas foram analisadas. Para efetuar a interpretação das respostas foi utilizada análise descritiva. Observou-se que, dos 13 vieses analisados, apenas 4 se mostraram com alta intensidade na escolha de investidores, sendo eles os vieses de excesso de confiança, excesso de negociação, contabilização mental e ancoragem, e que, para maioria dos vieses não se observa diferença significativa de intensidade entre sexo masculino e feminino, mas é possível perceber que para alguns dos vieses quanto maior o tempo como investidor, menor a intensidade do viés. As respostas de não investidores também foram analisadas, como forma de identificar a intensidade em que vieses estariam presentes em pessoas que poderiam em algum momento ser investidores, e percebeu-se que, comparativamente aos investidores, eles apresentaram maior intensidade dos vieses. Para pessoas que estão envolvidas com o mercado de capitais a intensidade dos vieses não foi tão alta, mas para aqueles que não são investidores a alta intensidade foi predominante para um maior número de vieses, o que poderia estar relacionado à experiência adquirida no mercado de alguma forma, e que mostra a necessidade de apresentar situações a que as pessoas poderiam estar expostas e cuidados a serem tomados para mitigar as influências que podem sofrer ao investir no mercado. / The Brazilian capital market has developed over the years, and with the end of a long inflationary period, there was the possibility of people making long-term plans, instead of only being worried about the loss of value of money in the short term. Some factors have led to the entry of investors in the capital market, which has been growing in recent years. In order to attract more and more investors to this market, and in a sustainable way, instructing them so they can be more aware when investing, it is essential to know the biases that influence their decisions, since, unlike what the Traditional and Modern Finance Theories describe, investors (and people in general) do not act completely rationally when they make choices and may be influenced more or less intensely by biases such as overconfidence, sunk cost fallacy, loss aversion, among others, which may affect these choices, and ultimately, the market. Thus, the aim of this study was to identify the intensity in which biases are present in the decisions of individual investors in the Brazilian capital market, and verify if this intensity is related to sex and time as market investor, providing information so that programs can be developed, focused initially on those biases that are more present in decisions, to instruct investors and potential investors of this influence, helping them to identify patterns in their choices that may be harmful to them. For this, the research instrument was a questionnaire with multiple choice questions, available in electronic research services provider, through which was collected the data. The link to the questionnaire was sent to the capital market related institutions, so they could send it to investors, and groups of investors through social networks. 178 people responded to the survey, of which 80 are investors in the capital market, whose responses were analyzed. To analyze the responses it was used descriptive analysis. It was observed that, of the 13 biases analyzed, only 4 showed up with high intensity in the choice of investors, namely the bias of overconfidence, excessive trading, mental accounting and anchor, and that for, most biases, no significant difference in intensity is observed between males and females, however, for some biases, it is possible to see that when higher the period the person has been an investor, lower the intensity of the biases. The responses of non-investors were also analyzed as a way to identify the intensity in which biases were present in people who might be, at some point, investors, and it was noticed that, compared to investors, they showed greater intensity of biases. For those people who are involved with the capital market the intensity of biases was not as high as for those who are not, for whom the high intensity was prevalent for a larger number of biases. This could be related, somehow, to the experience gained in the market, and shows the need to present situations that people could be exposed and be careful about, being aware of steps that could be taken to mitigate the influences that they can suffer when investing in the market.

La prémisse de la rationalité décisionnelle des agents, généralement définie comme la « prise de décision optimale sous l’information disponible », occupe une place centrale dans la théorie économique, à tel point que, lorsque la rationalité décisionnelle passe à la loupe, c’est souvent l’ensemble de la science économique qui l’est à sa suite. Pourtant, le quotidien abonde de cas ne se conformant pas à une certaine définition de l’optimalité; pour des décisions de faibles enjeux, les agents économiques pourront couper court à leur réflexion et se contenter d’une décision potentiellement sous-optimale afin de minimiser le coût de la prise de décision. Or, avec l’ubiquité d’Internet, l’information et la puissance de calcul sont plus disponibles que jamais, à tel point où le tri de l’information dans la surabondance d’Internet constitue une nouvelle forme de coût de décision. Ce mémoire s’intéresse donc à deux questions concernant la prise de décision avec échéances; la première est: à quels types de décisions, en fonction de la taille de leur enjeu, les agents économiques consacreront-ils le plus de temps de réflexion? La seconde est: à quels types d’informations les agents se fient-ils le plus? Essentiellement, nous testons si les individus optimisent l’effort qu’ils mettent à optimiser . Nous utilisons une méthodologie expérimentale basée sur un jeu-questionnaire, avec incitatifs financiers à la performance, simulant différentes décisions à caractère économique auxquelles des personnes sont appelées à être exposées au cours de leur vie. Comme résultats, nous trouvons une relation concave entre la taille de l’enjeu sur lequel porte une décision et les ressources allouées à la décision. Cette relation s’estompe lorsque les contraintes de temps se resserrent. Nous tentons également de vérifier s’il existe un lien entre et le choix de sources d’informations à des fins de décision et la familiarité de celles-ci pour l’agent, mais ne trouvons pas de relation significative. De plus, nous ne parvenons pas à établir de relation entre le temps consacré à prendre une décision et l’optimalité de celle-ci.

This paper proposes an agency model for headquarters subsidiary relationships in multinational organizations with headquarters as the principal and the subsidiary as the agent. As a departure from classical agency theory, our model is developed for the unit level of analysis and considers two root causes of the agency problem - self-interest and bounded rationality. We argue that in the organizational setting, one cannot assume absolute self-interest and perfect rationality of agents (subsidiaries) but should allow them to vary. We explain subsidiary level variation through a set of internal organizational and external social conditions in which the headquarters-subsidiary agency dyad is embedded. We then discuss several agency scenarios reflecting various levels of selfinterest and rationality that lead to different manifestations of the agency problem. The proposed framework can inform more relevant applications of the agency perspective in organizational studies and motivate future research.

A questão se a informação contábil é relevante para o mercado de capitais tem sido investigada, predominantemente, por meio dos pressupostos da Hipótese de Eficiência de Mercado (HEM). Para a HEM, toda informação relevante é refletida nos preços das ações de forma integral e instantânea, a partir da consideração de que as informações são analisadas e interpretadas por indivíduos plenamente racionais. Contudo, a literatura relacionada às áreas de Finanças Comportamentais e de Processos Decisórios tem indicado que os indivíduos, mesmo em condições de interação e de competição, como verificado nos mercados financeiros, são melhor caracterizados como detentores de racionalidade limitada ao tomar decisões. Nesse sentido, o objetivo deste estudo foi examinar a relevância da informação contábil para o mercado de capitais brasileiro sob o pressuposto da racionalidade limitada dos investidores. Para tanto, foram desenvolvidas escalas de complexidade específicas para as ações ordinárias e preferenciais. As escalas foram utilizadas como parâmetros para testar se níveis distintos de incertezas na estimação dos fluxos de caixa futuros estão associados à utilidade da informação contábil para o mercado de capitais. Além disso, no estudo, segregou-se a tomada de decisão nas dimensões dos ganhos e das perdas, tendo como objetivo identificar a relevância da informação contábil, segundo essa classificação. A amostra foi composta por informações de 232 empresas listadas na Bolsa de Valores, Mercadorias e Futuros de São Paulo (BM&FBOVESPA) no período de 2000 a 2015. Os resultados encontrados apontaram evidências de uma associação inversa entre os níveis de complexidade na avaliação das empresas e a relevância da informação contábil para os investidores. Foi identificado, também, que os preços das ações tendem a incorporar as informações contábeis relevantes de forma apenas gradual em condições de maiores níveis de incertezas. Esses resultados mostraram-se robustos para a dimensão dos ganhos. Além disso, os resultados obtidos sugerem que as normas contábeis do International Financial Reporting Standard (IFRS) reduziram os níveis de complexidade na avaliação das ações, o que resultou em um aumento da relevância da informação contábil para os investidores. De forma geral, as evidências obtidas corroboram a ideia de que os limites cognitivos dos indivíduos em processar informações pode ser um fator relacionado à magnitude com que os preços das ações refletem as informações contábeis. / The question whether accounting information is relevant to the capital market has been investigated predominantly through the assumptions of the Efficient Market Hypothesis (EMH). For EMH, all relevant information is reflected in stock prices in an integral and instantaneous way, considering that information is analyzed and interpreted by fully rational individuals. However, the literature related to the areas of Behavioral Finance and Decision Making has indicated that individuals, even in conditions of interaction and competition, as verified in financial markets, are better characterized as having limited rationality when making decisions. In this sense, the objective of this study was to examine the relevance of the accounting information to the Brazilian capital market, under the assumption of investors\' bounded rationality. Therefore, specific complexity scales were developed for common and preferred stocks. The scales were used as parameters to test if different levels of uncertainties in the estimation of future cash flows are associated with the usefulness of the accounting information for the capital market. In addition, the study segregated the decision making in gains and losses dimensions, aiming to identify the relevance of accounting information according to this classification. The sample consisted of information of 232 companies listed on the Brazilian Securities, Commodities and Futures Exchange (BM&FBOVESPA), from 2000 to 2015. The findings brought evidence of an inverse association between levels of complexity in the evaluation of the stocks and the relevance of accounting information to investors. It was identified that stock prices tend to incorporate the relevant accounting information only gradually in conditions of higher levels of uncertainties. These results were robust for the gain dimension. Furthermore, the results suggest that the accounting standards of International Financial Reporting Standard (IFRS) reduced complexity levels in stock valuation, which resulted in an increase in the relevance of accounting information for investors. In general, the evidence obtained corroborates with the idea that cognitive limits of individuals in processing information may be a factor related to the magnitude in which stock prices reflect the accounting information.

Cette thèse aborde le thème de la rationalité limitée à travers quatre chapitres, associant modèles théoriques, expériences en laboratoire et analyses statistiques et économétriques. Dans les deux premiers chapitres, nous testons la validité de nouveaux modèles en économie qui utilisent le formalisme mathématique de la mécanique quantique pour rendre compte de biais cognitifs. Au sein du chapitre 1, nous considérons des modèles expliquant l'effet d'ordre et en dérivons de nouvelles prédictions expérimentales. Dans le chapitre 2, nous proposons une expérience originale pour tester une large gamme de modèles quantiques qui rendent compte de l'erreur de conjonction. Les deux groupes de modèles échouent à nos tests empiriques. Nous discutons alors de possibles pistes d'améliorations de ces modèles. Le chapitre 3 explore la façon dont les individus traitent des informations économiques successives, complexes et abondantes. Nos résultats expérimentaux montrent l'inaptitude des sujets à combiner de telles informations, ce qui confirme la théorie de la trace floue. Enfin, le chapitre 4 relève de la finance expérimentale. Il étudie comment l'achat sur marge (respectivement la vente à découvert) augmente (diminue) le niveau des prix, la volatilité, l'hétérogénéité des marchés et les anticipations de prix des traders ainsi que la façon dont il modifie les stratégies de trading. Nos résultats mettent en évidence les nettes conséquences de chacune de ces techniques prises séparément, et identifient des phénomènes inattendus lorsqu'elles sont combinées. Nos analyses ouvrent la voie à une meilleure prise en compte de ces interactions déstabilisatrices par les autorités de régulation. / This thesis studies bounded rationality through four chapters, combining theoretical models, laboratory experiments and statistical and econometric analyzes. In the first two chapters, we test the validity of new models in economics which rely on the mathematical formalism of quantum mechanics to account for cognitive biases. In chapter 1, we consider models explaining the order effect and we derive new experimental predictions. In chapter 2, we propose an original experiment to test a wide range of quantum models that account for the conjunction fallacy. Both groups of models fail in our empirical tests and we then discuss possible ways to improve these models. The third chapter explores how individuals deal with successive, complex and abundant economic information. Our experimental results show the subjects' inability to combine such information, which confirms the fuzzy trace theory. Finally, the fourth chapter deals with experimental finance. It studies how margin buying (respectively short selling) increases (decreases) price levels, volatility, heterogeneity of markets, and traders' price expectations, as well as how it changes trading strategies. Our results highlight the clear consequences of each of these techniques used alone, and point to unexpected phenomena when both are combined. Regulatory authorities could take advantage of our analyzes to reduce the destabilization introduced by these techniques.

Economists have convulsively maintained the assumption that humans are able to arrive at decisions by perfect deductive rationality, despite the fact empirical evidences are showing otherwise. The contradicting evidences have resulted in a personal view that instead of finding a unified theory about decision-making, a sound approach would be to study how humans in fact are reasoning in specific contexts. The context of interest for this paper is where it could be assumed humans’ persistence of acting rational is determined by the perceived burden of the problem. In this work, the inductive way of arriving at decisions plays an important role, and the paper will present a way of describing this process in a consistent way. The process will be denoted as the actual level of behavioral change, and represent the core property of this paper. Applying the presented theory is most appropriate for situations where it could be assumed the burden of a problem, expressed as a prevalence rate, will drive the behavioral change. The line of reasoning in this paper will therefore be applied to the important arena of fighting the spread of HIV.

Den genomförda studien syftar till att undersöka transaktionsprocessen och identifiera var det finns störst risk för felbeslut och irrationellt agerande blad fastighetsaktörer i Sverige. Studien bidrar med att skapa en förståelse för läsaren gällande vilka risker som finns i processen samt att framställa var i processen riskerna är som störst. Materialet som ligger till grund för studien består av intervjuer och enkätundersökningar som delats ut till flertalet bolag i branschen. Intervjuerna har genomförts med personer som arbetar med fastighetstransaktioner på väletablerade bolag i branschen. Resultatet av studien indikerar på att det finns stora risker för felbeslut genom hela transaktionsprocessen vilket påverkas av flertalet faktorer i de olika delarna. För den säljande parten är risken för irrationella beslut som störst i början av processen. De förklaras av att det är den säljande parten som i början av processen exempelvis väljer vilken marknadsföringsprocess som ska genomföras, samt vem marknadsföringen ska rikta sig till. Riskerna i processen övergår därefter till köparen. De faktorer som påverkar irrationellt beslutsfattande kan förklaras av de teorier som presenteras i litteraturstudien och teoriavsnittet: ‘Bounded rationality’, ‘principal agent problem’ och ‘ankar effekter’. Det går inte att säkerställa vad som leder till att ett irrationellt beslut genomförs eftersom varje process är unik. Den här studien syftar endast till att identifiera var i transaktionsprocessen det finns risker för aktörerna att agera irrationellt. Nyckelord för studien är: ‘Transaktionsprocess’, ‘bounded rationality’, ‘principal agent problem’, ‘ankar effekt’, ‘due diligence’, ‘fastighetstransaktion’. / The following study aims to examine the transactional process for real estate transactions and establish where there is a risk for irrational decision making amongst real estate investors and advisors in Sweden. The method consists of questionnaires that have been distributed to real estate transaction consultants in Sweden. The study also used semi-structured interviews as a method. The interviews in the study were carried out with real estate consulting companies that focus on real estate transactions in Sweden. The results of the study indicate that there are risks in regards to investment decision making throughout the whole transactionprocess that are affected by various factors for the different parts of the transactionprocess. The underlying explanation is that most of the decision making is done in the initial part of the transactionprocess. The factors that cause irrational decision making can be explained by the theories presented in the literature part of the study and the theory part of the study. Those factors are: Principal agent problems, anchor effects and bounded rationality. It is not possible to ensure exactly what causes irrational behaviour in decision making but the study aims to discover and indicate where in the process there is a risk for the various factors causing irrational behaviour in decision making.