|
on Utility Models and Prospect Theory |
|
Issue of 2024‒02‒26
twenty-two papers chosen by |
| By: | Miles S. Kimball; Daniel Reck; Fudong Zhang; Fumio Ohtake; Yoshiro Tsutsui |
| Abstract: | How quickly does marginal utility fall with increasing consumption? It depends on the dimension along which we consider concavity of the utility function. This paper estimates the distribution of heterogeneous curvature parameters in individuals’ utility functions from hypothetical choice data, while accounting for survey response error. Types of curvature examined include relative risk aversion, intertemporal substitution, the reciprocal of the altruism elasticity, and a new measure of inequality aversion, which queries how much more a dollar means to a poor family than to a rich family. Median values of curvature parameters ranging from 0.6 to 13.2. Utility functions are most concave for situations involving altruism, followed by risk aversion, inequality aversion, and intertemporal substitution. Heterogeneity of curvature in the population also varies: altruism is the most heterogeneous, followed by risk aversion, the elasticity of intertemporal substitution, and inequality aversion. Nonetheless, curvature parameters are highly correlated (ρ > .8) over different elicitations within parameter type, and modestly correlated across dimensions in some cases, including inequality aversion and risk aversion (ρ ≈ 0.3), altruism and risk aversion (ρ ≈ 0.3), and altruism and inequality aversion (ρ ≈ 0.14). |
| JEL: | D01 D03 D15 D19 |
| Date: | 2024–01 |
| URL: | http://d.repec.org/n?u=RePEc:nbr:nberwo:32077&r=upt |
| By: | Joseph Y. Halpern; Evan Piermont |
| Abstract: | We show that it is possible to understand and identify a decision maker's subjective causal judgements by observing her preferences over interventions. Following Pearl [2000], we represent causality using causal models (also called structural equations models), where the world is described by a collection of variables, related by equations. We show that if a preference relation over interventions satisfies certain axioms (related to standard axioms regarding counterfactuals), then we can define (i) a causal model, (ii) a probability capturing the decision-maker's uncertainty regarding the external factors in the world and (iii) a utility on outcomes such that each intervention is associated with an expected utility and such that intervention $A$ is preferred to $B$ iff the expected utility of $A$ is greater than that of $B$. In addition, we characterize when the causal model is unique. Thus, our results allow a modeler to test the hypothesis that a decision maker's preferences are consistent with some causal model and to identify causal judgements from observed behavior. |
| Date: | 2024–01 |
| URL: | http://d.repec.org/n?u=RePEc:arx:papers:2401.10937&r=upt |
| By: | Jingyi Cao; Dongchen Li; Virginia R. Young; Bin Zou |
| Abstract: | We find the optimal indemnity to maximize the expected utility of terminal wealth of a buyer of insurance whose preferences are modeled by an exponential utility. The insurance premium is computed by a convex functional. We obtain a necessary condition for the optimal indemnity; then, because the candidate optimal indemnity is given implicitly, we use that necessary condition to develop a numerical algorithm to compute it. We prove that the numerical algorithm converges to a unique indemnity that, indeed, equals the optimal policy. We also illustrate our results with numerical examples. |
| Date: | 2024–01 |
| URL: | http://d.repec.org/n?u=RePEc:arx:papers:2401.08094&r=upt |
| By: | Avner Seror (Aix-Marseille Univ., CNRS, AMSE, Marseille, France) |
| Abstract: | This paper shows how to recover behavioral biases from revealed preference ranking implied by choices. The approach formalizes and unifies well-known behavioral models, including salience thinking, inattention, and logarithmic perception, thereby accounting for many well-documented choice puzzles. I show that this approach provides a way to filter out choice data from behavioral biases explaining rationality breaches before fitting parametric utility models. The approach is applied to workhorse data sets of the literature on choice under risk and scanner consumer choices. |
| Keywords: | Decision Theory, revealed preference, Behavioral Economics |
| JEL: | D91 D11 D81 |
| Date: | 2024–02 |
| URL: | http://d.repec.org/n?u=RePEc:aim:wpaimx:2401&r=upt |
| By: | Zongxia Liang; Sheng Wang; Jianming Xia; Fengyi Yuan |
| Abstract: | This paper addresses the continuous-time portfolio selection problem under generalized disappointment aversion (GDA). The implicit definition of the certainty equivalent within GDA preferences introduces time inconsistency to this problem. We provide the sufficient and necessary conditions for a strategy to be an equilibrium by a fully nonlinear ordinary differential equation (ODE). Through an exploration of the existence and uniqueness of solution to the ODE, we establish the existence and uniqueness of the equilibrium. Our findings indicate that under disappointment aversion (DA) preferences, non-participation in the stock market is the unique equilibrium. The numerical analysis reveals that, under GDA preferences, the investment proportion in the stock market consistently remains smaller than the investment proportion under the classical Expected Utility (EU) theory. |
| Date: | 2024–01 |
| URL: | http://d.repec.org/n?u=RePEc:arx:papers:2401.08323&r=upt |
| By: | Daniel Krähmer (Universität Bonn) |
| Abstract: | The paper studies the canonical hold-up problem with one-sided investment by the buyer and full ex post bargaining power by the seller. The buyer can covertly choose any distribution of valuations at a cost and privately observes her valuation. The main result shows that in contrast to the well-understood case with linear costs, if investment costs are strictly convex in the buyer’s valuation distribution, the buyer’s equilibrium utility is strictly positive and to tal welfare is strictly higher than in the benchmark when valuations are public information, thus alleviating the hold-up problem. In fact, when costs are mean-based or display decreasing risk, the hold-up problem may disappear completely. Moreover, the buyer’s equilibrium utility and total welfare might be non-monotone in costs. The paper utilizes an equilibrium characterization in terms of the Gateaux derivative of the cost function. |
| Keywords: | Information Design, Hold-Up Problem, Unobservable Information |
| JEL: | C61 D42 D82 |
| Date: | 2024–02 |
| URL: | http://d.repec.org/n?u=RePEc:ajk:ajkdps:278&r=upt |
| By: | Stelios Arvanitis (Athens University of Economics and Business - Department of Economics); O. Scaillet (Swiss Finance Institute - University of Geneva); Nikolas Topaloglou (Athens University of Economics and Business) |
| Abstract: | We develop and implement methods for determining whether relaxing sparsity constraints on portfolios improves the investment opportunity set for risk-averse investors. We formulate a new estimation procedure for sparse second-order stochastic spanning based on a greedy algorithm and Linear Programming. We show the optimal recovery of the sparse solution asymptotically whether spanning holds or not. From large equity datasets, we estimate the expected utility loss due to possible under-diversification, and find that there is no benefit from expanding a sparse opportunity set beyond 45 assets. The optimal sparse portfolio invests in 10 industry sectors and cuts tail risk when compared to a sparse mean-variance portfolio. On a rolling-window basis, the number of assets shrinks to 25 assets in crisis periods, while standard factor models cannot explain the performance of the sparse portfolios. |
| Keywords: | Nonparametric estimation, stochastic dominance, spanning, under-diversification, greedy algorithm, Linear Programming |
| JEL: | C13 C14 C44 C58 C61 D81 G11 |
| Date: | 2024–01 |
| URL: | http://d.repec.org/n?u=RePEc:chf:rpseri:rp2408&r=upt |
| By: | Hector Chade; Ilse Lindenlaub |
| Abstract: | How does the presence of risk sharing affect sorting patterns on productive attributes when there are complementarities among partners' skills in match output? We develop a matching model in which risk-averse agents, who differ in skills, match pairwise for productive purposes. Match output has stochastic returns and matched partners efficiently share this risk. We find that under plausible assumptions the risk-sharing benefit of marriage tends to push toward negative sorting on partners' skills. To obtain the prediction of positive skill sorting—a robust empirical feature of marriage markets—this force needs to be counteracted by sufficiently strong skill complementarities in match output. We provide a novel inequality that characterizes monotone (positive or negative) equilibrium sorting, balancing out skill complementarities and risk-sharing considerations in the right way. Several classes of primitives (utility and match output functions) render monotone sorting optimal. We then highlight a new implication of positive sorting on exogenous skills for matching patterns on endogenous differences in risk aversion: Positive sorting on skills translates into positive sorting on risk aversion—in line with the evidence from marriage markets. |
| JEL: | C78 D81 |
| Date: | 2024–01 |
| URL: | http://d.repec.org/n?u=RePEc:nbr:nberwo:32051&r=upt |
| By: | Ferreira, Susana (University of Georgia); Moro, Mirko (University of Stirling); Welsch, Heinz (University of Oldenburg) |
| Abstract: | A growing literature in economics uses subjective well-being data collected in surveys as a proxy for utility. Environmental economists have combined these data with the public goods experienced by respondents using a novel non-market valuation approach: the experienced preference approach. In this review, we take stock of what we know, including recent developments, and what we still need to learn about this new approach. We first present a conceptual framework that clarifies the relationship between experienced preference and conventional valuation approaches. We then discuss key challenges for its empirical application and identify areas where additional research would be fruitful. |
| Keywords: | subjective well-being, life-satisfaction, happiness, experienced utility, non-market valuation, willingness to pay, public goods |
| JEL: | Q51 I31 H41 |
| Date: | 2024–01 |
| URL: | http://d.repec.org/n?u=RePEc:iza:izadps:dp16718&r=upt |
| By: | Min Dai; Hanqing Jin; Xi Yang |
| Abstract: | We propose an innovative data-driven option pricing methodology that relies exclusively on the dataset of historical underlying asset prices. While the dataset is rooted in the objective world, option prices are commonly expressed as discounted expectations of their terminal payoffs in a risk-neutral world. Bridging this gap motivates us to identify a pricing kernel process, transforming option pricing into evaluating expectations in the objective world. We recover the pricing kernel by solving a utility maximization problem, and evaluate the expectations in terms of a functional optimization problem. Leveraging the deep learning technique, we design data-driven algorithms to solve both optimization problems over the dataset. Numerical experiments are presented to demonstrate the efficiency of our methodology. |
| Date: | 2024–01 |
| URL: | http://d.repec.org/n?u=RePEc:arx:papers:2401.11158&r=upt |
| By: | Roland Bénabou; Armin Falk; Luca Henkel |
| Abstract: | Choosing what is morally right can be based on the consequences (ends) resulting from the decision – the Consequentialist view – or on the conformity of the means involved with some overarching notion of duty – the Deontological view. Using a series of experiments, we investigate the overall prevalence and the consistency of consequentialist and deontological decision-making, when these two moral principles come into conflict. Our design includes a real-stakes version of the classical trolley dilemma, four novel games that induce ends-versus-means tradeoffs, and a rule-following task. These six main games are supplemented with six classical self-versus-other choice tasks, allowing us to relate consequential/deontological behavior to standard measures of prosociality. Across the six main games, we find a sizeable prevalence (20 to 44%) of nonconsequentialist choices by subjects, but no evidence of stable individual preference types across situations. In particular, trolley behavior predicts no other ends-versus-means choices. Instead, which moral principle prevails appears to be context-dependent. In contrast, we find a substantial level of consistency across self-versus-other decisions, but individuals’ degree of prosociality is unrelated to how they choose in ends-versus-means tradeoffs |
| Keywords: | morality, deontological, consequentialist, Kantian, ends-versus-means, trolley dilemma, prosocial, altruism, social preferences |
| JEL: | C91 D01 D64 |
| Date: | 2024–01 |
| URL: | http://d.repec.org/n?u=RePEc:bon:boncrc:crctr224_2024_499&r=upt |
| By: | KUMAGAI Junya; MANAGI Shunsuke |
| Abstract: | This paper critically assesses recent research on the policy application of well-being indicators and presents an overview of both their advantages and potential issues. Initially, we explore diverse methods for measuring well-being, including the Subjective Well-being (SWB) indicator, the OECD Better Life Index (BLI), the Inclusive Wealth Index (IWI) designed to sustain the well-being of future generations, and an approach that estimates the significance of each dimension through the stated preference method. We delineate the distinctive characteristics and merits of each indicator. Subsequently, drawing on recent studies suggesting a divergence between utility and SWB, we highlight methodological and normative concerns associated with utilizing SWB as an indicator of social well-being. We introduce strategies to mitigate these concerns. Additionally, through empirical analysis, we investigate the specific domains in which satisfaction measures derived from self-reported subjective responses exhibit a robust correlation with objective levels. Synthesizing insights from these reviews and data analysis, we pinpoint concerns that are inherent in well-being indicators and propose pragmatic solutions, offering actionable guidance for policymakers in the realm of well-being assessment. |
| Date: | 2024–02 |
| URL: | http://d.repec.org/n?u=RePEc:eti:rpdpjp:24001&r=upt |
| By: | Benjamin Enke; Thomas Graeber; Ryan Oprea; Thomas W. Graeber |
| Abstract: | A large literature shows that people discount financial rewards hyperbolically instead of exponentially. While discounting of money has been questioned as a measure of time preferences, it continues to be highly relevant in empirical practice and predicts a wide range of real-world behaviors, creating a need to understand what generates the hyperbolic pattern. We provide evidence that hyperbolic discounting reflects mistakes that are driven by the complexity of evaluating delayed payoffs. In particular, we document that hyperbolicity (i) is strongly associated with choice inconsistency and cognitive uncertainty, (ii) increases in overt complexity manipulations and (iii) arises nearly identically in computationally similar tasks that involve no actual payoff delays. Our results suggest that even if people had exponential discount functions, complexity-driven mistakes would cause them to make hyperbolic choices. We examine which experimental techniques to estimate present bias are (not) confounded by complexity. |
| Keywords: | hyperbolic discounting, present bias, bounded rationality, cognitive uncertainty |
| JEL: | C91 D91 G00 |
| Date: | 2023 |
| URL: | http://d.repec.org/n?u=RePEc:ces:ceswps:_10861&r=upt |
| By: | Roland Bénabou; Armin Falk; Luca Henkel |
| Abstract: | Choosing what is morally right can be based on the consequences (ends) resulting from the decision – the Consequentialist view – or on the conformity of the means involved with some overarching notion of duty – the Deontological view. Using a series of experiments, we investigate the overall prevalence and the consistency of consequentialist and deontological decision-making, when these two moral principles come into conflict. Our design includes a real-stakes version of the classical trolley dilemma, four novel games that induce ends-versus-means tradeoffs, and a rule-following task. These six main games are supplemented with six classical self-versus-other choice tasks, allowing us to relate consequential/deontological behavior to standard measures of prosociality. Across the six main games, we find a sizeable prevalence (20 to 44%) of nonconsequentialist choices by subjects, but no evidence of stable individual preference types across situations. In particular, trolley behavior predicts no other ends-versus-means choices. Instead, which moral principle prevails appears to be context-dependent. In contrast, we find a substantial level of consistency across self-versus-other decisions, but individuals’ degree of prosociality is unrelated to how they choose in ends-versus-means tradeoffs. |
| JEL: | C91 D01 D64 |
| Date: | 2024–01 |
| URL: | http://d.repec.org/n?u=RePEc:nbr:nberwo:32073&r=upt |
| By: | Thomas Epper (IESEG School of Management, Univ. Lille, CNRS, UMR 9221 - LEM - Lille Economie Management F-59000 Lille, France); Julien Senn (Department of Economics, Zurich University. Blümlisalpstrasse 10, 8006 Zurich, Switzerland); Ernst Fehr (Department of Economics, Zurich University. Blüumlisalpstrasse 10, 8006 Zurich, Switzerland) |
| Abstract: | The empirical evidence on the existence of social preferences—or lack thereof—is predominantly based on student samples. Yet, knowledge about whether these findings can be extended to the general population is still scarce. In this paper, we compare the distribution of social preferences in a student and in a representative general population sample. Using descriptive analysis and a rigorous clustering approach, we show that the distribution of the general population’s social preferences fundamentally differs from the students’ distribution. In the general population, three types emerge: an inequality averse, an altruistic, and a selfish type. In contrast, only the altruistic and the selfish types emerge in the student population. We show that differences in age and education are likely to explain these results. Younger and more educated individuals—which typically characterize students—not only tend to have lower degrees of other-regardingness but this reduction in other-regardingness radically reduces the share of inequality aversion among students. Differences in income, however, do not seem to affect social preferences. We corroborate our findings by examining nine further data sets that lead to a similar conclusion: students are far less inequality averse than the general population. These findings are important in view of the fact that almost all applications of social preference ideas involve the general population. |
| Keywords: | Social Preferences, Altruism, Inequality Aversion, Preference Heterogeneity, Subject pools, Sample Selection |
| JEL: | C80 C90 D30 D63 |
| Date: | 2024–01 |
| URL: | http://d.repec.org/n?u=RePEc:ies:wpaper:e202401&r=upt |
| By: | Kwabena Donkor; Lorenz Goette; Maximilian W. Müller; Eugen Dimant; Michael Kurschilgen |
| Abstract: | This paper examines how beliefs and preferences drive identity-conforming consumption or investments. We introduce a theory that explains how identity distorts individuals’ beliefs about potential outcomes and imposes psychic costs on benefiting from identity-incongruent sources. We substantiate our theoretical foundation through two lab-infield experiments on soccer betting in Kenya and the UK, where participants either had established affiliations with the teams involved or assumed a neutral stance. The results indicate that soccer fans have overoptimistic beliefs about match outcomes that align with their identity and bet significantly higher amounts on those than on outcomes of comparable games where they are neutral. After accounting for individuals’ beliefs and risk preferences, our structural estimates reveal that participants undervalue gains from identity-incongruent assets by 9% to 27%. Our counterfactual simulations imply that identity-specific beliefs account for 30% to 44% of the investment differences between neutral observers and supporters, with the remainder being due to identity preferences. |
| Keywords: | identity, investments, beliefs |
| JEL: | D91 G41 Z10 |
| Date: | 2023 |
| URL: | http://d.repec.org/n?u=RePEc:ces:ceswps:_10860&r=upt |
| By: | Huisheng Wang; H. Vicky Zhao |
| Abstract: | In this paper, we study the optimal investment problem involving two agents, where the decision of one agent is influenced by the other. To measure the distance between two agents' decisions, we introduce the average deviation. We formulate the stochastic optimal control problem considering herd behavior and derive the analytical solution through the variational method. We theoretically analyze the impact of users' herd behavior on the optimal decision by decomposing it into their rational decisions, which is called the rational decision decomposition. Furthermore, to quantify the preference for their rational decision over that of the other agent, we introduce the agent's investment opinion. Our study is validated through simulations on real stock data. |
| Date: | 2024–01 |
| URL: | http://d.repec.org/n?u=RePEc:arx:papers:2401.07183&r=upt |
| By: | Inessa BENCHORA; Aurélien LEROY; Louis RAFFESTIN |
| Abstract: | This article examines the impact of monetary policy (MP) on firms’ stock prices across CO2 emissions. We provide a theoretical model in which green firms are less sensitive to MP shocks than brown firms, because they are less exposed to transition risk and provide non-pecuniary utility to investors. We test this prediction by using a panel event-study regression approach on 857 US firms between 2010 and 2019. We find robust evidence that firms with high carbon intensity are significantly more affected by policy rate surprises. The sensitivity premium of brown firms remains significant when controlling for classic sources of MP heterogeneity, is persistent, and increases with climate awareness. Our results suggest that the market neutrality principle guiding the implementation of monetary policy could induce a bias toward brown firms. |
| Keywords: | Climate change, Transition risk, Carbon emissions, Monetary policy shocks, Risk premia. |
| JEL: | E44 O13 Q49 Q54 G12 G15 |
| Date: | 2023 |
| URL: | http://d.repec.org/n?u=RePEc:grt:bdxewp:2023-08&r=upt |
| By: | Yi-Hsuan Lin; Fernando Payró Chew |
| Abstract: | This paper models an agent that ranks actions with uncertain payoffs after observing a signal that could have been generated by multiple objective information structures. Under the assumption that the agent’s preferences conform to the multiple priors model (Gilboa and Schmeidler (1989)), we show that a simple behavioral axiom characterizes a generalization of Bayesian Updating. Our axiom requires that whenever all possible sources of information agree that it is more ’likely’ for an action with uncertain payoffs to be better than one with certain payoffs, the agent prefers the former. We also provide axiomatizations for several special cases. Finally, we consider the situation where the informational content of a signal is purely subjective. We characterize the existence of a subjective set of information structures under full Bayesian updating for two extreme cases: (i) No ex-ante state ambiguity, and (ii) No signal ambiguity. |
| Keywords: | updating, ambiguity, imprecise information, MaxMin |
| JEL: | D11 |
| Date: | 2024–01 |
| URL: | http://d.repec.org/n?u=RePEc:bge:wpaper:1424&r=upt |
| By: | Islas-Aguirre, Juan Francisco; Torres-Rojo, Juan Manuel; Venegas-Martínez, Francisco; Ríos-Bolívar, Humberto |
| Abstract: | Objetivo: Este trabajo tiene dos objetivos, primero, evaluar el efecto de las transferencias monetarias directas a productores agrícolas del Programa de Producción para el Bienestar (PPB) en el desarrollo humano de los beneficiarios en los municipios de México. Para ello se estiman regresiones cuadráticas separadas a nivel nacional y en las regiones agroalimentarias definidas por la Secretaría de Agricultura y Desarrollo Rural (SADER). El otro objetivo consiste en desarrollar un teórico del sector agrícola con agentes racionales idénticos para evaluar impacto del PPB en el bienestar económico (función de utilidad indirecta) y el crecimiento del sector agrícola a través de un análisis del estado estacionario de las razones de servicios públicos-capital privado y consumo-capital privado. Datos: Se emplea información a nivel municipal del Índice de Desarrollo Humano (IDH) para el año 2020 elaborado por el Programa de Naciones Unidas para el Desarrollo (PNUD) y del Sistema Único de Registro Institucional (SURI) del PPB. Resultados empíricos: Se encuentra un impacto positivo significativo de las transferencias sólo para los municipios con grado de desarrollo humano con especificación cuadrática, es decir, existen rendimientos crecientes a escala. Resultados teóricos: A partir del modelo teórico de equilibrio del sector agrícola propuesto, bajo un conjunto de supuestos convencionales, se de equilibrio del sector agrícolas se concluye que, en el largo plazo, el bienestar en el sector agrícola es mayor si se el gobierno invierte en servicios públicos para dicho sector en lugar de otorgar transferencias directas por un periodo de tiempo finito. // Objective: This paper has two main objectives, first, to evaluate the effect of direct monetary transfers to agricultural producers of the Productive Program for Well-being (PPB) on the human development of the beneficiaries in the municipalities of Mexico. For this, separate quadratic regressions are estimated at the national level and in the agrifood regions defined by the Ministry of Agriculture and Rural Development (SADER, Spanish acronym). The other objective is to develop a theory of the agricultural sector with identical rational agents to assess the impact of the PPB on economic welfare (indirect utility function) and growth of the agricultural sector through a steady state analysis of the ratios public services-private capital and consumption-private capital. Data: Information is used at the municipal level of the Human Development Index (IDH) for the year 2020 prepared by the United Nations Development Program (UNDP) and the Unique System of Institutional Registration (SURI) of the PPB. Empirical results: A significant positive impact of the transfers is found only for municipalities with a degree of human development with more quadratics, that is, there are increasing returns to scale. Theoretical results: Based on the proposed theoretical equilibrium model of the agricultural sector, under a set of conventional assumptions, it is concluded that the equilibrium of the agricultural sector, in the long term, welfare in the agricultural sector is greater if the government invests in public services for that sector instead of granting direct transfers for a finite time. |
| Keywords: | agricultores, desarrollo humano, estadística espacial, modelo de equilibrio del sector agrícola // farm households, human development, spatial models, equilibrium model of the agricultural sector |
| JEL: | Q14 |
| Date: | 2024–01–22 |
| URL: | http://d.repec.org/n?u=RePEc:pra:mprapa:119919&r=upt |
| By: | Tony Berrada (University of Geneva - Geneva Finance Research Institute (GFRI); Swiss Finance Institute); Peter Bossaerts (University of Cambridge); Giuseppe Ugazio (University of Geneva - Geneva Finance Research Institute (GFRI)) |
| Abstract: | In 1955, Herbert Simon proposed that economic agents do not optimize, but instead satisfice: they optimize up to some point of satisfaction. But Simon did not provide a formal model. Here, we develop a formal theory of a satisficing investor and consequent financial market equilibrium borrowing a technique from robust control in engineering, namely, Model Reference Based Adaptive Control (MRAC). Instead of optimizing a portfolio in terms of, say, a mean-variance trade off, the MRAC agent chooses portfolios that generate return distributions that minimize surprise with respect to a desired reference distribution. Surprisingly, the satisficing agent mostly acts “as if” optimizing, but we discover important – and realistic – deviations, such as willingness to accept risk even in the absence of a risk premium. This also implies that asset pricing may at times differ substantially from traditional theory. We motivate our modeling approach not only by pointing to benefits of robustness (robust control), but also with reference to recent developments in behavioral economics and decision neuroscience. |
| Date: | 2024–01 |
| URL: | http://d.repec.org/n?u=RePEc:chf:rpseri:rp2405&r=upt |
| By: | Ben Aoki-Sherwood; Catherine Bregou; David Liben-Nowell; Kiran Tomlinson; Thomas Zeng |
| Abstract: | A common theory of choice posits that individuals make choices in a two-step process, first selecting some subset of the alternatives to consider before making a selection from the resulting consideration set. However, inferring unobserved consideration sets (or item consideration probabilities) in this "consider then choose" setting poses significant challenges, because even simple models of consideration with strong independence assumptions are not identifiable, even if item utilities are known. We consider a natural extension of consider-then-choose models to a top-$k$ ranking setting, where we assume rankings are constructed according to a Plackett-Luce model after sampling a consideration set. While item consideration probabilities remain non-identified in this setting, we prove that knowledge of item utilities allows us to infer bounds on the relative sizes of consideration probabilities. Additionally, given a condition on the expected consideration set size, we derive absolute upper and lower bounds on item consideration probabilities. We also provide algorithms to tighten those bounds on consideration probabilities by propagating inferred constraints. Thus, we show that we can learn useful information about consideration probabilities despite not being able to identify them precisely. We demonstrate our methods on a ranking dataset from a psychology experiment with two different ranking tasks (one with fixed consideration sets and one with unknown consideration sets). This combination of data allows us to estimate utilities and then learn about unknown consideration probabilities using our bounds. |
| Date: | 2024–01 |
| URL: | http://d.repec.org/n?u=RePEc:arx:papers:2401.11016&r=upt |