Leader: Chiara Rapallini (UNIFI); Other collaborator(s): Erica Ordali
This task will investigate the effect of age on decision-making processes, particularly focusing on decisions requiring the assumption of risk (e.g., investing and savings). Analysis will include both a focus on the interaction between decision-making and healthy aging and a particular case of pathological aging (e.g., Parkinson’s disease).
As a first step, a meta-analysis of both the economic and the psychology literature will be conducted in order to test the research hypothesis that an aging population will take lower financial risk. Second, an experiment will be designed to be performed with samples of the population at different ages, distinguishing the assessment of risk preferences from the learning process. Experiments in the behavioral laboratory will give information for verifying if individuals of different ages are used to make different decisions. or if they follow different learning paths, and if these differences are larger in comparison with those usually verified among individuals of the same age in a given population.
Concerning the case of pathological aging, the focus will be on the measure in which dopaminergic medications alter the decision-making process of Parkinson’s patients with and without impulse control syndrome, either with a collection of behavioral data and by measuring the brain activity with EEG.
Brief description of the activities and of the intermediate results. The systematic investigation of the Economic and Psychology literature has been concluded ending up with more than 60 papers published between 1990 and 2023. Almost 300 partial effect sizes have been retrieved. Among the challenges in the coding process, the first is that the statistical methodology adopted by the two literature is distinct and only sometimes easily comparable. In fact, while in Psychology -and related fields- statistical comparisons among groups are mainly carried out with ANOVA models and t-tests, in Economics such comparisons are mainly investigated with OLS regression models, or non-linear models such as logit and probit. In addition, methodological differences are also related with the tasks and experimental designs used to assess preferences toward risk. Most papers pertaining to the economic field used surveys, and especially panel data, or incentivized lotteries, when using experimental protocols. In contrast, in psychology or neuroscience studies, the sample sizes are smaller, and the tasks used are taken from the clinical literature and rarely paid according to an incentive-compatible manner. For these reasons. we spent the last weeks to find a way for coding all the effect sizes using different but comparable coefficients. The preliminary results confirm our working initial hypothesis of an unclear picture of the relationship between aging and financial risk aversion, highlighting a cleavage between the Economics and Psychology literature. There are, in fact, evident positive associations between increasing age and risk aversion in Economic literature, even though covariates such as cognitive abilities, income, time preferences, and cohort effects tend to explain much of the significance. Confounding results are found in Psychology literature, where, depending on the task employed, we observe no relationship. Further analysis, mainly based on meta-regressions techniques, will be carried out to characterize better the methodological differences raised and the underlying cognitive mechanisms of task-related results.
Main policy, industrial and scientific implications. Several Western countries are experiencing decades of fertility decline, increasing life expectancy, and changes in the population's age structure. This change threatens long-term economic prosperity and the contract across generations, which is a foundation of the welfare state. One of the channels by which the aging population negatively affects economic growth and prosperity is the risk attitude that this population is willing to take. As such, understanding the relationship between ageing and financial risk attitude is crucial. This issue has been mainly investigated in the Economics and Psychology literature, with different methodologies for assessing this personality trait, and measuring individual differences.
Intermediate results: after completing the analysis of the results, the writing process for the draft of the study started.
Data Collection for a new experiment: Given the results obtained in the meta-analysis, the focus will now be on designing an original experiment to test the criticalities that emerged in the literature:
Scientific Publications: The writing of the first two articles (the meta-analysis and the Parkinson’s study) has been completed. The papers are either in the process of submission (meta-analysis) or accepted for publication with minor revisions (Parkinson’s study).
Policy Brief: A policy brief regarding the implications of our findings is currently under drafting.
Dissemination Activities: Dissemination efforts have begun since the beginning of September and are ongoing.
Brief description of the activities and of the intermediate results
The meta-analysis we conducted highlights significant conceptual and methodological differences in how risk preferences are assessed throughout the life course in both economic and psychological literature. These differences warrant deeper investigation, as risk preferences underpin numerous economic processes at both individual and group levels. Furthermore, changes in risk preferences across the life course represent a critical challenge for labor market patterns, economic dynamism, innovation, and the intergenerational contract at the foundation of welfare states. In particular, the methodology used to assess risk preferences is not the only factor that matters; cognitive abilities and health status also emerge as important determinants of risk preferences over time. However, these aspects are often overlooked in economic literature. To address this issue, our new project aims to identify differences between two experimental groups (young vs. old) regarding their attitudes toward risk in a standard lottery task conducted before and after a learning process. Specifically, we aim to experimentally investigate the role of learning in modifying baseline risk preferences. We will also examine whether this process, in interaction with aging and cognitive abilities, can better explain the observed discrepancies between self-reported risk-taking and risk behavior elicited through experimental tasks.
Brief description of the activities and of the intermediate results.
The meta-analysis titled “Aging and Financial Risk-Taking: A Meta-analysis” has been accepted for revision in the Journal of Economic Psychology, and we are currently in the process of revise and resubmit, addressing all the reviewers’ comments. The second project, preliminary titled “Measuring Changes in Risk Preferences Over the Life Cycle: Experimental Evidence from Economics and Psychology” is in the final design stage. We submitted the protocol to the local Ethical Committee to be evaluated and approved, and we are currently waiting for their output in order to proceed with pre-registration of the expected findings and data collection.
Publications
Dissemination Activities
Past
Forthcoming