Retirement saving behavior: Transtheoretical Model change processes and life cycle tenets

Abstract

This dissertation sought to examine predictors of retirement savings behavior by analyzing data from the 2019 Survey of Consumer Finances (SCF 2019). The paper considered potential predictors through the lens of the Transtheoretical Model of Change (TTM) and tenets of the Behavioral Life Cycle Hypothesis (BLC). It is common for researchers to focus primarily on the stages of change in TTM without scrutinizing the underlying constructs within TTM that define change. This research delved more deeply into the underlying constructs than is typical of TTM research efforts. Guided by prior literature, survey items were chosen to model the various processes of change within the TTM model. Reasonable proxies were determined for all constructs except for the stimulus control process of change, which was a common missing variable within the literature. These survey items from the SCF 2019 were further divided into those that were considered to be cognitive factors and those considered to be behaviors. Cognitive factors were defined as any type of opinion, attitude, knowledge, or belief that was held by the participants. Behaviors were defined as actions either already taken or reported as what action the respondent would take. This process was not a scale, but a checklist for each construct modeled. For each process of change or underlying construct, the subject either demonstrated the response that would be expected of a person in the precontemplation, contemplation, or preparation stage of change or what would be expected of a person in the action or maintenance stage of change. Once this was accomplished, a logistic regression, both unweighted and weighted, was obtained to examine whether there were differences in the two groups regarding owning any type of retirement savings accounts, thus whether they were saving for retirement. Whether each household was saving or not saving specifically for retirement was the binary outcome variable. In addition to the role of TTM factors, this research further searched for evidence of the influence of BLC elements in the presence or absence of retirement savings behaviors. Survey items were chosen to model the specific tenets of the doer versus the planner, the implied recognition of assets belonging either in the current asset or future asset mental account, and the impact of retirement saliency on temptation. These items were included in the overall logistic regression model, both unweighted and weighted, to look for any differences in retirement savings behavior for each of these BLC elements. Ultimately this dissertation found evidence of statistical significance when considering the TTM constructs whether considering cognitive factors or behaviors. This was true in both the unweighted and weighted regressions. Except for retirement saliency in the unweighted logistic model the BLC tenets were not found to be statistically significant. This research is important because the responsibility for retirement savings shifted dramatically from the employer to the employee in recent decades and evidence suggests that the challenge of saving is not being met. Results of this study could help financial planners, counselors, and educators, as well as employers wishing to support their employees, with valuable information to influence the development of effective interventions designed to increase retirement saving behaviors within the general population.