Sunday, September 20, 2015

Putting your money where you can’t touch it

In a recent conversation with a financial advisor, he told us that some of his customers prefer not to have the ability to withdraw money from their saving accounts at any time. Instead, they prefer having to contact the financial advisor in order to gain access to their money. The natural question to ask is: Why do some people want to "bind their hands" and intentionally make it more difficult to withdraw and spend their money?  
Credit Card in ice : Stock-Foto
Credit card in ice (Getty images)
Behavioural economic theory provides an answer: Individuals who predict that they will be tempted in the future to withdraw money and use it for unwise purposes will want to put their money where they can’t touch it. They have a demand for commitment. In this post, we will summarise a few recent insights from the economics and the psychology of commitment in the context of financial decision making.
A key insight in the economic literature on commitments is that there is individual heterogeneity; not everybody wants to reduce their access to their saving accounts. Some individuals are just not tempted by the prospect of withdrawing money and spending it impulsively. They have no demand for commitment simply because they do not need it as their preferences do not change over time and are dynamically consistent. Other individuals tend to be tempted by the prospects of early cash withdrawals as they are present-biased. If present-biased individuals, however, do not possess enough self-awareness to predict future temptation episodes and thus are naïve, they do not demand commitment strategies because they overestimate their future financial discipline. The group who will have a demand for commitment is composed of individuals who are present-biased and tend to be tempted by cash withdrawals, and anticipate these temptations. Behavioural economics calls them sophisticated.
Demand for commitment accounts and present bias
Without present bias
With present bias
No demand for commitment because of time-consistent preferences
No demand for commitment due to lack of awareness of potential self-control problems
Positive demand for commitment because of awareness of potential self-control problems

Hence, individual differences in saving decisions are not only determined by cognitive biases, but also by the degree to which individuals are aware of these biases. For example, Goda et al. (2015) suggest that self-awareness regarding one's biases can be a stronger determinant of financial behaviour than the biases themselves. They find that self-awareness of potential biases has a positive effect on retirement savings even after controlling for measures of IQ, financial literacy and socio-demographic characteristics.
In an experimental setting, Beshears et al. (2015) investigate how many individuals are present biased and sophisticated by testing whether individuals have a demand for illiquid commitment savings accounts. Individuals could choose how to allocate money to an illiquid commitment account or a normal savings account without penalty for withdrawal. The interest rates of both accounts were varied. Participants allocated around half of their endowments to the commitment account when there was no difference in interest rates between the two vehicles, and one-quarter of their money even when the interest rate paid by the commitment account was lower than the liquid account. These findings suggest the presence of sophisticated present-biased individuals in the U.S. adult population. Beshears et al. (2015), however, also find evidence that the U.S. adult population contains naïve present biased individuals and/or individual who have consistent time-preferences.
Further support for the existence of sophisticated present-biased individuals comes from psychology. As outlined here, psychologists are in the process of re-defining the nature of trait self-control. The conventional view that high trait self-control is related to a strong ability to resist temptations is weakened in favour of the view that individuals with high scores on the trait self-control scale avoid being exposed to the temptations in the first place. This proactive use of self-control is only possible if individuals are aware of their self-control problems, i.e. are sophisticated.
Policy and business implications
How can policy-makers and business people such as financial advisors make use of these insights? Most importantly, the insights suggest that a one-size-fits-all policy regarding penalties for early withdrawals is likely to be problematic. As Beshears et al. (2015) point out, higher penalties for early withdrawal may either discourage or encourage savings, depending on whether individuals are present-biased and whether they know about it. Setting high withdraw tax penalties might increase savings of sophisticated individuals, but might reduce savings of naïve and dynamically consistent individuals as they do not like to put their money where they cannot touch it.
Instead, financial advisors could explicitly consider individual heterogeneity when advising their customers about the best ways to save. Eliciting whether individuals tend to engage in impulsive purchases from time to time would allow financial advisors to suggest specific saving vehicles with and without penalties for early withdrawal. To elicit whether individuals are present-biased, naïve, or sophisticated, short questionnaires could be offered to the customers to improve the advice.

References and further reading:
Angeletos, G; Laibson, D; Repetto, D; Tobacman, J; Weinberg, S. (2001). The Hyperbolic Consumption Model: Calibration, Simulation, and Empirical Evaluation. Journal of Economic Perspectives, 15 (3), p. 47–68.
Beshears, J; Choi, J; Harris, C; Laibson, D; Madrian, B; Sakong, J. (2015). Self Control and Commitment: Can Decreasing the Liquidity of Savings Account Increase Deposits. NBER Working Paper Series, No 21474, August. Available at: .
Delaney, L; Lades, L. (2015). Present Bias and Everyday Self Control Failures. Stirling Economics Discussion Paper, 2015-01, University of Stirling, July. Available at: .
Ent, M. R; Baumeister, R. F; and Tice, D. M. (2015). Trait Self-control and the Avoidance of Temptation. Personality and Individual Differences, 74, p. 12–15.
Goda, G; Levy, M; Manchester, C; Sojourner, A; Tasoff, J. (2015). The Role of Time Preferences and Exponential-Growth Bias in Retirement Savings. NBER Working Paper Series, No 21482, August. Available at: .
Laibson, D. (1997). Golden Eggs and Hyperbolic Discounting. Quarterly Journal of Economics, , 62 (2), p. 443–477.
O’Donoghue, Ted; Rabin, M. (1999). Doing It Now or Later. American Economic Review, 89 (1), p. 103–124.
O’Donoghue, Ted; Rabin, M. (2001). Choice and Procrastination. Quarterly Journal of Economics, 116 (1), p. 121–160.
O’Donoghue, Ted; Rabin, M. (2015). Present Bias: Lessons Learned and to be Learned. American Economic Review: Papers & Proceedings, 105(5), p. 273–279. Available at: .

Blog post by Bernardo Nunes and Leonhard Lades


Anonymous said...

My main discomfort with this is that it ignores the contexts in which savings decisions are made; in particular, it ignores the very real costs that illiquidity can impose on persons struggling to save in positions of economic precarity.

Bernardo Nunes said...

Hi Anonymous.
That's true. Commitment is costly too, but if it is demanded by part of the market participants, we should understand why. That's what the table tries to elucidate from the experimental results of the paper.

Unknown said...

I take the point being made by this article, and particularly appreciate the modulation introduced by awareness of one's biases and limits. But I do object to the perpetuation of a myopic linkage between 'present bias' and 'self-control problems'. Let us say that present bias is one (plausible) consequence of self control problems. But it can also be a consequence of prudence under conditions of economic precarity, with or without deficits of self control.

Bernardo Nunes said...

Well observed.

Heterogeneous financial conditions are not observed in the lab. Survey or administrative data are helpful here.

I think one of the recent contributions of this literature is that we need to rethink how to measure impatience and self-control in surveys. It points out that higher scores in trait SC are linked with individuals who avoid temptations by demanding commitment devices ex-ante.