research

Empowering people to better forecast their future expenses and savings: the role of risk information

A TFI research project by Dilek Önkal, Wasim Ahmed & Shari de Baets
Posted on January 09, 2020

“Nudging for improved projections of future expenses and savings” is one of the short-term research projects supported by the Think Forward initiative.

What are the chances that your car breaks down or you suddenly lose your job? Are you overoptimistic, believing these events won’t happen to you? And do you rather think about the present than the future? You are not the only one! Many people find it difficult to save for a rainy day, risking future financial problems. So how could we help people make better predictions of their financial situation in the future? This is exactly what researchers Dilek Önkal, Wasim Ahmed & Shari de Baets asked themselves.

In their TFI research they studied how awareness of future unexpected shocks could help people make better forecasts of future expenses and savings. Do people become more prudent in their estimates when they know about the likelihood that they will experience a potential financial shock?

Download the report and find out!

SUMMARY

Many households lack the necessary savings to deal with unexpected events, like a car breaking down or a spouse being sacked (Grinstein-Weiss, Russell, Gale, Key, & Ariely, 2017; Hogarth, Anguelov, & Lee, 2003; Lusardi, Schneider, & Tufano, 2011). People buffer less than ever, whilst dealing with a burden of debt (OECD, 2013). Most of us suffer from economic insecurity, risking economic problems in the near or far future (Weller & Logan, 2009).

Forecasting savings and expenses are challenging tasks requiring realistic planning skills and a detailed assessment of our financial situation. To complicate things, biases may lead to savings and expenses fallacies: as human beings, we tend to be more oriented towards the present than the future (Frederick, Loewenstein, & O'Donoghue, 2002; Tam & Dholakia, 2011) and prefer instant gratification over long-run benefits (Ainslie, 1992; Angeletos, Laibson, Repetto, Tobacman, & Weinberg, 2001; O'Donoghue & Rabin, 1999).


Seeing life through rose-tinted glasses (or not)

Future rise in expenses also tend to be underestimated, and the same goes for the risk of unexpected expenses in the near future versus the near past (Howard, Hardisty, Sussman, & Knoll, 2016). People don't expect their car to break down soon, even though they know this might happen at any time, just like it might happen to anyone. Broader research on judgment and decision making teaches us that this so-called 'optimism bias' is true for most people (Weinstein & Klein, 1996). We tend to turn a blind eye to negative scenarios and focus on the positive ones (Newby-Clark, Ross, Buehler, Koehler, & Griffin, 2000).

We are also overly optimistic about the resources we expect to have at our disposal over time. When asked to think about the future, we tend to envision a very limited number of scenarios, including hopes and preferences, which leads to generally overoptimistic expectations (Newby-Clark et al., 2000).


The experiment: conceptual framework

We hypothesize that embedding plausible risk information in various scenarios will lead to changes in forecasts of savings and expenses. Facing potential risks on future events could act as a harsh reality check, resulting in adjustments of financial forecasts.

Secondly, we hypothesize that such adjustments can be affected by the context of each scenario. A scenario describing an unexpected expense situation, for instance, could lead to a revision of expense predictions, whilst an income loss situation could lead to changes in savings forecasts.

Thirdly, it is important to acknowledge that not everyone will change their forecasts after reading such scenarios. We hypothesize that such resistance to changing one's projections could depend on the risk level. If, for instance, one reads a risk scenario whilst not facing any immediate risks in real life, one would be more resistant to adjusting one's initial projections.


Implementation of the experiment

We ran a pilot study to identify key realistic events that could significantly influence expenses and savings plans. Findings showed that unexpected loss of income and unexpected expense situations were the most influential, and these scenarios were used in the main study (see Appendix A for scenario details).

Through an online study, we asked our 325 participants to give forecasts for expenses and savings, along with their estimates of emergency funds savings, retirement savings, personal savings, and target savings over the next three months. Afterwards, they were given a scenario (either describing an unexpected income loss or an unexpected expense situation) where the risk level of the unexpected event was high, low or no-risk.

After reading the scenarios, participants were asked to rate the likelihood of this scenario happening to them. To what extent would it impact their financial situation? At this point, participants could also make adjustments to their forecasts in light of the scenario information.


Finding 1 | forecasting expenses

Making realistic forecasts of expenses is critical to successful financial planning. Comparing people’s expense predictions before and after scenario information, we find two effects. Firstly, expense forecasts are significantly increased when individuals read scenarios describing an unexpected expense situation. Secondly, people reduce their expense forecasts after reading scenarios describing an unexpected loss of income.


Finding 2 | forecasting savings

Savings forecasts seem more resilient to change based on scenario context. Individuals appear largely committed to their initial forecasts. We do, however, see small changes in predicted savings after exposure to a high-risk scenario or after exposure to a no-risk scenario. After reading a high-risk scenario, people adjust their savings forecasts slightly downwards, as if they expect to be able to save less in anticipation of potential risks to their financial situation. Interestingly, people slightly increase their savings projections after reading a no-risk scenario, potentially due to a recognized need to save in case they are exposed to a negative event - even though this time they were lucky to get by with a near miss.

We also find that people set target savings targets that are significantly higher than their savings forecasts. Interestingly, if we add forecasts for different subcategories of savings (i.e., retirement savings, personal savings, emergency fund savings), this seems to be equivalent to target savings, while falling short of their overall savings forecasts. Basically, if you want to encourage savings, either work in categories, or work with targets!


Finding 3 | forecasting changes and risky scenarios

An important part of adjustment behaviour is not changing the initial forecasts, as it reveals acceptance/resistance to new information. In our case, not changing a forecast reflects how influential the risk information is for financial predictions. Figure 1 shows the total number of participants and the ‘no-changers’ per risk level.


Figure 1: Total number of participants and total number of no-changers per risk level

Figure 1: Total number of participants and total number of no-changers per risk level

Does this resistance to change depend on what is being forecasted? Figure 2 shows that predicted savings (PS) and predicted expenses (PE) may be changed after receiving risk information (with 55% of savings and 68% of expense forecasts adjusted after the risk scenarios). Retirement savings (RS) and emergency fund savings (EFS) remain the highest unchanged forecasts (with 89% and 70% unchanged projections). About 62% of forecasts for target savings (TS) and 60% of personal savings (PS) also remain untouched.


Figure 2: No-adjustments per category (expressed as percentage of total responses)

Figure 2: No-adjustments per category (expressed as percentage of total responses)


Potential implications

In this study, we set out to investigate how we could improve savings and expense forecasts, knowing that many households lack adequate backup funds (e.g., Hogarth et al., 2003; Lusardi et al., 2011). Our findings show that exposing people to financial vulnerability scenarios may serve as a reality check, leading to adjustments in personal financial predictions. Providing risk information through vulnerability scenarios offers a prolific toolbox in designing nudges towards better-informed financial forecasts and heightened financial awareness.

We suggest that further work on nudge designs for personal finance includes easily accessible smartphone apps, wearable gadgets and other smart devices. These will maximize the effectiveness and full integration into people's financial planning.


RainyDay: an app to improve people's financial wellbeing

Smartphone apps have a broad appeal, but can be more interactive and exciting, allowing users to experience a deeper level of engagement with the research. They may have the power to educate younger generations about better financial planning.

We used our research findings to develop a financial awareness app called RainyDay. This app provides advice and information, empowering users to save money for potential ‘rainy days’. It gives them the opportunity to engage with our research through interactive quizzes, native video viewing, and third-party plugins (please see Appendix B for details of our dissemination work and the RainyDay app).

The RainyDay app is an innovative tool to deliver positive societal impact with our research, helping people feel more in control their personal finances. It also provides a novel way to put research into practice, serving as a useful tool for local authorities, civil society organizations and financial institutions. Embedding effective nudges into our daily savings and expenses plans and projections promises to have a significant effect on people's financial wellbeing. Our findings - and our app - offer new ways of influencing a wider audience, and making a change.


References

  • Ainslie, G. (1992). Picoeconomics. Cambridge: Cambridge University Press.
  • Angeletos, G.-M., Laibson, D., Repetto, A., Tobacman, J., & Weinberg, S. (2001). The Hyperbolic Consumption Model: Calibration, Simulation, and Empirical Evaluation. Journal of Economic Perspectives, 15(47-68).
  • Frederick, S., Loewenstein, G., & O'Donoghue, T. (2002). Time Discounting and Time Preference: A Critical Review. Journal of Economic Literature, 40(2), 351-401.
  • Grinstein-Weiss, M., Russell, B. D., Gale, W. G., Key, C., & Ariely, D. (2017). Behavioral Interventions to Increase Tax-Time Saving: Evidence from a National Randomized Trial. The Journal of Consumer Affairs, Spring, 3-26.
  • Hogarth, J. M., Anguelov, C. E., & Lee, J. (2003). Can the poor save? Journal of Financial Counseling and Planning, 14(1), 1-18.
  • Howard, C., Hardisty, D., Sussman, A., & Knoll, M. (2016). Understanding the Expense Prediction Bias". In P. Moreau & S. Puntoni (Eds.), Advances in Consumer Research (Vol. 44, pp. 190-194). Duluth, MN: Association for Consumer Research.
  • Lusardi, A., Schneider, D. J., & Tufano, P. (2011). Financially fragile households: Evidence and implications. Retrieved from https://www.nber.org/papers/w17072
  • Newby-Clark, I. R., Ross, M., Buehler, R., Koehler, D. J., & Griffin, D. (2000). People focus on optimistic scenarios and disregard pessimistic scenarios while predicting task completion times. Journal of Experimental Psychology: Applied, 6(3), 171-182.
  • O'Donoghue, T., & Rabin, M. M. (1999). Doing it now or later. American Economic Review, 89(1), 103-124.
  • OECD. (2013). OECD economic outlook no. 94.
  • Tam, L., & Dholakia, U. M. (2011). Delay and duration effects of time frames on personal savings estimates and behavior. Organizational Behavior & Human Decision Processes, 114, 142-152.
  • Weinstein, N. D., & Klein, W. M. (1996). Unrealistic Optimism: Present and Future. Journal of Social and Clinical Psychology, 15(1), 1-8.
  • Weller, C. E., & Logan, A. M. (2009). Measuring Middle Class Economic Security. Journal of Economic Issues, 43(2), 327-336.