Learning from the lab: Optimism and the likelihood to overspend

Background

Budgets are thought to play an important role in shaping individual spending behavior—they provide rules that help us curb spending as we try to increase savings, repay debt, manage financial volatility, or to achieve any number of our financial goals. Furthermore, we often believe that budgets make us happier over time because they will help us spend more on things that make us happier and less on things that we regret. Budgets are often seen as essential for avoiding high-cost consumer debt, like credit cards.

However, there is plenty of evidence that we struggle to adhere to our budgets and that budgets are not all that effective in helping us reach our financial goals. People often find it difficult to rein in overspending, leading to revolving debt with high interest rates.

For this reason, we were interested in exploring how budgeting informs the ways individuals make singular spending decisions. Specifically, we were interested in better understanding how individuals incorporate budget reference points into their spending decisions, both within a single time period and across multiple time periods.

Key Insights

There are many behavioral and psychological factors that are believed to be why people struggle to adhere to the budgets they set for themselves. We wanted to isolate two specific factors in this experiment.

• The what-the-hell effect: When people break a rule such as overspending on a budget, people may feel licensed to spend even more. Once a rule has been broken, people may no longer feel constrained to engage in that behavior.

• Planning fallacy: People are often overly optimistic when assessing their capacity to achieve their goals in the future. One reason people may struggle to adhere to their budgets is they assume that they will compensate for temptation in the moment with strictly adhering to the rules they set for themselves in the future.

Experiment

To test these ideas, we created a simple hypothetical scenario. Participants were asked to imagine they are coming home from work. They were presented a choice: either to order take-out from their favorite restaurant or to cook at home. They are then told they have previously given themselves a budget of $30 to spend on food each week and shown a summary of their last ten expenses.

Participants are then randomly assigned into one of three weekly spending conditions.

• In the “under budget” condition, their spending on food this week adds up to $29, or just under their weekly budget.

• In the “at budget” condition, their spending on food this week adds up to exactly $30, or exactly their weekly budget.

• In the “over budget” condition, their spending on food this week adds up to $31, or just over their weekly budget.

Participants will also be randomly assigned into one of two future spending conditions.

• In the “reminder” condition, participants will be additionally told that even though ordering takeout will put them over their weekly budget, they have only spent half of their budget for the full month. They are also shown a “monthly budget” graphic depicting their spending showing opportunities to reduce future spending.

• In the “no reminder” condition, participants are not reminded of their monthly budget.

All participants were then asked, on a scale 0%-100%, how likely they are to purchase take-out even though they have groceries at home. Then, participants were also asked, on a scale 0%-100%, how likely they are to reduce spending on food in the following week.

Results

We recruited just under 500 participants to take our study. There was a wide distribution in ages, ranging from early 20s to mid-70s, about 66% of the sample was male, and the sample skewed lower income, with nearly 70% earning $60,000 or less.

We found that a majority (~67%) of participants said they would likely try to reduce their spending on takeout the following week. We did not find any effect of the reminder condition on either ordering food or reducing expenses in the future. However, the likelihood that a participant reported that they would try to reduce their spending was different between the budget conditions. As one might expect, participants who were informed that they had already overspent their budget were significantly more likely to report trying to reduce their spending in the future (p=0.04).

However, this did not change their likelihood of ordering take-out from their favorite restaurant. We found that across all conditions, about 40% of participants said they would order food and that this percentage was not meaningfully different between conditions. We also found that if someone said they would reduce their spending, they were also significantly more likely to say they would order food (p=<0.001).

Taken together, this suggests that people’s optimism about their behavior in the future exacerbates willingness to spend today. The more people feel like they will adjust their spending in the future, the more likely they are to overspend in the present.