The Underappreciated Costs of Chasing a Raise

People have outsized expectations of how much happiness money can buy

Would you be happier in a new job paying 20% more if it meant missing out on some Saturdays with your family? Or how about a transfer that offered the same raise but required getting up very early for a long commute?

In a working paper by UCLA Anderson’s Ricardo Perez-Truglia and Rafael Macedo Rubião, a Ph.D. candidate, subjects weren’t very good at making such forecasts. They expected almost seven times more life satisfaction from the money than the scientific evidence would suggest.

Learning that money typically buys only modest happiness gains — subjects were shown research based on national surveys and work with lottery winners — had a modest effect on decisions the subjects were facing in their own lives. They became somewhat less willing to trade free time, schedule flexibility and other things at stake in their specific circumstances for more money. 

Very Personal Trade-Offs 

The study’s 2,775 subjects were recruited through Prolific, an online service for researchers. Initially, each subject was asked to describe, to an AI interviewer, a real decision they were facing that trades income against something else. 

For example, one subject was trying to decide whether to take on more hours at work to help pay for their fiance’s desired wedding. Without the extra income, they would have a cheaper wedding. AI’s follow-up questions determined the either/or choice the individual was considering. In this case, as confirmed by the subject: earn more money for the wedding but spend less time with family and risk burnout.

All subjects were also asked how much a 20% raise would boost the average person’s happiness, and how much it would boost their own. 

As a group, they thought such an income boost would give other people a 5.2 point increase in life satisfaction on a 100-point scale. They said such a windfall would bring them, personally, even more happiness, with a 6.8 point boost.

People make these income/nonincome trade-offs in a multitude of settings. According to the personalized interviews with AI, about 25% of the study subjects were mulling job offers that might interfere with family time or other unpaid aspects important to them. Others were weighing things like postponing retirement, relocating for higher pay or staying in a financially secure but unhappy marriage, against schedule flexibility, proximity to family and friends or mental health generally. 

The authors analyzed these disparate personal dilemmas in a quantitative way; an approach that distinguishes this study from others in the field. Their results suggest that unrealistically rosy expectations for life with more money can lead to potentially disappointing real-world decisions, even when they do make us richer.

The Reality Check

Understanding how much more happiness money typically brings people — statistically called the “marginal satisfaction from income” — has been famously researched over decades. Perez-Truglia and Rubião sampled this prior work to benchmark their own study, comparing those findings with their subjects’ expectations.

The former research isn’t so optimistic about building happiness through more money. It finds that a 20% increase in income raises life satisfaction about 1 point on the scale, the authors report. The finding comes from both large-scale General Social Survey data that correlates income and life satisfaction, as well as a natural experiment that followed life satisfaction reports from lottery winners five to 20 years beyond their windfalls. 

Half of the subjects, the treatment group, were shown these research findings. The control group did not see them. Through three generic examples, all subjects were then asked to choose between more money and something else, like shorter commute or more sleep. 

Their fourth question described the trade-off they were facing in real life and asked them to choose. Working extra hours means less time with family and risks burnout, our example was reminded. So will you, A) work extra hours for more wedding money? Or B) not work extra hours?

Subjects that had seen the research findings rejected the additional money in each scenario more frequently than those in the control group. While the difference between reactions in the treatment and control groups was a few percentage points, the findings suggest that the information did affect some real life choices, even a month later.

Storytelling, Quantified

Research in this field by nature usually involves surveys, and these typically are analyzed by answers subjects give when presented with check-box options. They measure hypothetical reactions to hypothetical scenarios that may seem grossly irrelevant to the subject. (Not everyone gets offered a job with a 20% raise.) 

Notably, research Perez-Truglia collaborated on in 2025, looking at 1,396 U.S. tech workers weighing an actual job offer, found them on average willing to accept 25.6% less pay in exchange for partly or fully remote work, a far larger trade-off than prior studies had found.

Recent studies have outlined ways to use large language models to collect descriptive data from subjects, much like the way a social scientist traditionally does by interviewing and surveying subjects. They gather very personal stories, each with circumstances unlikely to match any other in the study. It’s difficult to generate data that can be aggregated and tested based on this kind of qualitative data alone. 

By conducting the AI interviews and creating the personalized trade-off questions, Perez-Truglia and Rubião essentially made these stories quantifiable. Each story created a response that could be compared against thousands of others.