Does Money Really Make Us Happier?

It is often easy to equate money with happiness and success. But does money really buy us happiness?

One study suggests it might. The author, Matt Killingsworth, claimed the study suggested that “higher incomes are associated with both feeling better day-to-day and being more satisfied with life overall” (p. 1). The study was then published in a top journal and made headlines around the world.

Some might take this study as evidence that money really does make us happier. But while this study has value, it might be difficult for the average person to interpret: the article is replete with concepts like logarithms, slopes and z-scores—concepts which many are unfamiliar with.

To help us get a better sense of what Killingsworth’s data really says, I want to re-analyze it here using what I call “plain stats” approach—an approach which focuses on presenting the data as clearly, comprehensibly and accessibly as possible.

To understand the data, we first have to understand how it was collected. The study involved an app which asked 33,391 adults “How do you feel right now?” at random points during their waking hours. The study collected 1,725,994 responses ranging from “Very bad” to “Very good.” Each individual’s reports were averaged into an average score representing their “emotional wellbeing”. This score was then analyzed in relation to their income, as reported in response to the question “What is your total annual household income before taxes?”.

So what did Killingsworth find once he did this study? To visualize the data, we can use a scatter plot—that is, a graph which, in this context, depicts where every individual is with respect to both their income and their wellbeing. In my view, scatter plots are one of the best plain stats tools for understanding how two things relate to each other, like income and wellbeing in this case. But since there’s 33,000 people, it would be too messy to present data for all of those people; so we can instead select a random sample of, say, 4,000 people which is representative of the results.

And when we do that, we get the following scatter plot (with code and data available at my website):

From Killingsworth's (2021) data

Source: John Eric Wilcox

In the plot, each dot represents an individual—albeit with some random noise added to make sure that the dots don’t all overlap. The so-called x-axis represents their income: individuals further to the right reported a higher yearly income than individuals to the left, spanning from those earning approximately $15k per year all the way up to millionaires earning up to $625k per year. The y-axis represents their average response to the question “How do you feel right now?”: individuals further to the top reported feeling better on average than individuals below.

The blue line is a so-called “regression line”: it is an idealized representation of the relationship between income and well-being. This line and the dots around it tell us many things.

The first is that wellbeing is influenced by many other things aside from income. If income was the only thing that impacted wellbeing, then every single dot would be on the line, and the only way to have higher wellbeing would be to have higher income.

But this is not what we see. Instead, we find that wellbeing differs drastically regardless of income. On the one hand, we see some individuals with income around $731k who on average report feeling closer to “Very bad” than to “Very good”. On the other hand, we also see some individuals with incomes closer to $18k who report feeling very close to “Very good”. This is not surprising, since we know from both science and experience that many things impact our wellbeing: our relationships, our work, our health, our leisurely activities and so on.

The other thing that the line shows is that even though other things influence wellbeing, there is still some positive relationship between income and wellbeing, but this relationship is very small. If income and wellbeing were entirely unrelated, that line would be flat. However, the slight slope of the line indicates that there is some relationship, albeit a very small one.

But one thing the line does not show is whether income is primarily responsible for this small relationship. Even if income is associated with higher wellbeing, this might be for other reasons: for example, perhaps individuals have higher incomes because they have the self-control necessary to pursue profitable avenues, and the evidence shows self-control is associated with a range of other positive outcomes in life which might affect one’s wellbeing.

Happiness Essential Reads

Japanese Wisdom for a Good Life

Why Thinking About the Future Makes Your Life Better Now

To tie all this together, the scatter plot indicates that the relationship between income and wellbeing is like the relationship between being tall and being in New Zealand versus Norway. As it turns out, some websites state that people are taller in Norway than in New Zealand, but because Norwegians are taller by only less than an inch, the relationship is small. And beyond that, it is not clear that being in Norway makes people taller: perhaps Norwegians on average just happen to be taller for other reasons, like genes or different patterns of immigration.

The scatter plot says income and wellbeing is like this too. There is a relationship between income and wellbeing, but it is also small, many other things matter too and it is not clear that income is responsible for the relationship—as opposed to other things that might be correlated with income, like self-control and its other consequences.

So while income is undoubtedly important and necessary for wellbeing to some extent, when we accurately analyze the data with a plain stats approach, it suggests we should not lose sight of its importance in the scheme of many other things that matter—and potentially much more so.