Why subjective views matter: the case of incomes and life satisfaction

In my experience, there are two camps when it comes to perceptions and reality. Those who think what matters is “objective” reality: whether people have jobs, reasonable incomes, decent houses, etc. And those, like myself, who think perceptions matter more. Why is that?

I could probably write a whole book about this, but let’s just mention the top three reasons why I think people’s subjective (consciouos or subconscious) sense of a situation is important — and probably more important than many decision makers think.

First, there is no such thing as “objective” reality. Certainly not in social sciences. Exactly what you decide to measure, exactly how you calculate the components of it, exactly what data gathering strategy you deploy, exactly what adjustments you make, all have an enormous impact on whatever it is that you are measuring. [My article on levelling up provides some specific examples of all these questions in that context.] There are hundreds, if not thousands, of small process and value judgements that go into measuring something like, say, real GDP.

Second, people’s behaviour and choices — which in turn deliver some outcomes we might care about, such as economic output— are by definition driven by their perceptions, not by any “objective” reality. [In some cases, such as someone seeking shade because the direct sunlight feels too hot, the perception and reality (in as much as it can be measured) can be quite close to each other. In other cases, such as someone being unhappy with their income because they think everyone else makes more, perception might be quite far removed from reality.] If social sciences are trying to understand human behaviour, and meet human needs, then there is no escaping the centrality of subjective thinking and feeling.

Third, I’ve never understood why some “thinker” or “leader” — typically sitting in the proverbial ivory tower — could claim to have a better understanding of someone’s life than that someone themselves. Of course, scientists will have complementary information about general human tendencies, and will also often be able to observe more about someone’s situation. And of course, decision makers need to have some idea about what does help people live a better life. But ultimately, when it comes to assessing the quality of that life, I would say that the individual’s view matters more.

Today’s chart illustrates this point. In line with my argumentation above, I wanted to understand how someone’s income contributes to their life satisfaction. The latter is explicitly a subjective variable, asking people to assess how satisfied with their lives they are, all things considered. Here, I’ve contrasted two metrics. [I’ve not had time to do a careful “all other things equal” analysis, but I think it’s still instructive.]

On the left, we see the relationship between actual incomes and life satisfaction in the UK (in the 2018–20 wave of the Understanding Society survey). On the right is the relationship between each respondent’s subjective statement about their financial situation and life satisfaction. The widths of the bars indicate the number of adults in the general population that typically fall into each category (based on the Understanding Society survey results and cross-sectional weights).

This juxtaposition reveals a couple of important points. First, at this aggregate level (within one country, the UK), there is practically no relationship between someone’s income and life satisfaction (left hand panel of the chart). Careful econometric analyses do reveal a link, but it is often weaker than you would expect and logarithmically shaped (meaning that a fixed increase in income tends to lead to a lower and lower increase in life satisfaction as the starting-point income increases).

Second, the link between someone’s subjective financial situation and life satisfaction is strong (right hand panel of the chart). As I mentioned in a previous post, in a machine learning analysis of around 400 variables, someone’s subjective financial situation was the second most important variable predicting life satisfaction. So this clearly does matter to people’s feelings about their overall situation. In other words, money matters, but only in the context of people’s own lives.

Third, and I will return to this in a future post, there is a surprisingly weak relationship between someone’s “objective” income (say, as measured here) and their subjective assessment of their income. Just to give you a sneak preview into the data, the average net equivalised income per person in the sample is around £2,000 per month. Yet, there are many, many people whose income is twice as much but say that they are “Finding it quite difficult” or “Just about getting by” financially. At the other end, there are also a lot of people whose income is less than half the average and who say they are “Doing alright” or “Living comfortably”.

So, context and expectations matter enormously to people’s overall sense of wellbeing and satisfaction. That makes analysing things like “social welfare” or “social impact” much harder — but just because it’s hard doesn’t mean we shouldn’t try.