What do we mean by “good jobs”? Moving beyond the basics
By now, pretty much everyone has heard of “anchoring bias” and “availability bias”: the tendency of people’s thoughts and assessments to depend on what they might previously have thought. I suspect some of the famous experiments that were initially used to “prove” the existence of this pattern may have been discredited. Based on a quick googling, though, the overall phenomenon seems valid. For example, the following has been reliably replicated: when doctors diagnose a patient, their diagnoses are significantly biased by which other patients, and diagnoses, they’ve seen previously.
At a much higher level, I think such anchoring effects are a hugely important, and challenging, factor at play when we talk about how to make the world, and people’s lives, better. In my experience, discussants’ mental models play a strong anchoring role. This is one of the reasons why I’m often frustrated by debates about “good jobs”, and indeed the broader metrics “beyond GDP”. It is as if, by starting with economic concepts such as “jobs” and “GDP”, we can’t bring ourselves to talk about the real essence of human wellbeing, and are only able to expand our sights a little from the original concept.
For example, most studies I’ve come across that talk about “good jobs” still focus on the monetary, contractual and physical aspects. Do the jobs pay well enough? Are they precarious? Are they dangerous? Do they involve anti-social working hours? Even when analysis goes beyond this, for example looking at the impact of one’s boss on one’s job satisfaction and life, the questions tend to be quite technocratic or technocratically worded. (E.g., in this study — the closest I’ve found to my earlier piece on “the boss factor” — of the 6 dimensions, none are worded so as to elicit feelings or emotions about work. The study is still very interesting; just limited by the undelrying survey.)
Yet, as I’ve argued before, in today’s economy, it is the psychological safety of jobs that is paramount. According to the Health and Safety Executive (HSE), in the year ending March 2021, 142 people in Great Britain were killed in fatal workplace accidents. That’s 0.0005% of the 31,278,200 people in employment in the same period. According to the (amazing!) Understanding Society survey, only around 4% people said they felt unsafe at work (in the wave that was conducted in 2019–2021).
However, when we look at some of the more emotional factors, the picture is quite different. I will return to this theme, but for now, I’ve focused on two variables in the Understanding Society survey (wave 10, conducted from January 2018 to June 2020). The left hand panel of the chart shows how often people reported feeling worried by their work. The exact question was worded as: “Thinking of the past few weeks, how much of the time has your job made you feel worried?” The good news is that almost half — 44% — answered “Never”, and only 6% said “Most of the time” or “All the time”.
However, quite a few — around a third — said that work made them feel worried “Some of the time”. More importantly, the higher the frequency at which someone said work worried them, the lower their satisfaction with their overall life. So not just satisfaction with work; satisfaction with life. For the moment, this is just a correlation, and omitted variable bias is almost certainly at play: someone who is of a neurotic tendency is more likely to say work made them feel worried and, separately, to be less satisfied with their lives (because a lot of other stuff also makes them feel worried, and in general, being worried is not a great state to be in, physically or mentally).
Nevertheless, it is my strong hypothesis that there is causality in the other direction, too: i.e., all other things equal, having work make you feel worried is likely to have a negative effect on your overall wellbeing and, hence, your overall life satisfaction. I will try to return with a better analysis that controls for personality and other relevant characteristics, soon.
[If anyone already has a thought-through DAG on “life satisfaction”, please send it my way! It will ensure that I control for confounders but not colliders. And if that’s greek to you, but you are interested in causal inference, I cannot recommed highly enough “The Book of Why” by Judae Pearl and Dana Mackenzie.]
