Most research on happiness relies on surveys that ask people to reflect back on and evaluate their experiences ‘these days’ or ‘nowadays’. In doing so, respondents usually attach weight to events that are related to their overall sense of wellbeing or satisfaction with their lives.

These studies find consistent evidence that paid work contributes substantially to overall life satisfaction and general happiness (for example, Blanchflower and Oswald, 2011). The research also finds that becoming unemployed results in a precipitous decline in wellbeing, from which – unlike most other changes in personal circumstances – people do not recover, sometimes even after they are no longer unemployed (Clark et al, 2008).

Paid work is a central part of most people’s lives, so it should not be surprising to find that it is critical to the way we feel about ourselves and our sense of wellbeing. Perhaps it is because work generates a feeling of being worthwhile, leading to a sense of meaning or purpose in life.

But how does this square with the utility that individuals derive from the activities they perform day-to-day? After all, it is standard in economics to assume that there is a ‘disutility’ attached to work, which is why labour supply is responsive to wages – if the pay is higher, people will typically put in more hours and more effort.

Answering this question requires measures of momentary wellbeing that capture the feelings that individuals express at the time they are undertaking an activity rather than on reflection afterwards. We are able to do this using a new data source called Mappiness, which permits people to record their wellbeing via a smartphone.

The data contain more than a million observations on tens of thousands of individuals in the UK, collected since August 2010. We can analyse these data together with information about the personal, work and household characteristics that individuals provide when registering for Mappiness.

People who have downloaded the Mappiness app receive randomly timed ‘dings’ on their phone to request that they complete a very short survey. They are asked to rate how happy they feel and how relaxed they are; whether they are alone and, if not, whom they are with; whether they are indoors, outdoors or in a vehicle; and whether they are at home, at work or elsewhere. Finally, they are asked what they were doing ‘just now’.

The respondent chooses all that apply out of 40 response options, including working. Together with the responses to the survey, the app transmits the location of the individual (via satellite positioning) and the precise time at which the survey was completed. It also records the time elapsed between the random ‘ding’ and the response, thus making it possible to distinguish between immediate, ‘random’ responses and delayed responses.

Although there are drawbacks to this method of data collection, such as non-random sampling, it has considerable advantages over the more traditional ‘day reconstruction method’ (DRM), in which individuals are asked to reconstruct their activities and experiences of the preceding day. DRM responses may be subject to recall bias and retrospective distortion. In contrast, Mappiness obtains instantaneous responses so that people report their feelings at the time they are undertaking the activity.

We use these unique data to run analyses that isolate the independent association between episodes of paid work and momentary happiness and anxiety. On average people respond 60 times, so we focus primarily on variance within individuals over time, thus avoiding comparisons across individuals, which might be confounded by differences in their propensity for work and happiness.

So what do we find? First, and perhaps most strikingly, we find that paid work is ranked lower than any of the other 39 activities people engage in, with the exception of being sick in bed. A similar result was obtained in a small DRM study in the United States (Kahneman et al, 2004), but this is the first time the association has been uncovered in a large-scale study. The effect is equivalent to a 7-8 per cent reduction in happiness relative to circumstances in which someone is not working. Paid work has a similar though slightly larger negative impact on whether an individual feels relaxed.

Second, precisely how unhappy or anxious someone is while working depends on the circumstances. Wellbeing at work varies significantly with where you work (at home, at work or elsewhere); whether you are combining work with other activities; whether you are alone or with others; and the time of day or night at which you are working. Many of these circumstances can be influenced by public policies to facilitate working conditions conducive to ‘happier’ working – something that economists have noted can also improve productivity (Oswald et al, 2015).

We are left with the question as to why work appears to have such an adverse effect on people’s momentary wellbeing. We know that part of the answer is related to anxiety at work. Even though people are positive about paid work when reflecting on the meaning and value of their lives, actually engaging in paid work comes at some personal cost in terms of the pressures and stress they face while working.

But this is not the whole story. Working continues to be negatively correlated with happiness, even when it is combined with other activities that are pleasurable. Furthermore, even when we condition on feelings of relaxation, working continues to be negatively associated with momentary wellbeing.

Instead, it appears that work is negatively associated with hedonic wellbeing, such that we would rather be doing other things. In other words, work really is a disutility as economists traditionally conceive of it, one that requires some form of monetary reward to induce effort.



alex-brysonAlex Bryson is Professor of Quantitative Social Science at UCL’s Department of Social Science. He is also a Research Fellow at IZA, Rutgers and at NIESR, where he was previously Head of the Employment Group. Prior to that he was Research Director at the Policy Studies Institute.  Follow him on Twitter @AlexanderBryson. Email a.bryson@ucl.ac.uk



george-mackerronGeorge MacKerron is Lecturer in The Economics of Environment, Energy & Climate Change (Economics) at the University of Sussex and is co-creator of Mappiness. He is a Chief Technical Officer (CTO) at Psychological Technologies, which makes apps to support mindfulness and experience-sampling techniques. He’s a geo/web/app developer specialising in survey techniques and spatial data. He codes mainly in Objective-C/Cocoa, Ruby, CoffeeScript & PostGIS. He has worked on projects for the BBC and the University of Cambridge, amongst others. Email george@mackerron.com, Twitter @jawj, or find him at github.com/jawj.