The value of work extends beyond a paycheck. I expect that few (if any) readers will be surprised by this statement. The notion that there are non-pecuniary benefits to employment is not new. Unemployment (at least in high income countries) has been linked to depression, for example. But to what extent does this link exists net of the income effect of work? Here we have less evidence in low income settings. Reshmaan Hussam, Erin Kelley, Gregory Lane, and Fatima Zahra address this gap in their forthcoming AER study of Rohingya refugees. They study the multiple dimensions of the value of work beyond income. And they find big (big!) effects.
The setting is three Rohingya refugee camps situated upon the southern tip of Bangladesh which offer very few work opportunities. Lack of work is pervasive: just one in ten adults 18-45 worked in the past month. This is partly by design. For these refugees in Bangladesh, formal employment is illegal and informal work in cities is difficult to obtain due to mobility restrictions. In additional to extreme un/under-employment, this population suffers in many other dimensions. Three quarters of the sample studied qualified as depressed.
Hussam et al. surveyed 745 residents of working age (one-third female by design) across these 3 camps. The sample was assigned to one of three arms for the 8-week intervention. The employment arm entailed part-time work in the form of a surveying assignment (data collection exercise) over 8 weeks. Take-up of the employment offer was 100%. The cash arm was designed to isolate the psychosocial value and other benefits of work net of the income effect. The residents in this arm were given cash equivalent to the employment arm, in exchange for weekly survey participation. A third group of residents received a small fee for weekly survey participation, and served as a control arm. Psychological well-being was measured with a psychosocial index -- an aggregation of the standardized measures of depression, stress, life satisfaction, self-worth, sociability, locus of control, and sense of stability.
There are lots of additional details about the study in the paper: that participants were aware of the randomization process, that the selection of blocks within camps was designed to minimize spillovers, details about the value of the payments ($0.60/week for the control survey group and $5.30/week for employment and cash arms, described as sizeable). The paper also describes deliberate consideration of the attributes of the employment offered and how work completion was tracked. The study was informed by sociological literature from the early 1970s on key features of work in regards to having a nominal social component and a purpose. Lastly, the work was designed to not require literacy to complete the task (a visual worksheet is used for data collection).
Work improved psychological wellbeing. The psychosocial index of refugees employed significantly increased compared to the control group (no work offered). These effects were much larger than those found from a year-long psychoeducation program (44 weekly sessions of counseling) for Rohingya refugee mothers implemented around the same time in the same camps (0.24 SD compared to 0.15 SD, respectively).
Is this about the money? No. Employment improved psychosocial well-being at a magnitude four times greater than impact from receipt of an equivalent amount of cash received but with no work. The subcomponents of the psychosocial index also significantly increased from the employment offer beyond the income effect. Cash transfers have been shown to increase happiness (see Markus’ last blog of a cash transfer study in Uganda, among other studies), but maybe a job can do even more.
And it is not just psychosocial measures which improved with work net of income effects. Work resulted in improvements in memory and math tests, and reduced rates of felling physically ill. It also lowered risk aversion. These effects could be because the work itself made workers healthier – that is, exercise is good for us. But weekly data on days healthy did not suggest that physical health drives these results. There was no increase after the first week of work on healthy days; it remained steady in subsequent weeks (we would expect the exercise effect to grow over time). They also find that these effects were not simply the result of becoming employed (the impact of having a label of being employed), rather, the experience of the work itself mattered – additional days of work (net of the income earned) increased these effects.
Notably (though I won’t say more on this), this 8 weeks of work for women significantly increased their beliefs that women should be permitted to work outside the home.
To learn more about why employment brings a psychosocial value, they measured changes in self-worth with regards to family and community. Employment significantly increased a worker’s perception of how valuable they are to their family. This reminds me of the emphasis in the 2013 World Development Report on Jobs (on which I worked) that jobs can be a critical source of self-respect and social identity, and can contribute positively to how people view themselves and their relations with others, beyond the paycheck received. On the other hand, Hussam et al. find suggestive results of a disempowering effect of cash on men’s (not women’s) perceived status in the family.
So jobs can mean much more than the value of the paycheck. To further understand how much more, they further offered workers an extra week of work at varying wage rates (following the incentivized Becker-DeGroot-Marschak, BDM, method). Almost everyone accepted the additional work offer….And three quarters of them were willing to work for zero pay… And about three quarters of those willing to work for free are also willing to forego cash for no work-- in order to work for free.
The authors are over-cautious in describing their study as “proof-of-concept”. Even with the caveats about this specific context (to be sure, it is an incredibly difficult setting), I concur with them that their study offers insights to policy makers choosing between cash and workfare. Of course, it is important to keep in mind that the employment they offer is quite different from the attributes of typical public works (or workfare) in low-income countries, or at least the ones with which I am familiar. It is a less physical job (as compared to road rehabilitation, irrigation/dam repair/building, and afforestation) and more conveniently located near the participants dwelling (compared to more distant work sites). There are also gender dimensions which differ. Women bear the burden of arranging childcare which is arguably easier to arrange in this study where the work is closer to home and is 2.5 hours a day, and the worker chooses when in the assigned day. Nonetheless, this innovative study is a valuable addition to the evidence on cash and “cash plus”.
Join the Conversation