Does Money Buy Happiness? A new answer to an old question
By Johannes Haushofer Jeremy Shapiro and Catherine Thomas
In 2011, Caren, a young woman from Western Kenya, received a text message saying that she had received an “unconditional cash transfer” from a charity called GiveDirectly. The NGO had sent her $1,500 to spend as she wished – no strings attached. Two years later, Caren explained that her relationships with her family and her community had changed dramatically. With the cash transfer, she and her husband had been able to move out of her mother-in-law’s house, as is customary for the eldest son, and into a house with a tin instead of a thatched roof. Additionally, Caren had begun to work outside the house. Consequently, Caren’s mother-in-law’s respect for her grew and the insults she had heard from her mother-in-law almost daily began to subside. No longer incited by his mother, Caren’s husband’s harsh words and physical abuse abated. Even in her community, her neighbors started coming to her to borrow tools. Caren felt less stressed and depressed, and reported being much happier than before.
Caren had received the GiveDirectly transfer as part of a randomized controlled trial that two of us (JS & JH) conducted in Western Kenya in 2011-2013. This setting allows us to ask whether receiving these cash transfers had similar effects on other people as it had on Caren. Would the average person experience a similar improvement in their psychological well-being as Caren?
The cycle of poverty and poor psychological wellbeing
Given that poverty is often marked by a state of chronic stress and is associated with poor psychological wellbeing on a range of other indicators, it would not be too surprising if cash transfers had increased psychological well-being. Indeed, depression and unhappiness disproportionately burden low-income populations, both within and across countries. In addition, our recent research reveals that poverty is also associated with other negative psychological outcomes, including greater feelings of meaninglessness and loneliness, and lower levels of trust and intrinsic motivation.
However, such observed associations between poverty and poor psychological wellbeing still don’t tell us which causes which, and it is here that a randomized controlled trial such as the one we conducted can help: in contrast to correlational studies, it establishes causality. In our study we also went beyond self-reports of psychological well-being by measuring changes in the stress hormone cortisol following a reduction in poverty levels. What did we find?
In this population, the burden of mental illness was remarkable. The average score on the CES-D depression questionnaire was 26.5, well above the clinical cut-off point for depression of 16 established in Western populations. While this significant burden of depression could partly be attributed to an imperfect translation of the clinical concept across cultures, it is unlikely to be the whole story. A reduction of poverty through cash transfers resulted in improvements in this measure: recipients scored one point lower on the depression scale than non-recipients, and also showed improvement on several other measures of psychological wellbeing, such as perceived stress, life satisfaction, happiness, and optimism. As predicted, these positive psychological effects were even greater when households were given large ($1,100) compared to small transfers ($300). Cortisol levels were significantly reduced in several treatment arms, including in households that received large transfers compared to small transfers, female-recipient compared to male-recipient households, and transfers delivered in a lump-sum compared to in monthly installments.
Although it is possible that giving cash to only some but not other households in a village would have negative externalities on the psychological wellbeing of non-recipient households, we find no such effects on average – in fact, optimism was actually boosted among control households in such villages. Furthermore, we find trends toward improvement in a number of other psychological outcomes and no significant negative psychological outcomes among these spillover households.
Interestingly, these improvements in psychological well-being were dependent on whether the husband or the wife in the household received the money: when women received the cash transfers, both males and females alike showed greater improvements in psychological well-being compared to when transfers went to men. Female empowerment may partly account for these observed differences between female- and male-recipient households: female empowerment, as measured in part by the absence of physical and sexual violence against women, was improved in the treatment group, and particularly in female recipient households. In the cross-section, female empowerment is a strong predictor of psychological well-being.
Thus, our findings suggest that cash transfers can improve psychological well-being – contrary to popular notions that such windfalls may lead to conflict and unhappiness. The fact that a reduction of poverty reduced symptoms of depression and stress suggests that poverty may indeed be a strong determinant of psychological wellbeing. An important implication of these results is that “treating poverty” may potentially contribute to alleviating the burden of poor psychological wellbeing among low-income populations. Future research is needed to determine whether addressing mental health directly through a psychological intervention, or addressing a root cause of poor mental health—financial strain—is most cost-effective for improving mental health. Additionally, future research should aim to determine whether enhancing psychological wellbeing in turn might have positive effects on household wealth.