The field experiment was conducted between 2008 and 2010. 845 households in Tangier whose occupants couldn't afford to connect to the city's water supplies were selected. Of these, 434 were randomly chosen and targeted in a door-to-door campaign to raise awareness of the possibility of getting an interest-free loan to have running water installed, and were given help with administrative procedures. The remaining participants - who had access to the same loan, but weren't targeted in this campaign - acted as a control group. The random allocation is key in insuring that both groups are initially similar so that we can attribute any difference afterwards to the intervention.
The results of the study illustrate the large impact of 'nudge' policies (i.e. policies which facilitate but don't mandate certain behaviours): in the first group, 69% had running water installed in the first 6 months of the campaign, as opposed to only 10% in the control group. One year later, 44% of the participants in the first group reported an improvement in quality of life, while only 22% of the others felt the same way.
Financially, this was a sustainable policy: as of mid-2010, 44% of households had repaid all of the loan, 28% owed less than 20% of the original amount, only 5% were late on more than 50% of the loan, no household had defaulted.
It must be noted that this policy resulted neither in income gain - in fact the average water bill increased from 11 USD to 21 USD - nor in better health, as all recipients already had access to clean water. As a result, this kind of intervention is unlikely to hasten self-perpetuating improvements in quality of life, nor beat our current best recommendations.
It seems that the main factor in increase in well-being was convenience - a legitimate desire in terms of life improvement. This is no small issue: in urban Morocco, people living in households without running water spend on average 7 hours a week collecting it from public taps. For almost two-thirds of them, access to water is a major concern, and for almost 30% it has been the cause of conflicts with family and neighbours. Thus this program contributed to reduction social tension, and freed up a significant amount of time. Interestingly, this spare time wasn't used to increase income nor to improve school performance by working and studying longer hours, but in social and leisure activities.
This highlights another point: improvements to people's quality of life are often 'consumed' right away, rather than 'invested' in raising children, further education or expanding a business. For this reason we can't count on the help we provide to necessarily trigger long-lasting economic growth.
This study shows that improving convenience of access to water greatly increases well-being. The value of installing water access can be increased by providing it to individual households. More generally, we may extrapolate that it is important, when measuring the impact of a policy in welfare, to consider convenience - and factors such as time for leisure and socializing - in addition to income and health. Furthermore, the study indicates that 'nudge' policies have a large effect on improving people's quality of life at a reduced public cost, and without interfering unduly with their lives. By making it easier for individuals and families to improve their own lives, it is possible to contribute to individual happiness and all-round social development in a cost-effective way.