Principal Investigators: Sendhil Mullainathan (Harvard University), Heather Schofield (University of Pennsylvania)
Research Team: Nick Swanson
LEAD Centre: Small Enterprise Finance Centre (SEFC)
Focus Area:
Project Geography: Tamil Nadu, India
Status: Ongoing


One seventh of the world’s population remains below the level of caloric intake recommended by health professionals (FAO 2011). Given that food is not only a consumption good, but also an input into production, these low levels of intake have the potential to hinder labor market productivity, a fact modeled in a long line of theory literature (e.g., Bliss and Stern 1978; Stiglitz 1976; Dasgupta and Ray 1986). Yet, a revealed preference argument hinging on the fact that there is good evidence that the poor can afford to consume enough calories suggests that returns to higher caloric intake may be low. The inherent challenges posed by the endogeneity of caloric intake and cleanly measuring productivity have hindered empirical work in this area. Hence, it remains an open question whether there is an economically significant calorie-productivity gradient at the levels of caloric intake observed among the world’s poor today.

We have already run a relatively large-scale pilot RCT on the causal effect of improved nutrition on productivity and cognitive outcomes with over 200 low-income individuals. The preliminary results are encouraging: treated participants increase both labor supply and earnings roughly 10 percent by the fifth and final week of the pilot study as well improving performance on physical and cognitive laboratory tasks by 7 percent and 12 percent, respectively (Schofield 2014). However, to address concerns about the longer-run path of productivity changes, we are currently conducting an additional RCT designed to add rigorous evidence regarding the impact of increased caloric intake on both physical and mental productivity in the lab and labor market outcomes in the field over the course of roughly two months. This next round of the study also explores reasons why investment in caloric intake may remain low in the face of apparently high returns.

The study population will consist of 400 cycle rickshaw drivers with a BMI under 20. Participants will be randomized to a Control condition, receiving only cash compensation for their participation, or a Treatment condition, receiving an equal amount of cash and an additional 700 calories of food per day. Additionally, because survey evidence suggests that incorrect beliefs about the returns to calories and the caloric content of different foods may play a role in the low observed consumption, participants will be cross-randomized to a knowledge treatment at the end of the fifth week. This treatment will take the form of a motivational interview regarding the returns to caloric intake as well as the caloric density of foods. This cross-randomization, combined with experimental tasks described below, will shed light on important aspects of the decision-making surrounding food choices and allow us to determine whether there is learning from the experience of increased caloric intake, whether this information can be conveyed effectively through education, and whether there is an interaction between experience and information.