Training Effective Altruism

• In Event: Experiments in South Asian Politics

Abstract

We conduct a randomized evaluation implemented through a partnership with the Civil Service Academy (CSA) of Pakistan. A randomized soft-skill training of elite civil servants suggests that emphasizing the utilitarian value of empathy increases pro-social behavior in the lab and field. The elevated levels of altruism and charitable donation are reflected in the increased willingness to donate blood and make a field visit to an orphanage. Three results suggest the mechanism of effective altruists and theory of mind. First, blood donation only increased for those who were told their blood type was in need. Second, treatment increased cooperation and coordination in strategic dilemmas. Third, treatment increased theory of mind (in the guessing game). Redistributive preferences, trust, competitiveness, risk, and time preferences were unaffected. Linking to civil servants’ social media we find evidence that soft-skill training increases the use of “we” as opposed to “I” in tweets. Emphasizing the malleability of empathy did not lead to effects measurably different from the control, raising a cautionary note on extrapolating soft-skills training methods across cultural systems.
A total of four treatments were delivered while the control group received a placebo training unrelated to utility or malleability of empathy. This placebo group received a macroeconomics lecture taught in economics undergraduate programs. The training lecture that this placebo group underwent covered macroeconomic concepts that include the Gross Domestic Product, Gross National Product, Purchasing Power Parity, macroeconomic identities and the Solow model.
The impact of the two stand-alone utilitarian and malleability trainings and the joint training is evaluated by comparing outcomes across groups in a simple regression framework where outcome is regressed on dummy variables for various treatments and a set of control variables. We find that only stand-alone utilitarian treatment increases empathy.
We also investigate if the civil servants that obtain our treatments want to learn more about empathy. In a more revealed preference setting we offer to send one of the two books to each participant. The first book is on empathy (Mindsight: Transform Your Brain with the New Science of Empathy by Daniel J. Siegel) and the other is our control book, a book on econometrics (Mastering Metrics by Joshua Angrist and Jörn-Steffen Pischke). We find the group allocated the stand-alone utilitarian treatment is about 20 percentage points more likely to choose the empathy book over the econometrics book over the group that obtained a placebo training lecture. These results suggest that our training also made the utilitarian group more curious to learn about empathy.
We leverage unique information on blood groups of the bureaucrats and randomized phone calls to bureaucrats. The phone operator randomized the phone calls to the bureaucrats where half of the bureaucrats (106 participants) were randomly told that their particular blood group was in urgent need, while the other half (107 participants) were just made an urgent request to donate blood but without mention that their exact blood group is needed. These requests for blood donations were made after one month following the roll-out of the training lectures. The estimations reveal a large effect of the utilitarian treatment: the stand-alone utilitarian group is about 25 percentage points more likely to both agree to donate blood and set up a definite appointment with a blood bank relative to the group that received the placebo training. This is a substantial effect and equivalent to about 80% increase over the placebo mean. These results indicate that results from behavioral games map well to empathic behavior in the field: only the stand-alone utilitarian treatment has a qualitatively and statistically different effect, both on altruism scores in behavioral games and blood donations in the field relative to the placebo group.
Finally, we obtain data from Civil Service Academy (CSA) on their “syndicate field trip” at the conclusion of the Common Training Program (CTP). These entrant civil servants are given the option by the Academy, every year, to either visit a prominent Orphanage (Dar-ul-Aman) or attend lectures on a specific government program from a senior bureaucrat and do a corresponding field visit on the project site. We find consistent with the results on blood donations, the group assigned the stand-alone Utilitarian treatment is about 20 percentage points more likely to visit the Orphanage relative to attending the lecture by a senior bureaucrat. This is equivalent to about 80% increase over the placebo mean. Substantively, this result is particularly interesting for two key reasons: (1) the field visit took place at the end of January, that is, over 3 months after our training lectures were rolled out (2) These data come directly from the Pakistan Civil Service Administration and are part of their regular training curriculum, providing an “external” corroboration of our results.

Authors

• Shaheen Naseer, Lahore School of Economics

• Sultan Mehmood, University of Paris-Dauphine