Training Effective Altruism: Experimental Evidence from Pakistan

Last registered on March 13, 2023

View Trial History

Pre-Trial

Trial Information

General Information

Title

Training Effective Altruism: Experimental Evidence from Pakistan

RCT ID

AEARCTR-0006655

Initial registration date

October 22, 2020

Initial registration date is when the trial was registered.

It corresponds to when the registration was submitted to the Registry to be reviewed for publication.

First published

October 26, 2020, 2:13 PM EDT

First published corresponds to when the trial was first made public on the Registry after being reviewed.

Last updated

March 13, 2023, 9:41 AM EDT

Last updated is the most recent time when changes to the trial's registration were published.

Locations

Country

Pakistan

Region

Primary Investigator

Name

Daniel Chen

Affiliation

Toulouse School of Economics

Contact Primary Investigator

Other Primary Investigator(s)

PI Name

Shaheen Naseer

PI Affiliation

Lahore School of Economics

Contact Investigator

PI Name

Sultan Mehmood

PI Affiliation

New Economic School

Contact Investigator

Additional Trial Information

Status

Completed

Start date

2020-10-31

End date

2021-10-14

Keywords

Education, Governance

Additional Keywords

Civil service, social-emotional learning, empathy

JEL code(s)

Secondary IDs

Prior work

This trial does not extend or rely on any prior RCTs.

Abstract

Empathy is said to be the antidote to in-group bias. In this study we evaluate the causal effects of empathy training on elite civil servants in collaboration with the civil service administration and assess its impact on social preferences, bureaucratic performance, and thought leadership. Recent research suggests that emphasizing malleability of empathy as being key to behavioral change. In a factorial design, we also assess the impacts of emotional intelligence training, the impact of reading material whose content application is assessed via social emotional learning exercises. We also randomize econometrics book (mastering metrics) across the civil servants subject to their selection of the metrics book, where book on empathy serves as a placebo.

Update March 13, 2023: We now also are collecting data on the universe of spending decisions by the civil servants and evaluating the impact of the training workshop on empathy in Pakistan on policy outcomes in year 1 and year 2 post-workshop as they perform their official duties. Specifically, we will use government audit documents to investigate the spending decisions of treated versus placebo ministers across 12 government departments (which represent the universe of their spending choices). Our main hypothesis is that altruism training will increase spending on social policies (i.e. will raise ministers’ budgetary recommendations to education and health) and reduce spending on self-oriented policies (i.e. budgetary recommendations for personal security and office maintenance).

External Link(s)

Registration Citation

Citation

Chen, Daniel, Sultan Mehmood and Shaheen Naseer. 2023. "Training Effective Altruism: Experimental Evidence from Pakistan." AEA RCT Registry. March 13. https://doi.org/10.1257/rct.6655-2.1

Sponsors & Partners

There is information in this trial unavailable to the public. Use the button below to request access.

Request Information

Experimental Details

Interventions

Intervention(s)

In this study, we aim to test whether training in empathy affects civil service quality, measured in terms of social preferences in experimental games, performance on the job, and thought leadership footprint.

Intervention Start Date

2020-10-31

Intervention End Date

2021-08-20

Primary Outcomes

Primary Outcomes (end points)

Job performance
Update March 13, 2023: We are collecting data on the universe of spending decisions by the civil servants as they have now spent about 2 years in the field post-intervention. This data will become available through performance audits of deputy ministers which contain information on their spending on various government departments. There are 12 government departments on which they can choose to recommend a budgetary allocation. These 12 policy outcomes are:
1) health,
2) education,
3) personal security,
4) office budget,
5) IT,
6) development,
7) horticulture,
8) sports,
9) entertainment,
10) law and order,
11) climate encroachment,
12) disaster.

Primary Outcomes (explanation)

Main Hypotheses. —In our pre-analysis, we outline two main categories of policies that are most likely to be impacted by altruism training. For instance, we expect social policies, i.e. spending on health and education departments, to be positively impacted and selfish (self-oriented) policies, i.e. personal security and office budget, to be negatively affected (as they potentially crowd-out government funds) or unaffected (if government budget is not binding and there is no crowd-out). In the case of crowd-out, these selfish policies would be negatively impacted by increased value of altruism. Our main hypothesis, therefore, is that spending on social policies will increase and spending on selfish policies will decrease, while the other 8 policies will be unaffected.

Main Hypotheses. —In our pre-analysis, we outline two main categories of policies that are most likely to be impacted by altruism training. For instance, we expect social policies, i.e. spending on health and education departments, to be positively impacted and selfish (self-oriented) policies, i.e. personal security and office budget, to be negatively affected (as they potentially crowd-out government funds) or unaffected (if government budget is not binding and there is no crowd-out). In the case of crowd-out, these selfish policies would be negatively impacted by increased value of altruism. Our main hypothesis, therefore, is that spending on social policies will increase and spending on selfish policies will decrease, while the other 8 policies will be unaffected.

Secondary Outcomes

Secondary Outcomes (end points)

Secondary Outcomes (explanation)

Experimental Design

Lectures are delivered over zoom and randomized at the individual level. Factorial design: empathy malleable vs. empathy utilitarian vs. both vs. control (focusing on leadership) in online lectures. Class-wide lectures will also be randomized, focusing on emotional intelligence vs. macro-economics lectures.

Update March 13, 2023: We now also are collecting data on universe of spending decisions by the civil servants. Specificially, the study experimentally evaluates three methods of cultivating prosocial behaviour (i) Utilitarian Value of Empathy (ii) Malleability of Empathy (iii) Joint Utilitarian and Malleability Treatment (iv) Placebo training.

We outline two main categories of policies that are most likely to be impacted by altruism training. For instance, we expect social policies, i.e. spending on health and education departments, to be positively impacted and self-oriented policies, i.e. personal security and office budget, to be negatively affected (as they potentially crowd-out government funds). Our main hypothesis, therefore, is that spending on social policies will increase and spending on selfish policies will decrease, while the other 8 policies will be unaffected. For more details, see the pre-analysis plan.

Experimental Design Details

Randomization Method

Randomization done by computer

Randomization Unit

Individual

Was the treatment clustered?

Yes

Experiment Characteristics

Sample size: planned number of clusters

200 civil servants

Sample size: planned number of observations

200 civil servants

Sample size (or number of clusters) by treatment arms

50 in treatment arms

Minimum detectable effect size for main outcomes (accounting for sample design and clustering)

A single sided power test using as the outcome variable of interest from the pilot of empathy in a list experiment with a mean of 2.319 using one treatment group with 100 observations against a control group with 100 observations at a power .90 and a standard deviation of .920 shows that a minimum detectable effect would be a change of .230 of the standard deviation. The new minimum detectable mean would need to reach 2.53 to be detected.

Supporting Documents and Materials

IRB