Motivating Bureaucrats: Autonomy vs Performance Pay for Public Procurement in Pakistan

Last registered on February 10, 2017

View Trial History

Pre-Trial

Trial Information

General Information

Title

Motivating Bureaucrats: Autonomy vs Performance Pay for Public Procurement in Pakistan

RCT ID

AEARCTR-0000610

Initial registration date

January 24, 2015

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

January 24, 2015, 12:09 AM EST

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

Last updated

February 10, 2017, 6:44 PM EST

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

Locations

Country

Pakistan

Region

Punjab

Primary Investigator

Name

Michael Best

Affiliation

Columbia University

Contact Primary Investigator

Other Primary Investigator(s)

PI Name

Oriana Bandiera

PI Affiliation

London School of Economics

Contact Investigator

PI Name

Adnan Khan

PI Affiliation

London School of Economics

Contact Investigator

PI Name

Andrea Prat

PI Affiliation

Columbia University

Contact Investigator

Additional Trial Information

Status

On going

Start date

2014-07-01

End date

2017-04-30

Keywords

Governance, Labor

Additional Keywords

JEL code(s)

Secondary IDs

Abstract

This project works directly with the government of Punjab, Pakistan to improve the effectiveness of public procurement. We develop two novel methods to measure procurement effectiveness and leakages based on detailed, item-level characteristics data and random field audits of purchases. We then implement a field experiment to study how these are affected by two policy reforms targeting bureaucrats’ extrinsic and intrinsic motivations to achieve value for money in procurement. The first treatment awards financial bonuses for good procurement performance. The second treatment empowers bureaucrats to exercise discretion in procurement to achieve better value for money by giving them a larger cash-in-hand budget, making pre-audit procedures simpler and more predictable, and releasing their procurement budget earlier. A third treatment combines the two to identify complementarities.

External Link(s)

Registration Citation

Citation

Bandiera, Oriana et al. 2017. "Motivating Bureaucrats: Autonomy vs Performance Pay for Public Procurement in Pakistan." AEA RCT Registry. February 10. https://doi.org/10.1257/rct.610-4.0

Former Citation

Bandiera, Oriana et al. 2017. "Motivating Bureaucrats: Autonomy vs Performance Pay for Public Procurement in Pakistan." AEA RCT Registry. February 10. https://www.socialscienceregistry.org/trials/610/history/13916

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)

Intervention Start Date

2015-07-01

Intervention End Date

2016-06-30

Primary Outcomes

Primary Outcomes (end points)

Our main outcome variables are twofold.
1-Quality-adjusted prices paid
2-Quality, quantity and price discrepancies between products procured as self-reported and as revealed through audit.

Primary Outcomes (explanation)

Quality adjusted prices will be constructed from extremely detailed data on the goods being purchased, with a focus on “generic” goods where quality can be determined by using a relatively small number of attributes of a good. With this detailed data, we can adjust prices paid for the quality of the goods purchased to uncover differences in the prices different procurement officers pay for the same good. We are collecting this data through an online portal, the Punjab Online Procurement System (POPS), designed and maintained by us, into which procurement officers enter the full set of details of every purchase of a generic good they make.

Secondary Outcomes

Secondary Outcomes (end points)

Secondary Outcomes (explanation)

Experimental Design

Drawing and Disbursing Officers (the government officers legally responsible for public procurement) are randomly allocated to 1 of 5 treatments, stratifying by district and by line department (Health, Higher Education, Communication & Works, Agriculture). Treatment arm 1: Financial bonuses for good procurement performance; data collected through POPS. Treatment arm 2: Increased discretion for bureaucrats; data collected through POPS. Treatment arm 3: Combined treatments 1 & 2. Treatment arm 4: Control group; data collected through POPS. Treatment arm 5: Pure control group; data not collected through POPS, but through field visits.

Experimental Design Details

Randomization Method

Computer program

Randomization Unit

Drawing and Disbursing Officer (the bureaucrat with the legal responsibility and authority for conducting public procurement)

Was the treatment clustered?

Yes

Experiment Characteristics

Sample size: planned number of clusters

1074 Drawing and Disbursing Officers (DDOs)

Sample size: planned number of observations

Approx 1074*50=53,700

Sample size (or number of clusters) by treatment arms

202 in Incentive payment arm
203 in Discretion arm
228 in Combined treatments arm
320 in Control group
121 in Pure Control group

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

Using Bloom's (2005) expression for the MDE, our estimate of the intracluster correlation from baseline data of 0.067, and an estimate of an average of 75 observations per bureaucrat, we estimate that a sample of 572 bureaucrats, or 42,882 observations, will allow us to detect a 5% effect with 95% probability.

Supporting Documents and Materials

IRB