Civil Service Decision-making. Do advisor and advisee gender matter?

How to cite

Dissanayake, Ranil; Serneels, Pieter; Musillo, Benedetta. 2022. "Civil Service Decision-making. Do advisor and advisee gender matter?." AEA RCT Registry. AEARCTR-0010391. https://doi.org/10.1257/rct.10391-2.1

Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.

Ranil Dissanayake* , Pieter Serneels , Benedetta Musillo

* Primary investigator

Registry ID
AEARCTR-0010391
DOI
https://doi.org/10.1257/rct.10391-2.1
First registered
Last updated
Status
In development
Trial dates
2022-11-14 – 2023-11-30
Countries
Private
JEL codes
J16, D73, D91
Registry record
http://www.socialscienceregistry.org/trials/10391

Abstract

Under-representation by gender or other characteristics is receiving increased attention, both in policy-making and research (Allgood et al. 2019). Diversity is seen as highly desirable, for reasons of fairness and representation, and because it brings with it more varied perspectives, which in turn enhances innovation (Ann Hewlett, Marshall, and Sherbin 2013; Heaton and Aminossehe 2020). It can improve the quality of policymaking on the supply side. Constraints to diversity on the demand side, and their consequences, have received less attention. We investigate whether such constraints on the demand side exist, and need addressing to realize the benefits of diversity. Using a survey experiment we assess whether perceived, quality, valuation, and take-up of professional advice depends on the identity of the provider, and test and explore channels through which this works. We also investigate whether the supply of advice depends on the gender of the commissioning party.

Our setting is a UK Government department that represent a high-stakes policy environment. We use experimentally varied survey vignettes to investigate if gender of the involved parties matters for civil service decision making. Specifically, we examine if decisions made on the basis of new information and analysis depends on the identity of the source providing this information, as well as or instead of its perceived quality. We investigate two sets of questions. Firstly, do decisions made in the organization vary according to the gender of technical advisers? Secondly, if there is such a bias, is it mitigated by the style and type of advice given, specifically if advice is more or less conscientious and detailed?

Experimental design

We implement a randomized survey experiment, using vignettes designed to investigate if decisions made on the basis of new information and analysis depends on the identity of the source providing this information, as well as or instead of its perceived quality.

Design details

Randomization method
Randomization done in office by a computer
Randomization unit
Individual
Clusters
3000 individuals
Observations
3000 individuals
Arms
750 per arm
Primary outcomes
Demand side analysis: - Likelihood of take up, measured by updating of decisions: We measure how the response to the question “How likely are you to make a medium-sized investment in this programme?” changes after the receipt of advice from an adviser (gender and conscientiousness of the advice randomized), The question is answered on a 4 point scale (Very unlike to very Likely), and updating is measured on a 7 point scale where 0 is no change -3 is shifting three spots towards less likely, and 3 is a three point increase in likeliness). - Likelihood of take up, measured by updating of confidence in decision: We measure how the response to the follow-up question “How confident are you in this decision” changes after the receipt of advice from an adviser (gender and conscientiousness of the advice randomized), The question is answered on a 5 point scale (Very unlike to very Likely, with a neutral option), and updating is measured on a 8 point scale where 0 is no change -4 is shifting four spots towards less confident, and 4 is a four point increase in confidence). - Willingness to pay for advice: respondents are asked how much they would pay for the advice provided, had it been commissioned from a paid external adviser (0-10000). - Perceived advice quality, measured on a 0-10 scale. Supply side analysis: - Confidence in offering advice: in our supply of advice vignette, we ask respondents how confident they feel to volunteer a response to a question made by their management board. This is measured on a 1-5 scale, from not confident at all to very confident, with a neutral option.

Cited by

No citation data available for this registration yet.

Keywords