Does revealing salary expectations affect employment prospects?

How to cite

Falco, Paolo; Gatti, Roberta; Islam, Asif. 2025. "Does revealing salary expectations affect employment prospects?." AEA RCT Registry. AEARCTR-0016692. https://doi.org/10.1257/rct.16692-1.0

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

Paolo Falco* , Roberta Gatti , Asif Islam

* Primary investigator

Registry ID
AEARCTR-0016692
DOI
https://doi.org/10.1257/rct.16692-1.0
First registered
Last updated
Status
In development
Trial dates
2025-09-08 – 2025-11-30
Countries
Private
JEL codes
C93, D83, J31, J71, M51
Registry record
http://www.socialscienceregistry.org/trials/16692

Abstract

We conduct a correspondence (CV) study in Egypt to test whether revealing salary expectations affects employers’ responses. We create four Egyptian-sounding applicant identities (two male, two female) and cross them with three salary expectation conditions: (i) no expectations stated, (ii) “low” expectation, and (iii) “high” expectation—yielding 12 arms. We apply once to each vacancy harvested from the main job platforms operating in Egypt during a two-week window and active at the start of the intervention. All CVs are standardized for qualifications, experience, and application materials; only the name (gender signal) and the salary-expectation line vary. The primary outcome is employer callback, tracked via dedicated job-portal accounts, unique email addresses, and phone numbers. We plan an estimation comparing callback rates across arms and testing whether stating expectations—and their level—affects employer engagement overall and by gender signal. We will explore pre-specified heterogeneity by employer characteristics (sector, customer interaction, posted salary, and proxies for competitive pressure).

Experimental design

Setting and sampling: We scrape vacancies from the main job platforms operating in Egypt over a two-week period. We apply only to postings verified as active at launch and send at most one application per vacancy.

Treatments: 4 applicant identities (two male, two female) × 3 expectation conditions (None, Low Expected Monthly Salary, High Expected Monthly Salary) = 12 arms, assigned in equal proportions. Names are common Egyptian names; CV content and cover text are otherwise identical.

Implementation: Applications are submitted through job-portal accounts created for the study. Each identity has a dedicated email and phone number. We standardize sending times and avoid repeat contacts with the same employer.

Outcomes and tracking: Callbacks are captured from (a) job-portal inboxes, (b) dedicated email accounts, and (c) phone logs (missed calls/voicemails/WhatsApp if applicable).

Analysis: Comparisons of callback rates across arms using OLS or logit with robust SEs. We will report pooled “any expectations” vs. “none,” low vs. high expectations, and interactions with gender signal. Pre-specified heterogeneity analyses use employer characteristics collected from postings. No multiple applications to the same vacancy.

Design details

Randomization method
Random lottery using a pseudo-random number generator in code to assign vacancies in equal proportions to the 12 arms.
Randomization unit
Individual vacancy. Each active vacancy is randomly assigned to exactly one of the 12 treatment variations in equal numbers, with one application per vacancy.
Clusters
N/A
Observations
We have approximately 7,500 vacancies that have been harvested and will apply to those that are still active at launch. We anticipate at least 3,600–4,800 applications.
Arms
Approximately 300–400 vacancies per arm (12 arms), totaling 3,600–4,800 applications.
Primary outcomes
Callback (binary): Any employer-initiated positive response within the follow-up window (e.g., interview/test invitation, request for additional information) recorded via job-portal messages, dedicated email, or phone. Ancillary: Type of callback (interview/test vs. information request) and time-to-first-callback.
Secondary outcomes
Heterogeneity in treatment effects by employer characteristics: sector/industry, whether the job is customer-facing, posted salary level/band (if available), and a proxy for market competition faced by the firm (e.g., industry concentration or within-platform posting density). Additional non-primary measures: time-to-callback; number of contact attempts; message “positivity” (interview/test vs. generic reply).

Cited by

No citation data available for this registration yet.

Keywords