The Impacts of a Pilot-Run Virtual Job Fair for Recent College Graduates in Bangladesh

How to cite

Hayashi, Ryotaro. 2021. "The Impacts of a Pilot-Run Virtual Job Fair for Recent College Graduates in Bangladesh." AEA RCT Registry. AEARCTR-0007294. https://doi.org/10.1257/rct.7294-1.0

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

Ryotaro Hayashi*

* Primary investigator

Registry ID
AEARCTR-0007294
DOI
https://doi.org/10.1257/rct.7294-1.0
First registered
Last updated
Status
On going
Trial dates
2020-11-25 – 2021-06-30
Countries
Bangladesh ()
Registry record
http://www.socialscienceregistry.org/trials/7294

Abstract

We conduct a pilot run of a virtual job fair in Bangladesh. The fair is organized for recent graduates from a large public agricultural university and participated by agro-industry firms. During the fair, each participating firm holds a live video session in which the firm explains its business and job openings and answers questions from participating jobseekers. Job applications and interviews are made through the online platform of the fair. This RCT adopts an oversubscription design: we first ask graduates if they are interested in the fair and then randomly select a half of the interested graduates to invite to the fair. We collect baseline and follow-up survey data of the college graduates to estimate the impacts of the fair on their employment outcomes.

Intervention

The intervention is a virtual job fair that is organized for recent graduates from a large public agricultural university and participated by agro-industry firms. Each participating firm holds a live video session in which the firm explains its business and job openings and answers questions from jobseekers. On the virtual job fair website, participating jobseekers can view and apply for participating firms’ job openings.

Adopting an oversubscription design, we randomize the intervention across the graduates who expressed the interest in the fair. Jobseekers in the treatment group are invited to the fair, while control-group jobseekers are not.

Experimental design

1. We contact recent graduates over the phone and ask if they are interested in the virtual job fair and also if they have access to an electronic device, such as computer and smartphone, and the internet that are necessary to join the fair. A total of 1248 graduates express their interest and have the necessary digital environment. (1255 graduates are interested, only 7 of whom do not have the digital device or the internet.)

2. We conduct a baseline survey of all the 1248 interested graduates over the phone. The baseline survey interview is conducted in the same call as the recruitment call (step 1): that is, if a graduate is interested and has the necessary digital environment, an enumerator continues to ask the baseline survey questions.

3. The 1248 interested graduates are randomly assigned to the treatment and control groups. Treatment graduates are invited to the fair, while control graduates are not. The randomization is at the individual level. There are 625 and 623 graduates in the treatment and control groups, respectively.

4. The job fair’s organizer, which is a firm that runs an online job portal, opens a virtual job fair website. On the website, jobseekers can view participating firms’ profiles and job advertisements and apply for jobs. Twenty-four agro-industry firms participate and advertise a total of 38 positions for the fair.

5. Of the 625 graduates in the treatment group, 283 sign up for the job fair.

6. Approximately three months after the job fair, we conduct a follow-up survey of all the 1248 graduates over the phone.

Design details

Randomization method
Randomization done in Stata. The randomization is at the jobseeker level and stratified by (i) gender, (ii) geographical division, (iii) type of place of residence (i.e., urban or rural), and (iv) whether a jobseeker has never used any online job portal, or has used the online job portal of the job fair’s organizer, or has used online job portals other than the organizer’s job portal.
Randomization unit
The unit of randomization is individual.
Clusters
1248 jobseekers
Observations
1248 jobseekers
Arms
625 and 623 jobseekers in the treatment and control groups, respectively.
Primary outcomes
Note: The letters and numbers in parentheses indicate the questions in the follow-up questionnaire based on which the outcome variables will be constructed. - Being employed (C1) - Being employed for a permanent full-time job (C2, C3) - Earnings (C4)
Secondary outcomes
Notes. (1) Since these outcomes are secondary and intend to understand additional and/or detailed effects, multiple hypothesis testing adjustment will be conducted separately for the primary and the secondary outcomes. (2) Analyses on the outcome variables for those who are employed and those who are unemployed will use the subsamples of the employed and the unemployed, respectively. Hence, the analyses will be descriptive. [All graduates in the sample] - Whether looking for a job in last 4 weeks (D2) - Reservation wage and wage aspiration (D7, D8) - Life satisfaction (D9) [For those who are employed] - Formal employment (being employed on the basis of a written contract, C6) - Job satisfaction (C11) - Job match quality (C9 and C10) [For those who are unemployed] - Self-perceived job search prospects: likelihood of getting a job in next 6 months (D5); the expected number of weeks until having a permanent job (D6)

Cited by

No citation data available for this registration yet.

Keywords