Featured:

3-2

Turn AI anxiety into action

Watch now →

Featured:

Mega menu – Featured Image Template (1)

Bolstering technical skills development in MENA

Read the case study →

Featured:

Christian Martinez 4.19-thumb-1

Explore new digital career pathways

Read more →

Featured:

2-Sep-04-2024-01-24-22-6638-PM

Take action on generative AI

Get my copy →

Featured:

1-Sep-04-2024-05-44-06-9704-PM

Hear what the experts have to say

Tune in →
Tech for Jobs — Jordan

1,751 Jordanians in new or better work — and $10M a year in new household earnings.

 

Jordan's largest employment program of its kind, made possible by USAID and the U.S. Department of State, whose commitment carried a national-scale intervention across every one of the country's twelve governorates. Built from the outset with the Abdul Latif Jameel Poverty Action Lab (J-PAL) MENA as a randomized controlled trial, designed to answer what participants earn, where they end up working, and how much of that is attributable to the program itself.

Outcomes ledger OUTCOMES STILL ACCRUING
1,751
Participants in new or better work — 1,297 new jobs, 454 improved
TARGET 2,160 BY SEPT 2026
~$10M
Additional earnings per year into Jordanian households, across 1,540 placements
FROM 536 PAIRED SALARY RECORDS
74%
Of placements went to participants who had no work at baseline
1,297 OUT OF 1,751
53%
Women among enrolled participants — 2,300 of 4,360
TARGET WAS 50%
How it was built

A three-arm randomized evaluation, specified with J-PAL MENA before applications opened.

 

Jordan produces more qualified graduates than its domestic labor market absorbs. The constraint facing educated youth is demand. Tech for Jobs tests whether skills training combined with active placement support moves people into paid work against that constraint.

Two treatment arms sit against a pure control group. One arm receives technical training, career coaching, and job matching; the other receives everything except the technical training. That structure identifies the marginal contribution of the technical component specifically, rather than the effect of the package as a whole. Randomization is at the individual level, post-baseline, stratified by gender.

Standard Training Model

  • Program built first, evaluation added later if budget allows
  • Everyone who enrolls receives the same thing
  • No comparison group, so no counterfactual
  • Success measured at course completion
  • Outcomes estimated from whoever answers a survey
  • Reporting stops when the training stops

Tech for Jobs

  • Randomized design fixed with J-PAL MENA before applications opened
  • Two distinct interventions, individually randomized
  • A pure control group, specified before enrollment began
  • Success measured at verified employment and wages
  • Every participant tracked individually and repeatedly
  • Tracking continues 18 months past the final class
 
Results

1,751 people in new or better work, and 189 employers that hired them.

 

Every figure below is a count of named individuals, verified through employment surveys issued to the full participant population and checked against each person's employment status at application. Figures are current to June 2026 and continue to rise through the end of the program.

Measure Women Men Total
Enrolled Selected from 7,231 qualified applicants, across all 12 governorates 2,300 2,044 4,360 53% women — target was 50%
Graduated Completed the attendance and assignment requirements for their track 1,868 1,708 3,587 Against a target of 3,600 by September 2026
New or better employment 1,297 participants who had no work at baseline, and 454 who moved up from the job they had 847 895 1,751 Against a target of 2,160 by September 2026
EMPLOYER DEMAND

Thirty-two American companies have hired Tech for Jobs participants, among 189 distinct employers in total against a life-of-project target of 50. They span software and customer experience, cloud infrastructure, defense engineering, advanced manufacturing, and frontier AI — including Fortune 500 firms and AI-native companies. Those are the same sectors Jordan has named as priorities under its Economic Modernisation Vision, which means the hiring is building domestic capability in the areas the country has chosen to grow, not drawing talent away from them. 

Measurement

The RCT study design

 

A three-arm randomized evaluation of an active labor market program connecting young Jordanians to offshore data and technology roles, fielded with the Abdul Latif Jameel Poverty Action Lab (J-PAL) MENA.

Design

T1

Full intervention

Technical training Career coaching Job matching
T2

Reduced intervention

Technical training Career coaching Job matching
C

Control

Technical training Career coaching Job matching
Sample
6,199 enrolled applicants, stratified by gender
Assignment
Individual-level randomization, post-baseline
Tracking
Baseline → end of training → 12-month → 24-month phone surveys

Primary outcomes

Employment

Any paid work in the past 7 days, ILO definition

Earnings

Monthly net earnings, IHS-transformed

Secondary outcomes

  1. 01Job quality
  2. 02Subjective wellbeing
  3. 03Gender attitudes
  4. 04Intra-household decision-making
  5. 05Reservation wages
  6. 06Reservation working conditions
  7. 07Remote and flexible work
  8. 08ICT occupations

Contribution to the literature

This evaluation tests whether hands-on data analytics skills paired with offshore remote placement can unlock good jobs that local labor markets cannot supply. It directly addresses the binding demand constraint facing educated youth in Jordan, and offers a replicable model for economies where local job creation lags local human capital.

Hypotheses (pre-specified)

H1

Any treatment improves employment, earnings, and formality

H2

Full treatment (T1) outperforms reduced treatment (T2)

H3

Effects are larger for women than for men

6,199
Randomized sample
3
Arms — T1 · T2 · Control
24mo
Endline horizon
8–11pp
Minimum detectable effect
Lessons learned

Placement rates were not evenly distributed. The gaps are the most useful thing in the data.

 

Disaggregating outcomes by gender, geography, and age exposes three patterns that a headline placement rate conceals entirely — and each one points at labor market access rather than at capability.

Finding 01

Women advanced faster than men, but entered more slowly

Among participants who arrived already employed, 56.9% of women moved to better work against 53.4% of men. Among those who arrived unemployed, 42.6% of women found work against 52.0% of men — a 9.4 point gap in the opposite direction. The barrier is at the point of entry, not at the point of performance.

What changes

Direct employer matching is being weighted toward women entering the labor market for the first time, where the open channel performs worst.

Finding 02

Remote-first training reached everyone. Remote-first hiring can too

Among participants unemployed at baseline, 50.6% of those living in Amman found work, against 42.2% of those elsewhere in Jordan. Delivery already proved that training reaches every governorate equally well. What has not yet followed is the matching — connecting participants outside the capital to employers willing to hire them where they are. Remote roles are what decouple where a person lives from where they can work.

What changes

Employer outreach now prioritizes fully remote roles, extending to placement the reach that delivery already achieved.

Finding 03

Participants over 30 placed at two-thirds the rate of those under 30

50.1% of participants aged 18 to 29 who arrived unemployed found work, against 32.4% of those aged 30 to 49. The same age pattern appeared independently in Correlation One's cybersecurity program in Jordan and Egypt, which suggests a regional hiring norm rather than a program-specific effect.

What changes

Mid-career participants are being routed to employers with demonstrated senior hiring, and away from entry-level pipelines that screen on age.

Learner testimonies

Four of the 1,751.

 

Selected to show the range of what an outcome looks like in this program: a first professional role, a career change out of a different field entirely, a promotion inside the same company, and a placement with a global firm.

New employment · Full intervention

Ahmad N.

Data and Research Assistant, United Nations agency

Ahmad completed the full technical track and moved into his first professional role in August 2025, working on data and research inside a UN agency. It is as close as an individual outcome gets to the purpose the program was funded to serve.

No prior professional role → Data and research, multilateral agency
Career change · Full intervention

Maya A.

Data Conversion Specialist, enterprise technology firm

Maya trained as a pharmacist and was tutoring freelance when she enrolled. She completed the full technical track, went on to a master's in data science, and now works in data conversion at an enterprise technology firm in Amman. Pharmacy to production data work inside two years.

Pharmacy graduate, freelance tutor → Data specialist, enterprise tech
Better employment · Reduced intervention

Nadine D.

Insights Manager, analytics firm

Nadine had been an insights analyst at a London-based analytics company for close to seven years. She enrolled in the reduced arm — career coaching and job matching, without the technical track — and was promoted to Insights Manager in February 2025. Same employer, better title, nothing else in her circumstances changed.

Insights Analyst → Insights Manager, same firm
New employment · Full intervention

Leen H.

Consultant, global professional services firm

Leen completed the full technical track and was hired as a consultant at one of the large global professional services firms in August 2025. That is a demanding screen to clear, and hers is one of several outcomes in this cohort that landed with international employers rather than domestic ones.

Recent graduate → Consultant, global firm
Scale

What it takes to move four thousand people into work at once.

 

The largest line in the delivery model is not instruction. It is one-to-one time: 751 individual coaching sessions covering job search strategy, CV work, and interview preparation, running through training and continuing well past it.

 

12

Governorates reached — the whole of Jordan, not the capital alone

751

One-to-one career coaching sessions delivered to participants

102

Live technical and professional development sessions

500+

Alumni and staff activated as job referrers for their peers



How coaching stays current

Coaching is not generic employability advice, and it is not written once. Correlation One's employer engagement team works directly with hiring managers to establish what each open role actually requires — the stack, the seniority, the screening criteria — and feeds that back into the coaching sessions. A participant preparing for an interview is prepared against the requirements that employer has stated, in that hiring round. The same intelligence determines which three to five candidates are referred for each role, so the coaching and the matching run off one shared picture of demand rather than two.

FAQ

Everything you need to know.

 
What is Tech for Jobs? Tech for Jobs is the largest employment program of its kind in Jordan, implemented by Correlation One and made possible by USAID and the U.S. Department of State. It combines data analytics training, professional development, individual career coaching, and job matching for job seekers aged 18 to 40 across all twelve governorates. 4,360 people enrolled in October 2024 and 3,587 have graduated. The enrolled cohort was 53% women, against a target of 50%.
 What employment outcomes has the program achieved? As of 30 June 2026, 1,751 participants had gained new or better employment: 1,297 who had no job at baseline are now working — 74% of all placements — and 454 who were already working moved to better roles through higher pay, greater seniority, or improved conditions. Graduates already in the workforce saw an average wage gain of 30%. At the 1,540-placement milestone in May 2026, the program was adding an estimated $10 million a year in household earnings to Jordan's economy, or roughly $520 more per graduate per month. 189 distinct employers have hired a participant against a life-of-project target of 50, 32 of them American companies. The employment target is 2,160 outcomes by the end of the program period in September 2026.
 How were employment outcomes and earnings measured? Across the full participant population rather than a sample. A structured employment survey went to all 4,360 participants capturing status, role, salary, contract type, and start date, and has been re-run three times — November 2025, March 2026, and July 2026. Each participant's employment status at the point of application is on file, so a new job can be distinguished from an improved one against that individual's own starting position rather than a cohort average. Wage figures rest on 536 paired pre- and post-program salary observations.
 Was Tech for Jobs designed as a randomized trial? Yes. Correlation One co-designed the program with the Abdul Latif Jameel Poverty Action Lab (J-PAL) MENA as a three-arm randomized controlled trial. From 7,231 qualified applicants, 6,199 completed a baseline survey and were individually randomized post-baseline, stratified by gender, into a full intervention arm (technical training, career coaching, job matching), a reduced intervention arm (career coaching and job matching only), and a pure control group. The two-arm treatment structure identifies the marginal contribution of the technical training specifically, rather than the effect of the package as a whole. Primary outcomes are employment, defined as any paid work in the past seven days under the ILO definition, and monthly net earnings. Tracking runs from baseline through end of training to 12-month and 24-month phone surveys.
 Who was eligible, and how were participants selected? Eligibility required residency in Jordan, an age between 18 and 40, working English proficiency, availability for a 16-week training program, and interest in full-time roles requiring data analytics skills. A bachelor's degree in any subject qualified; applicants without a degree who scored in the top 40th percentile on the technical assessment were also considered. A national campaign produced 15,525 applications — 50% from women and 46% from outside Amman — of which 11,291 completed a full application and technical assessment, and 7,231 were identified as top qualified on assessment and interview performance.
 Did the program benefit all participants equally? No, and the gaps are instructive. Among participants who arrived already employed, women outperformed men — 56.9% moved to better work against 53.4% of men. Among those who arrived unemployed, the pattern reversed: 42.6% of women found work against 52.0% of men. Participants living in Amman placed at 50.6% against 42.2% for those elsewhere in Jordan, and participants aged 18 to 29 placed at 50.1% against 32.4% for those aged 30 to 49. All three gaps point to labor market access rather than participant capability, and each is changing how placement support is targeted.
 Who funded Tech for Jobs?Who funded Tech for Jobs? Tech for Jobs is made possible by the United States government, through a cooperative agreement awarded by USAID and subsequently administered by the U.S. Department of State, Bureau of Near Eastern Affairs. That backing is what allowed a single intervention to run at national scale across all twelve governorates simultaneously. The performance period runs from May 2024 to September 2026.
 Can this model be delivered in another country? Yes. The design is country-agnostic: aptitude-based selection rather than credential-based, remote-first delivery that reaches participants outside the capital, employer engagement running in parallel with training rather than after it, and individual verification of every outcome. The binding requirements are a recruitment network capable of reaching qualified candidates at national scale and an employer base with real hiring demand in the target labor market. Correlation One has delivered comparable programs in Jordan, Egypt, Colombia, and across Latin America.
THE FULL EVALUATION · MINDSET, 2026

Placeholder

 Download the independent endline evaluation Mindset prepared for Correlation One: 107 participants surveyed six to seven months after the program, with employment, earnings, job-quality, and wellbeing outcomes broken out by country, age, and prior experience — methodology and limitations stated in full. 
Sponsored by the US

  • Tech for Jobs is funded by the United States Government through the U.S. Agency for International Development and the U.S. Department of State.

    The contents of this page are the responsibility of Correlation One and do not necessarily reflect the views of the United States Government.

Bring the model to your region.

Tech for Jobs was built for one national labor market and one skill set. The design — aptitude-based selection, randomized measurement, placement support running alongside training, and individual verification of every outcome — transfers to others.