FINAL REPORT FOR THE 
WORKFORCE DEVELOPMENT 
STUDY 
Submitted April 1, 2021 
This document was produced for review by the United States Agency for International 
Development El Salvador Mission (USAID/El Salvador) by Carlos Carcach and Gustavo Quan, 
consultants of ME&A.
FINAL REPORT FOR THE WORKFORCE 
DEVELOPMENT STUDY
Submitted: April 1, 2021
Contract Number: AID-OAA-I-15-00024/AID-519-TO-16-00002 
Activity Start Date and End Date: August 1, 2016 – July 31, 2021
Contracting Officer’s Representative (COR): Annie de Valencia, avalencia@usaid.gov
Submitted by: Randal Joy Thompson, Chief of Party
ME&A 
4350 East-West Highway, Suite 210
Bethesda, MD 20814 USA
Email: rthompson@engl.com
CONTENTS 
Executive Summary ........................................................................................................................................... i
OBJECTIVES OF THE STUDY, PURPOSE AND QUESTIONS......................................................... i
PROJECT BACKGROUND ....................................................................................................................... i
DATA, METHODS AND LIMITATIONS ..............................................................................................ii
FINDINGS, CONCLUSIONS AND RECOMMENDATIONS .........................................................iv
1.0 Study Purpose and Questions........................................................................................................... 1
1.1 STUDY PURPOSE ........................................................................................................................... 1
1.2 RESEARCH QUESTIONS.............................................................................................................. 1
2.0 Project Background................................................................................................................................... 1
2.1 PROJECT CONTEXT..................................................................................................................... 1
2.2 PRELIMINARY FINDINGS FROM THE SISPUENTES DATA MANAGEMENT SYSTEM
 …………………………………………………………………………………………...2
3.0 Data, methods and limitations................................................................................................................ 3
3.1 DATA ................................................................................................................................................. 3
3.2 METHODS ........................................................................................................................................ 3
Assessing the factors affecting the likelihood of employment among BE participants ............................ 3
Assessing the factors affecting the salaries of the BE participants............................................................ 3
3.3 LIMITATIONS .................................................................................................................................. 4
4.0 Findings and conclusions .......................................................................................................................... 4
4.1 FINDINGS ......................................................................................................................................... 4
4.1.1 Q1: By key sectors, how can participants’ specific technical, academic and vocational 
skills (alone and/or combined) be most effectively leveraged for being hired into jobs above 
the minimum wage?.................................................................................................................................. 4
4.1.2 Q3: By sector, how much does having a 1) high school degree or 2) at least some 
high school education contribute to employment? Is there a wage differential?....................... 5
4.1.3 Q4: What data knowledge discovery was identified by sector and beneficiary profiles?
...................................................................................................................................................................... 7
4.1.4 Q5: How do formal or informal minimum academic requirements by the private 
sector affect the potential for youth to be employed? Based on the data at hand is there a 
minimum academic level required by the private sector to employ youth?............................... 7
4.1.5 Q6: How do formal or informal minimum academic requirements by the private 
sector affect the potential for youth to be employed? Based on the data at hand is there a 
minimum academic level required by the private sector to employ youth?............................... 8
4.2 CONCLUSIONS.............................................................................................................................. 9
4.2.1 Q1 Conclusion................................................................................................................................ 9
4.2.2 Q3 Conclusion................................................................................................................................ 9
4.2.3 Q4 Conclusion................................................................................................................................ 9
4.2.4 Q5 Conclusion................................................................................................................................ 9
4.2.5 Q6 Conclusion............................................................................................................................. 10
4.2.6 Q6 Recommendations - How might workforce development programs complement 
this potential? ......................................................................................................................................... 10
APPENDICES .................................................................................................................................................. 11
APPENDIX I: STATEMENT OF WORK (SOW) .............................................................................. 12
APPENDIX 2: TABLES AND GRAPHS ............................................................................................... 20
APPENDIX 3: BIBLIOGRAPHY OF DOCUMENTS REVIEWED.................................................. 23
LIST OF TABLES
Table A1: Summary of Findings and Conclusions iv 
Table A2: Recommendations viii
Table 1: MARS Basis Functions and Their Coefficients – Model for the Logarithm of Salary 5
Table 2: Logistic Regression Models Coefficients for the Likelihood of Employment by Economic 
Sector.................................................................................................................................................................. 6
ACRONYMS
APE Average Partial Effect
Bridges Bridges to Employment Project
CDCS USAID Country Development Strategy
INSAFORP Spanish acronym for Instituto Salvadoreño de Formación Profesional (Salvadoran
Institute for Professional Training)
MARS Multivariate Adaptive Regression Splines 
SAS/STAT A software for data analysis
USAID United States Agency for International Development
i 
Executive Summary
OBJECTIVES OF THE STUDY, PURPOSE AND QUESTIONS
The purpose of this study is to conduct a statistical analysis of the data collected under the United 
States Agency for International Development (USAID) Bridges to Employment (Bridges) Project; 
to provide evidence and identify trends to inform future interventions under the newly approved 
Country Development Cooperation Strategy (CDCS); and to identify beneficiary professional 
and academic profiles in the key business sectors of agroindustry, retail trade and tourism, energy, 
manufacturing, and information technology.
The study seeks to answer the following research questions:
1. By key sectors, how can participants’ specific technical, academic, and vocational skills 
(alone and/or combined) be most effectively leveraged for being hired into jobs above the 
minimum wage?
2. How do sector-specific academic and skills requirements act as a barrier to either men 
or women looking for employment typically sought by Bridges participants? 
3. As specifically as possible, by sector, how much does having a 1) high school degree or 
2) at least some high school education contribute to employment? Is there a wage 
differential? 
4. What data knowledge discovery was identified by sector and beneficiary profiles?
5. How do formal1 or informal2 minimum academic requirements by the private sector affect 
the potential for youth to be employed? Based on the data at hand, is there a minimum 
academic level required by the private sector to employ youth?
6. Based on the data at hand, which economic sectors would be the best value for investment 
to increase youth employment and support improved income for youth? How might 
workforce development programs complement this potential? 
PROJECT BACKGROUND
The objective of the Bridges Project was to increase employment within targeted, high-growth 
economic sectors by strengthening the technical and soft skills of at-risk youth in high-crime 
municipalities. The activity promoted formal and non-formal vocational and technical training that 
responded to market demand and improved workforce development services to help at-risk 
youth successfully compete for work in targeted growth sectors. Other key focus areas included:
● Improve the collection and analysis of labor market demand data for public use.
● Improve workforce development programs to promote a skilled workforce.
● Promote youth entrepreneurship and self-employment to increase employment and 
create jobs.
1
Academic level certifications approved by the Ministry of Education. 
2
Academic level achievements not approved by the Ministry of Education, but recognized by INSAFORP 
(Instituto Salvadoreño de Formación Profesional).
ii
● Advocate for effective corporate social responsibility programs to advance respect for 
inclusive and fair human resource practices related to young workers.
In addition, Bridges invested in the development of a beneficiaries’ data management system called 
SisPuentes. Bridges provided support and technical assistance to adopt this customized system 
which was also used to report data back to USAID. This platform allows training centers to track 
youth beneficiaries during the course of the training, i.e., to monitor enrollment, attendance and 
performance through pre- and post-tests; monitor their employment status after they graduate; 
and evaluate progress toward the targets for the project indicators. Data extracted from the 
SisPuentes system had more than 20,000 sanitized records from young people who benefited 
from this USAID-funded workforce development project as well more than 1,200 records from 
companies that hired these young people. 
Through Bridges, USAID supported the following economic sectors:
1. Manufacturing (including plastics and textiles);
2. Tourism, commerce, and services (high end services as logistic services, and lower end 
services as waiters, cashiers, bartenders, sales, display, etc.);
3. Agroindustry (including the food and beverages sector);
4. Energy; and
5. Information Technology and Communications
Moreover, the program also supported the following additional sectors:
● Positive Youth Development;
● Private Sector Engagement;
● New Partners Initiative Development;
● Female and women empowerment; and
● Diversity and inclusion (emphasis in youth with disabilities). 
DATA, METHODS AND LIMITATIONS
Research questions pulled data from three SisPuentes databases: one containing data on 26,393 
beneficiaries of Bridges, a second with 15,981 records with data on all the courses attended by 
13,535 beneficiaries, and a third with 4,708 records containing data on employment outcomes. A 
single data set was created by merging the beneficiaries, courses, and employment files.
The main factors affecting employability were assessed by fitting separate logistic regression 
models to estimate the probability of employment for each of the key economic sectors. The 
technique known as Multivariate Adaptive Regression Splines (MARS)3 was used to assess the 
factors affecting youth salaries.
3 MARS is a non-parametric regression method that builds multiple linear regression models across the 
range of predictor values. The MARS algorithm builds a model in two steps. First, it creates a collection of 
so-called basis functions (BF). In this procedure, the range of predictor values is partitioned in several 
groups. For each group, a separate linear regression is modeled, each with its own slope. The connections 
between the separate regression lines are called knots. In the second step, MARS estimates a least￾squares model with its basis functions as independent variables. (Friedman, 1991; Hastie, Tibshirani & 
Friedman, 2009). 
iii
The codes, analyses and outputs for this report were generated using SAS/STAT software, 
Version 9.4 of the SAS®4 System for Windows.
Research Question 2 could not be answered from the Bridges data as the variables included in 
the files refer to the jobs made available to the project beneficiaries and not to the type of jobs 
typically sought by Bridges participants.
Similar challenges were faced to answer Research Question 4 on the identification by sector of 
data knowledge discovery and beneficiary profiles. The factors affecting youth employability and 
salaries were those predicted by the standard economic theory of employment and earnings 5
. 
The relatively small number of observations for each of the economic sectors prevented the team 
from pursuing a meaningful exercise of extracting previously unknown and potentially useful 
information from data stored in databases. 
The findings from this study are valid to describe the flow from training to employment for the 
youth registered as beneficiaries of Bridges. The findings are not generalizable to the training￾to-employment experience of the whole youth population, nor are they valid to provide 
answers to evaluation questions. 
 
4 SAS and all other SAS Institute Inc. product or service names are registered trademarks or trademarks of 
SAS Institute Inc. in the USA and other countries. ® indicates USA registration. Copyright © 2016 by SAS 
Institute Inc., Cary, NC, USA.
5 Human Capital Theory (Becker, 1993) deals with the effects of education and training on earnings and 
other labor outcomes. "Human capital" can be defined as knowledge, skills, attitudes, aptitudes, and other 
acquired traits contributing to production. There are two main components of human capital with strong 
complementarity: early ability (whether acquired or innate) and skills acquired through formal education or 
training (Blundell et al, 1999). The stock of human capital is the major contributor to the earnings of 
individuals (Ishikawa & Ryan, 2002; Mincer, 1974). The human capital theory predicts that earnings 
increase with formal education, training and experience, after controlling for local labor market conditions. 
Research indicates that formal education, training and experience account for most of the observed 
variation in salaries (Mincer, 1989). Better educated and more skilled workers receive a premium, 
resulting to higher earnings. Also, education has a positive effect not only in facilitating access to 
employment, but also in improving the chances of gaining quality employment. 
iv
FINDINGS, CONCLUSIONS AND RECOMMENDATIONS
Table A1: Summary of Findings and Conclusions
Questions
Q1: By key sectors, how can participants’ specific technical, academic, and vocational skills 
(alone and/or combined) be most effectively leveraged for being hired into jobs above the 
minimum wage?
Q1.1: How do such factors as level of education, quantity and quality of Bridges training, 
social skills influence the salaries at which Bridges beneficiaries are hired?
Q1.2: Do the effects of these variables vary across sectors of capacitation, gender, or area 
of residence of the Bridges beneficiary? 
Q1.3: Are there any other factors that affect the earnings of the Bridges beneficiaries?
Findings Conclusions
Q1.2 There are no salary differentials due to the 
economic sector, after accounting for the effect of 
the other variables in the model.
1. Salaries did not differ across economic 
sectors. (Q1.2)
2. Salaries for men and women were similar. 
(Q1.2) 
3. Living in a rural area did not make a difference 
on youth salaries as compared to living in an 
urban area. (Q1.2)
4. As expected, formal education and training 
emerged as the main factors affecting Bridges 
beneficiaries’ salaries. (Q1.1)
5. Experience matters for youth earnings. (Q1.1)
6. Employment conditions have an impact on 
youth salaries. Youth on full-time or part-time 
employment and being paid on a fortnightly or 
monthly basis, have higher salaries than those 
on other forms of employment or payment 
period. (Q1.3)
Q1.2 There are no salary differentials due to either 
sex or area of residence, after accounting for the 
effect of the other variables in the model.
Q1.1 Salary grows 2.7 percent for each additional 
year of education beyond high school, and increases 
by 1.3 percent for each additional year of experience 
(as measured by whether the youth was 22 years 
and older).
Q1.1 Each additional training course a youth attends 
increases his/her salary by 15 percent. 
Q1.3 Not being on a temporary job, or not being 
hired on an hourly basis, or not being an 
entrepreneur, either formal or informal, increases 
the salary by 5.9 percent.
Q1.3 Being paid monthly or fortnightly increases the 
salary by 25 percent.
v 
Findings Conclusions
Q3: By sector, how much does having a 1) high school degree or 2) at least some high 
school education contribute to employment? Is there a wage differential?
Having an educational level above a high school certificate 
increased a youth likelihood of employment in the energy, 
manufacturing, and information technology and communications 
sectors. 
1. The sectors of energy, 
manufacturing, and information 
technology require an educational 
level above high school to hire 
youth. This is expected as the 
relative sophistication of the 
business and production processes 
in these sectors demand workers 
equipped with competencies and 
skills beyond those provided by 
the basic and middle education 
system. 
2. In the agroindustry and retail 
trade/tourism sectors, holding a 
high school degree seems to be 
enough for employment.
3. Having a high school certificate 
makes little difference in earnings
when a special level of effort is 
provided by implementers to 
support beneficiaries with less 
than high school degree 6. 
The strongest effect of having an educational level above a high 
school certificate was observed for the energy sector. A high 
school graduate was 6 times as likely to be employed as a non￾high school graduate. 
The information technology and communications sector 
recorded the second strongest effect of having an educational 
level above a high school certificate. A high school graduate was 
twice as likely to be employed as a non-high school graduate. 
An effect was also observed for the manufacturing sector. A 
youth with an educational level above high school was 1.6 times 
as likely to be employed as a non-high school graduate. 
Having an educational level above a high school certificate did 
not affect the likelihood of employment in both the retail/trade 
and agroindustry sectors.
Salaries grow 2.7 percent for each additional year of education 
beyond high school. 
Findings Conclusions
Q4: What data knowledge discovery was identified by sector and beneficiary profiles?
No factors explaining youth employability or salaries beyond 
those predicted by standard economic theory were identified 
from the data.
The information extracted from the 
data is consistent with the actual 
implementation of the Bridges 
6 The findings from this study are valid to describe the flow from training to employment for the youth 
registered as beneficiaries of Bridges. The findings are not generalizable to the training-to-employment 
experience of the whole youth population, nor are they valid to provide answers to evaluation questions. Out of every 10 young people that were placed through the Bridges Project, eight have a high school 
degree and two do not have a high school degree. The total universe assessed was the +/- 5,000 youth 
who attained employment as documented in the SisPuentes database.
vi
The information contained in the databases reflects the way the 
Bridges Project was implemented to meet employment, salary, 
gender equity, and private sector engagement goals. 
Project and with its theory of 
change.
Q5: How do formal or informal minimum academic requirements by the private sector 
affect the potential for youth to be employed? Based on the data at hand, is there a 
minimum academic level required by the private sector to employ youth?
Having a high school degree increased a youth likelihood of 
employment in the energy, manufacturing, and information 
technology and communications sectors.
1. The sectors of energy, 
manufacturing, and information 
technology demand an 
educational level above high 
school to hire youth. 
2. In the agroindustry and retail 
trade/tourism sectors, holding a 
high school degree seems to be 
enough for employment.
3. In the manufacturing sector, 
having completed a course 
recognized by INSAFORP has a 
strong beneficial effect on youth 
employability beyond the benefits 
due to having a formal 
certification. 
The strongest demand for high school graduates was observed 
for the energy sector. A high school graduate was 6 times as 
likely to be employed as a non-high school graduate. 
The information technology and communications sector 
recorded the second strongest demand for high school 
graduates. A high school graduate was twice as likely to be 
employed as a non-high school graduate. 
The manufacturing sector ranked third in demand for high school 
graduates. A high school graduate was 1.6 times as likely to be 
employed as a non-high school graduate. 
Both the retail/trade and agroindustry sectors demanded youth 
with education levels below high school.
Only for the manufacturing sector, did having completed a 
course backed by Instituto Salvadoreño de Formación Profesional
(INSAFORP) increase a youth likelihood of employment beyond 
the positive effect of a high school certification. Independent of 
the other variables in the model, the probability of employment 
for a youth with high school certificate was assessed at 0.60. This 
probability increased to 0.77 when the youth completed a course 
certified by INSAFORP, a 17 percent increase in the likelihood 
of employment.
vii
Findings Conclusions
Q6A: Based on the data at hand, which economic sectors would be the best value for 
investment to increase youth employment and support improved income for youth?
Q6B: How might workforce development programs complement this potential? (Which 
economic sectors offer the highest rates of employment, higher earnings, and lower 
gender differentials?)
Bridges training effects on youth employability:
●Retail Trade/Tourism and Manufacturing were the sectors 
where youth employability was most positively impacted by 
Bridges training. The average partial effect (APE) of training on 
the probability of employment was assessed at 0.45 for retail 
trade/tourism, and 0.44 for manufacturing. These effects 
indicate how much the probability of employment increases by 
each additional training course completed.
●In the information technology and communications sector, the 
APE of training was assessed at 0.25.
●The weakest average effects of training were observed in the 
energy and agroindustry sector. The APE for the former was 
only 0.03 and it was null in the agroindustry sector. 
Formal education effects on youth employability: 
●Energy was the sector where formal education impacted the 
most on employability. The APE of an additional year of formal 
education beyond high school on the probability of 
employment was 0.26.
●The APE was assessed at 0.10 and 0.09 for the sectors of 
information technology and manufacturing, respectively.
●The smallest APE on the probability of employment was 
observed for the retail trade/tourism sector (0.04). 
The Bridges training premium: A measure for USAID 
value for investment to increase youth employment. 
●The training premium was assessed from the ratio of training 
to formal education APE effects on the probability of 
employment. 
●Retail trade/tourism was the sector where the training 
premium was larger. The APE of training was 11.25 times the 
APE for formal education beyond high school.
●Manufacturing and information technology followed with 
training premiums of 4.9 and 2.5, respectively.
●Energy recorded a training premium of 0.12, higher only to 
agroindustry where this premium was 0.
1. In terms of efficacy of Bridges 
training, retail trade/tourism was 
the sector representing the best 
value for investment, followed by 
manufacturing and information 
technology.
2.Whether a youth has attained a 
level of education of high school 
or above is key to obtaining 
employment in the energy, 
manufacturing, and information 
technology sectors.
3. Education influences both the 
prospects of youth employment 
and their salaries. It provides 
youth with the basic knowledge 
that is required for a successful 
completion of any training 
program.
4.Regardless of the economic 
sector being considered, 
providing youth with incentives to 
complete high school and to 
undertake further studies is a key 
factor for the efficacy of any 
workforce development 
program. 
Salary and sex differentials:
• No salary differentials due to both sector and sex were 
detected.
viii
Table A2: Recommendations
Recommendations
Q6A: On the economic sectors that offer the best value for USAID support to increase 
youth employment and improve income for youth:
In terms of youth employment:
• Energy, manufacturing, and information technology are the sectors offering the greatest potential 
to employ youth with a minimum of 13 years of formal schooling.
• Retail commerce/tourism is the sector with the most potential to employ youth with completed 
high school education or lower.
In terms of youth earnings:
• Earnings are sensitive to the formal education level achieved and the quantity and quality of training. 
Therefore, USAID workforce development programs might focus in the sectors of energy, 
manufacturing, and information technology that require levels of education above high school and 
technical skills. 
Q6B: How might workforce development programs complement this potential?
Initiatives to increase youth educational attainment:
• Combining flexible vocational education with part-time apprenticeships through public-private 
partnerships. 
• Combining technical education scholarships, both formal (MINED) and professional (INSAFORP), 
with part-time apprenticeships.
Initiatives to increase youth employability:
• Supporting INSAFORP training and certification programs for youth who complete high school.
• Designing programs to improve soft and labor market skills among the youth.
• Providing grants to increase the youth intake by businesses. 
Initiatives to increase youth income: 
• Supporting relevant technical and tertiary education retention programs (i.e., scholarships, 
internships, etc.).
1 
1.0 Study Purpose and Questions
1.1 STUDY PURPOSE
This report presents primary findings, conclusions, and recommendations from the analysis of 
data collected under the United States Agency for International Development (USAID) Bridges 
to Employment (Bridges) Project. The purpose of the study is threefold. First, it finds insights, 
patterns, relationships, and trends through diverse data mining analysis methods, not previously 
identified. Second, the study aims at providing evidence and identifying trends that might inform 
future interventions under the newly approved CDCS, and third, to support knowledge discovery 
in data processes that identifies beneficiary professional and academic profiles in the business 
sectors participating in the Bridges Project.
The main audience of this study will be USAID El Salvador, particularly the Office of Economic 
Growth and Education (EGE) and the Regional Program Office (RPO).
1.2 RESEARCH QUESTIONS
The study seeks to answer the following research questions:
1. By key sectors, how can participants’ specific technical, academic, and vocational skills 
(alone and/or combined) be most effectively leveraged for being hired into jobs above the 
minimum wage?
2. How do sector-specific academic and skills requirements act as a barrier to either men 
or women looking for employment typically sought by Bridges participants? 
3. As specifically as possible, by sector, how much does having a 1) high school degree or 
2) at least some high school education contribute to employment? Is there a wage 
differential? 
4. What data knowledge discovery was identified by sector and beneficiary profiles?
5. How do formal or informal minimum academic requirements by the private sector affect 
the potential for youth to be employed? Based on the data at hand, is there a minimum 
academic level required by the private sector to employ youth?
6. Based on the data at hand, which economic sectors would be the best value for investment 
to increase youth employment and support improved income for youth? How might 
workforce development programs complement this potential?
2.0 Project Background
2.1 PROJECT CONTEXT
The objective of the Bridges Project was to increase employment within targeted, high-growth 
economic sectors by strengthening the technical and soft skills of at-risk youth in high-crime 
municipalities. The activity promoted formal and non-formal vocational and technical training that 
responded to market demand and improved workforce development services to help at-risk 
youth successfully compete for work in targeted growth sectors. Other key focus areas included: 
2 
1) improve the collection and analysis of labor market demand data for public use; 2) improve 
workforce development programs to promote a skilled workforce; 3) promote youth 
entrepreneurship and self-employment to increase employment and create jobs; and 4) advocate 
for effective corporate social responsibility programs to advance respect for inclusive and fair 
human resource practices related to young workers. 
In addition, Bridges invested in the development of a beneficiaries’ data management system called 
SisPuentes. This platform allows training centers to track youth beneficiaries during the course 
of the training. Data extracted from the SisPuentes system have more than 20,000 records from 
young people who benefited from USAID’s Bridges workforce development project as well more 
than 1,200 records from companies that hired these young people. 
Through the Bridges program, USAID supports the following economic sectors: 1) 
manufacturing (including plastics and textiles); 2) tourism, commerce and services (high end 
services as logistic services, and lower end services as waiters, cashiers, bartender, sales, 
display, etc.); 3) agroindustry (including the food and beverages sector); 4) energy; and 5) 
information technology and communications. Moreover, the program also supported the 
following additional sectors: 1) positive youth development; 2) private sector engagement; 3) 
new partners initiative development; 4) female and women empowerment; and 5) diversity and 
inclusion (emphasis in youth with disabilities).
2.2 PRELIMINARY FINDINGS FROM THE SISPUENTES DATA 
MANAGEMENT SYSTEM
Some preliminary findings from the SisPuentes data management system are listed: 
• The training completion rate of young people with or without a high school degree is similar. 
Both groups tend to finish at an average rate of 80 percent. 
• Out of every 10 young people that were placed through the Bridges Project, eight have a high 
school degree and two do not have a high school degree. The total universe assessed was the 
+/- 5,000 youth who attained employment as documented in the SisPuentes database. 
• More than half of the youth employed without a high school degree have at least an academic 
level of 9th grade.
• USAID Bridges had a wide number of companies (i.e., over 1,200) hiring participating youth. 
In fact, 88 percent of the companies hired five youth or less, and less than 1 percent hired 50 
youth or more. 
• There are significant differences in the beneficiary profile by sector. This results in a varied 
employment and wage rate. Sector jobs have specific academic requirements. Youth with a 
specific profile were consistently able to attain employment and/or earn wages higher than 
the legal minimum (+/- $305 per month). Similarly, youth who attained a job, but did not 
meet the wage threshold established in the Monitoring, Evaluation, and Learning (MEL) Plan 
for the USAID Bridges Project had a well-defined profile (these were youth in the textile 
industry where the minimum wage required by law is +/- $295 p/month).
3 
3.0 Data, methods and limitations
3.1 DATA
Research questions pulled data from three SisPuentes databases: one containing data on 26,393 
beneficiaries of the Bridges program, a second with 15,981 records with data on all the courses 
attended by each beneficiary, and a third with 4,708 records containing data on employment 
outcomes.
For data analysis, the course file only included data on the last course attended by each youth 
besides data on the total number of courses taken and demographic and geographical variables. 
This file contained data on 13,535 observations. A single data set was created by merging the 
beneficiaries, last course, and employment files. This analysis data set included 
26,393 beneficiaries, 13,535 youth enrolled in a course, and 4,708 employed youth. The variables 
contained in this data set are listed in Appendix 1.
The codes, analyses, and outputs for this report were generated using SAS/STAT software, 
Version 9.4 of the SAS®7 System for Windows.
3.2 METHODS
3.2.1 Assessing the factors affecting the likelihood of employment among Bridges 
participants
This analysis addressed Quarter 4 (Q4) (part 1), Q5, and Q6 — all related to factors affecting 
the likelihood of employment. Data on the 13,535 youth enrolled in at least one course were 
used to fit separate logistic regression models to estimate the probability of employment by key
economic sector. The predictor variables were the following: 1) whether the youth had 
completed high school prior to participation in Bridges; 2) whether the youth had education 
above high school prior to participation in Bridges; 3) whether the youth was male or female; 4) 
whether the area of residence was urban or rural; 5) cohort of admission into Bridges; 6) age at 
time of enrolment in a Bridges course; 7) whether the youth completed the last course attended; 
8) whether the youth improved his/her level in the last course attended; 9) whether the course 
was part of the Instituto Salvadoreño de Formación Profesional (INSAFORP) curricula (informal), or 
not (formal); 10) whether received life-skills and completed input and output assessments; and 
11) whether the youth returned to study.
3.2.2 Assessing the factors affecting the salaries of the Bridges participants
This analysis wholly addressed Q2 and the second part of Q4. Salaries and earnings depend upon 
factors such as education, experience, quality of relational networks, employment duration, job 
tenure, employment shocks, shocks to general skills, and draws of new job opportunities offering 
different hours and wages (Altonji, Smith Jr. & Vidangos, 2013). That the relationship between 
salaries and most of these variables is nonlinear is a well-known fact. However, there is enough 
evidence to suggest that this relationship is roughly linear in the logarithm of salaries. The 
7 SAS and all other SAS Institute Inc. product or service names are registered trademarks or trademarks of 
SAS Institute Inc. in the USA and other countries. ® indicates USA registration. Copyright © 2016 by SAS 
Institute Inc., Cary, NC, USA.
4 
standard approach is to regress the logarithm of salaries on education, experience, and other 
labor market variables. The working hypothesis is that salary increases are larger among the most 
educated and experienced individuals. A key assumption is that salaries vary widely both across 
individuals and over the values of the intended explanatory variables. 
The apparent absence of a linear relationship between most of the candidate explanatory 
variables and the logarithm of salaries, led us to fit a prevented us Multivariate Adaptive 
Regression Splines (MARS) model for the natural logarithm of wage instead of the traditional 
regression model. 
3.3 LIMITATIONS
Research Question 2 could not be answered from the Bridges data as the variables included in 
the files refer to the jobs made available to the project beneficiaries and not to the type of jobs 
typically sought by Bridges participants.
Similar challenges were faced to answer Research Question 4 on the identification by sector of 
data knowledge discovery and beneficiary profiles. The factors affecting youth employability and 
salaries were those predicted by the standard economic theory of employment and earnings. The 
relatively small number of observations for each of the economic sectors prevented the team 
from pursuing a meaningful exercise of extracting previously unknown and potentially useful 
information from data stored in databases.
The COVID-19 global pandemic, staff rotation in the private sector due to COVID-19 impacts, 
and job seasonality were not factored in at the time of implementation. Thus, the data studied in 
this report does not contemplate these external factors, per the SOW. Readers are advised to 
consider these external factors when reading the report. 
The findings from this study are valid to describe the flow from training to employment for the 
youth registered as beneficiaries of Bridges. The findings are not generalizable to the 
training-to-employment experience of the whole youth population, nor are they 
valid to provide answers to evaluation questions at a country-level. Data is focused 
on the successful beneficiary profile attained by the USAID Bridges to Employment 
interventions. 
4.0 Findings and conclusions
4.1 FINDINGS
4.1.1 Q1: By key sectors, how can participants’ specific technical, academic and 
vocational skills (alone and/or combined) be most effectively leveraged for being 
hired into jobs above the minimum wage?
No salary differentials due to the economic sector of employment were detected. 
Once the variation in salaries was controlled for the effects of education, experience, training, 
and working conditions, no differential effects due to the economic sector of employment were 
detected. The economic sector was not selected for inclusion in the set of final basis functions 
in the MARS model (refer to Table 1).
5 
Table 1: MARS Basis Functions and Their Coefficients – Model for the Logarithm of 
Salary 
Transformed Variable (Basis Function) Coefficient
Intercept 5.834 
Attended one training course -0.280
Attended two training courses 0.137
Attended three training courses 0.366 
Attended four training courses -0.073
First Bridges cohort -0.189
Second or third Bridges cohort -0.067
Fourth Bridges cohort 0.206 
Paid monthly or fortnightly 0.253 
Hired on a casual or hourly basis, or 
entrepreneur (formal or informal) -0.059
More than 13 years of schooling 0.027
Being older than 22 years of age 0.013
Mean square error 0.073 
R Square 0.207
No salary differentials due to either sex or area of residence were detected. Neither 
sex nor area of residence were included among the set of final basis functions in the MARS model 
(refer to Table 1).
Salary grows with formal education beyond high school, and with experience. As 
expected, salary grew 2.7 percent for each additional year of education beyond high school, and 
increased by 1.3 percent for each additional year of experience, as measured by whether the 
youth was 22 years and older (refer to Table 1).
Bridges training had a strong impact on youth salaries. Each additional training course a 
youth attends increases his/her salary by 15 percent (refer to Table 1).
Working and payment conditions had an effect on youth salaries. Not being on a 
temporary job, or not being hired on an hourly basis, or not being an entrepreneur, either 
formal or informal, increased the youth salary by 5.9 percent. Besides, being paid monthly or 
fortnightly increased the salary by 25 percent. (refer to Table 1).
4.1.2 Q3: By sector, how much does having a 1) high school degree or 2) at least 
some high school education contribute to employment? Is there a wage 
differential?
Having an educational level above a high school certificate increases youth likelihood 
of employment in the energy, manufacturing, and information technology and 
communications sectors. The strongest effect of having an educational level above a high 
6 
school certificate was observed for the energy sector. A high school graduate was 6 times as 
likely to be employed as a non-high school graduate8 (refer to Table 2). 
Table 2: Logistic Regression Models Coefficients for the Likelihood of Employment 
by Economic Sector
 Agro￾industry
Retail 
Trade/
Tourism
Energy Manufacturing Information 
Technology
Being a male 1.01** 0.21 
Age at Bridges 
registration 0.19** 0.07** 0.11** 
Education above high 
school prior to 
Bridges
-3.80 1.85** 0.41** 0.68** 
Number of courses 
attended 0.16** 0.89** 
Cohort -0.21** 
Last course was on 
integral skills -5.63** -0.71** 
Completed the last 
course 
1.99** 0.79** 
Improved in the last 
course 
3.24** 
Course was part of 
INSAFORP curricula 
(informal)
1.21** 
Intercept -2.06 -3.80 -5.72 -0.91 -4.78 
Number of 
observations 252 2,774 216 2,426 2,611 
Pearson Chi-Square 8.8 191.5 48.6 411.7 729.8 
Average Partial Effect (APE)
Education above high 
school 0.04 0.26 0.09 0.10 
Training 0.45 0.03 0.44 0.25 
** Significant at the 5% level.
8 The number of times a high school graduate was more likely to obtain employment than a youth with no 
high school is known as the odds of employment. It is equal to the exponential of the logistic regression 
coefficient. An odds value of 1 means that high school graduates are equally as employable than non-high 
school graduates ones. An odds value greater than 1 means that a high school graduate is more likely to 
be employed than a youth with no high school certificate, and an odds value lower than 1 means the 
opposite. 
7 
The information technology and communications sector recorded the second strongest effect of 
having an educational level above a high school certificate. A high school graduate was twice as 
likely to be employed as a non-high school graduate. A weaker effect was observed for the 
manufacturing sector. A youth with an educational level above high school was 1.6 times as likely 
to be employed as a non-high school graduate. Having an educational level above a high school 
certificate did not affect the likelihood of employment in both the retail/trade and agroindustry 
sectors (refer to Table 2). 
Salary grew 2.7 percent for each additional year of education beyond high school (refer to 
Table 1). Having a high school certificate makes little difference in earnings when a special level 
of effort is provided by implementers to support beneficiaries with less than high school degree. 
The monthly salary differential due to high school completion was $10. The monthly salary for a 
youth with high school was assessed at $313. The salary for a youth with education below high 
school was assessed at $303. 
4.1.3 Q4: What data knowledge discovery was identified by sector and beneficiary 
profiles?
No factors explaining youth employability or salaries beyond those predicted by 
human capital theory were identified from the data. The information contained in the 
databases reflects the way the Bridges Project was implemented to meet employment, salary, 
gender equity, and private sector engagement goals (refer to Tables 1 and 2).
4.1.4 Q5: How do formal or informal minimum academic requirements by the 
private sector affect the potential for youth to be employed? Based on the data at 
hand is there a minimum academic level required by the private sector to employ 
youth?
Having a high school degree increased youth’s likelihood of employment in the 
energy, manufacturing, and information technology and communications sectors. 
The strongest demand for high school graduates was observed for the energy sector. A high 
school graduate was 6 times as likely to be employed as a non-high school graduate. The 
information technology and communications sector recorded the second strongest demand for 
high school graduates. A high school graduate was twice as likely to be employed as a non-high 
school graduate. The manufacturing sector ranked third in demand for high school graduates. A 
high school graduate was 1.6 times as likely to be employed as a non-high school graduate. Only 
for the manufacturing sector, having completed a course backed by INSAFORP increased a 
youth’s likelihood of employment beyond the positive effect of a high school certification. 
Independent of the other variables in the model, the probability of employment for a youth 
with a high school certificate was assessed at 0.60. This probability increased to 0.77 when the 
8 
youth completed a course certified by INSAFORP, a 17 percent increase in the likelihood of 
employment (refer to Table 2).9
4.1.5 Q6: How do formal or informal minimum academic requirements by the 
private sector affect the potential for youth to be employed? Based on the data at 
hand is there a minimum academic level required by the private sector to employ 
youth?
The Bridges training effects on youth employability. Retail trade/tourism and 
manufacturing were the sectors where youth employability was most positively impacted by 
Bridges training. The average partial effect (APE) of training on the probability of employment 
was assessed at 0.45 for retail trade/tourism, and 0.44 for manufacturing.10 In the information 
technology and communications sector, the APE of training was assessed at 0.25. The weakest 
average effects of training were observed in the energy and agroindustry sector. The APE for the 
former was only 0.03 and it was null in the agroindustry sector (refer to Table 2).
Formal education effects on youth employability. Energy was the sector where formal 
education impacted employability the most. The APE of an additional year of formal education 
beyond high school on the probability of employment was 0.26. The APE was assessed at 0.10 
and 0.09 for the sectors of information technology and manufacturing, respectively. The smallest 
APE on the probability of employment was observed for the retail trade/tourism sector (0.04). 
(refer to Table 2).
The Bridges training premium: a measure for USAID value for investment to 
increase youth employment. The training premium was assessed from the ratio of training 
to formal education APE effects on the probability of employment. Retail trade/tourism was the 
sector where the training premium was larger. The APE of training was 11.25 times the APE for 
formal education beyond high school. Manufacturing and information technology followed with 
training premiums of 4.9 and 2.5, respectively. Energy recorded a training premium of 0.12, higher 
only to agroindustry where this premium was 0 (refer to Table 2).
Salary and sex differentials. No salary differentials due to sex were detected (refer to Table 
2).
Economic sectors that offer the best value for USAID support to increase youth 
employment and improve income for youth:
a. In terms of youth employment: Energy, manufacturing, and information technology are 
the sectors offering the greatest potential to employ youth with a minimum of 13 years 
of formal schooling. Retail trade/tourism is the sector with potential to employ youth with 
9 The findings from this study are valid to describe the flow from training to employment for the youth 
registered as beneficiaries of Bridges. The findings are not generalizable to the training-to-employment 
experience of the whole youth population, nor are they valid to provide answers to evaluation questions. Out of every 10 young people that were placed through the Bridges Project, eight have a high school 
degree and two do not have a high school degree. The total universe assessed was the +/- 5,000 youth 
who attained employment as documented in the SisPuentes database. 10 The APE indicates by how much the probability of employment increases by each additional training 
course completed. For the logic model, the APE is equal to the regression coefficient for training multiplied 
by the average of the probabilities of employment assessed at the means of the explanatory variables 
across all observations.
9 
a high school education or lower (refer to Table 1).
b. In terms of youth earnings: Earnings are sensitive to the formal education level achieved 
and the quantity and quality of training. Therefore, USAID workforce development 
programs might focus on the sectors of energy, manufacturing, and information 
technology that require levels of education above high school and specific technical skills 
(refer to Table 2).
4.2 CONCLUSIONS
4.2.1 Q1 Conclusion
Salaries do not differ across economic sectors, gender, or the area of residence. Salary 
differentials are due to factors such as formal education and training. These are the main factors 
affecting Bridges beneficiaries’ salaries. Consistent with expectations, experience matters for 
youth earnings. Also, employment conditions have an impact on youth salaries. Youth with full￾time or part-time employment who are paid on a fortnightly or monthly basis have higher salaries 
than those on other forms of employment or payment periods.
4.2.2 Q3 Conclusion
The sectors of energy, manufacturing, and information technology require an educational level 
above high school. This is expected as the relative sophistication of the business and production 
processes in these sectors demand workers equipped with competences and skills beyond those 
provided by the basic and middle education system. In the agroindustry and retail trade/tourism 
sectors, holding a high school degree seems to be enough for employment.
Having a high school certificate with no further studies makes little difference in earnings, when 
vulnerable youth receive support from USAID programs. The high school premium was 
assessed at $10 a month. The result here refers to youth who participated in the Bridges to 
Employment program who were employed and therefore included in the salaries model. Within 
this group, employed youth with only a high school certificate represented 62% of the total, and 
those with studies above high school represented 14.4%. The predicted salary for a youth with 
high school but no further studies was $313. This salary was $10 higher than the salary for a 
youth with education below high school. The average academic level of the employed youth 
without a high school certificate was equivalent to first year of high school. The predicted salary 
was $330 for a youth with education level above high school. This confirms our expectation 
that salary increases with education.
4.2.3 Q4 Conclusion
The information extracted from the data is consistent with the actual implementation of the 
Bridges Project and with its theory of change.
4.2.4 Q5 Conclusion
The sectors of energy, manufacturing, and information technology require that youth possess an 
educational level above high school in order to be hired. In the agroindustry and retail 
trade/tourism sectors, holding a high school degree seems to be enough for employment. In the 
10
manufacturing sector, having completed a course recognized by INSAFORP has a strong beneficial 
effect on youth employability beyond the benefits due to having a formal certification.
4.2.5 Q6 Conclusion
In terms of the efficacy of Bridges training, retail trade/tourism was the sector representing the 
best value for the investment, followed by manufacturing and information technology. Whether 
a youth has attained a level of education of high school or above is key to obtaining employment 
in the energy, manufacturing, and information technology sectors. Education influences both the 
prospects of youth employment and their salaries. It provides youth with the basic knowledge 
that is required for the successful completion of any training program. Regardless of the economic 
sector being considered, providing youth with incentives to complete high school and to 
undertake further studies is a key factor for the efficacy of any workforce development program.
4.2.6 Q6 Recommendations - How might workforce development programs 
complement this potential?
Although the SisPuentes data do not enable answering this question, the findings seem to provide 
evidence to suggest some initiatives that might be included as part of a revised new version of 
the Bridges Project. These initiatives are grouped in three areas: 1) initiatives aimed at increasing 
youth educational attainment; 2) initiatives to increase youth employment; and 3) initiatives to 
increase youth income.
Initiatives to increase youth educational attainment
• Combining flexible vocational education with part-time apprenticeships through public￾private partnerships.
• Combining technical education scholarships with part-time apprenticeships.
Initiatives to increase youth employability
• Supporting INSAFORP training and certification programs for youth who complete general 
high school.
• Designing programs to improve soft and labor market skills among the youth.
• Providing grants to increase the youth intake by businesses.
Initiatives to increase youth income 
• Supporting relevant technical and tertiary education retention programs (i.e., scholarships, 
internships, etc.).
11
APPENDICES
12
APPENDIX I: STATEMENT OF WORK (SOW)
USAID/El Salvador
Workforce Development Study
A. Purpose of the Study
The purpose of this study is to analyze data collected under the USAID Bridges to Employment 
Project. Using data mining the contractor is expected to find insights, patterns, relationships, and 
trends through diverse analytical methods, not previously identified. The data to be analyzed have 
been collected in the beneficiaries’ data management system called SisPuentes which was 
developed under this Project. The study will extract usable data to expand on findings. Usable 
data should include, but not be limited to beneficiary profile, employment by educational level, 
education related dropout rates and completion rates, possible wages related to specific sectors 
(e.g., agroindustry, commerce and tourism, energy, manufacturing, and information technology 
and communications (ITC)), and demonstrating potential wages to varying educational levels. 
This study will also provide evidence and will identify trends to inform future interventions under 
the newly approved Country Development Cooperation Strategy (CDCS). The study shall 
support knowledge discovery in data processes11 that identifies beneficiary professional and 
academic profiles in diverse key business sectors—at a minimum, the study should describe 
beneficiary profiles12 for each sector identified in the data. The beneficiary profiles should also 
reference El Salvador’s National Policy for diversification and productive transformation and 
consider key investments sectors that promote economic growth.
The main audience of this study will be USAID El Salvador, particularly the Office of Economic 
Growth and Education (EGE) and the Regional Program Office (RPO).
The study is scheduled to occur during the first quarter of FY 2021.
B. Background Information about USAID Bridges to Employment Project 
The objective of the Mission’s Bridges to Employment program was to increase employment 
within targeted, high-growth economic sectors by strengthening the technical and soft skills of 
at-risk youth in high-crime municipalities. The activity promoted formal and non-formal 
vocational and technical training that responded to market demand and improved workforce 
development services to help at-risk youth successfully compete for work in targeted growth 
sectors. Other key focus areas included:
● Improve the collection and analysis of labor market demand data for public use.
● Improve workforce development programs to promote a skilled workforce.
● Promote youth entrepreneurship and self-employment to increase employment and 
create jobs.
● Advocate for effective corporate social responsibility programs to advance respect for 
inclusive and fair human resource practices related to young workers.
11 Knowledge Discovery in Databases refers to the nontrivial extraction of implicit, previously unknown, and 
potentially useful information from data stored in databases. 
12By sector, each beneficiary profile should include weighted variables such as, but not limited to age, 
academic level, etc. The Study Team must review the variables included in the database and select those 
ones that are statistically significant.
13
In addition, Bridges invested in the development of a beneficiaries’ data management system called 
SisPuentes. 
Bridges provided support and technical assistance to adopt this customized system which Bridges 
also used to report data to USAID. This platform allows training centers to track youth 
beneficiaries during the course of the training, i.e., to monitor enrollment, attendance, and 
performance through pre- and post-tests; monitor their employment status after they graduate; 
and evaluate progress toward the targets for the project indicators. Data extracted from the 
SisPuentes system has more than 20,000 sanitized records from young people who benefited 
from USAID’s Bridges workforce development project as well more than 1,200 records from 
companies that hired these young people. 
Through the Bridges to Employment program, USAID supports the following economic sectors:
● Economic Sector 1: Manufacturing (including plastics and textiles)
● Economic Sector 2: Tourism, commerce and services (high end services as logistic 
services, and lower end services as waiters, cashiers, bartender, sales display, etc.) 
● Economic Sector 3: Agroindustry (including the food and beverages sector)
● Economic Sector 4: Energy 
● Economic Sector 5: Information Technology and Communications 
Moreover, the program also supports the following additional sectors:
● Positive Youth Development
● Private Sector Engagement
● New Partners Initiative Development
● Female and women empowerment
● Diversity and inclusion (emphasis in youth with disabilities)
Preliminary data findings from the SisPuentes data management system 
● The training completion rate of young people with or without a high school degree is 
similar. Both groups tend to finish at an average rate of 80 percent. 
● Out of every 10 young people that were placed through Bridges, eight have a high school 
degree and two don't have a high school degree. The total universe assessed was the 
+/- 5000 youth who attained employment as documented in the SisPuentes database. 
● More than half of the youth employed without a high school degree have at least an 
academic level of 9th grade.
● USAID Bridges to Employment had a wide range of companies (i.e., over 1,200) hiring 
participating youth. In fact, 88 percent of the companies hired 5 youth or less, and less 
than 1 percent hired 50 youth or more. There are significant differences in the beneficiary 
profile by sector. This results in a varied employment and wage rate. Sector jobs have 
specific academic requirements. Youth with a specific profile were consistently able to 
attain employment and/or earn wages higher than the legal minimum (+/- $305 per 
month). Similarly, youth who attained a job, but didn't meet the wage threshold 
established in the MEL Plan for the USAID Bridges to Employment project had a well￾defined profile (these were youth in the textile industry where the minimum wage 
required by law is +/- $295 p/month).
14
C. Research Questions
Research questions include, but are not limited to the following questions:
1. By key sectors, how can participants’ specific technical, academic, and vocational skills 
(alone and/or combined) be most effectively leveraged for being hired into jobs above the 
minimum wage? (Please rank the importance or value of these skills, either individually or 
combined. Please disaggregate the findings, to the degree possible, by gender).
2. How do sector-specific academic and skills requirements act as a barrier to either men 
or women looking for employment typically sought by Bridges participants? (Please rank 
these barriers and disaggregate the findings by gender.) 
3. As specifically as possible, by sector, how much does having a 1) high school degree or 2) 
at least some high school education contribute to employment? Is there a wage 
differential? 
4. What data knowledge discovery was identified by sector and beneficiary profiles?
5. How do formal13 or informal14 minimum academic requirements by the private sector 
affect the potential for youth to be employed? Based on the data at hand, is there a 
minimum academic level required by the private sector to employ youth?
6. Based on the data at hand ,which economic sectors would be the best value for investment 
to increase youth employment and support improved income for youth? How might 
workforce development programs complement this potential?
D. Methodology, Data Collection, and Analysis
Data to answer the research question above will be analyzed from the following SisPuentes 
databases which will be delivered to the Expert(s). USAID/El Salvador will provide the SisPuentes 
databases which includes all beneficiaries assisted under the Bridges to Employment Program. 
The main variables of the SisPuentes databases include:
● Database of Beneficiaries:
1. Código Joven
2. Departamento 
3. Municipio 
4. Área (rural o urbana)
5. Sexo 
6. Rango de edad
7. Rango PPR (17 o menos/18 o mas)
8. Discapacidad (si o no)
9. Nivel Educativo
10. Tipo de Evento (del primer contacto)
11. Nombre del Evento (del primer contacto)
12. Municipio del evento (del primer contacto)
13. Fecha de Inicio del evento (del primer contacto) 
13 Academic level certifications approved by the Ministry of Education. 14 Academic level achievements not approved by the Ministry of Education, but recognized by INSAFORP 
(Instituto Salvadoreño de Formación Profesional).
15
14. Fecha de Fin del evento (del primer contacto) 
15. Periodo del evento por trimestre
16. RFA (para que grant se realizó el onboarding (inducción) del jóven)
17. Centro de Formación del joven
18. Vulnerabilidad (En base al índice de vulnerabilidad de Puentes fue o no aceptado 
el joven)
● Training-Employment database main variables:
1. Código Jóven
2. Edad 
3. Rango de edad
4. Sexo 
5. Nivel Educativo más alto alcanzado
6. Clasificación nivel educativo
7. Municipio (del joven)
8. Área (rural o urbana) 
9. Matricula (en principio un joven podía tomar dos cursos con Puentes ejemplo: 
terminar bachillerato y un entrenamiento específico, o habilidades blandas para la 
vida y un entrenamiento técnico)
10. # de Cursos
11. Completado 
12. Mejorado
13. Nombre del curso
14. Clasificación del curso
15. Clasificación de formación
16. Sector
17. Fechas (inicio y final - varias columnas)
18. RFA (para que grant se realizó el onboarding (inducción) del jóven)
19. Centro de Formación del joven 
20. Municipio del curso
21. Retorno a estudiar (si o no)
22. Evento 
23. Empleo (1er empleo, nuevo empleo o mejor empleo más detalles de ubicación y 
sector - varias columnas)
24. Empresa (más detalle - tres columnas)
25. Salario
It is expected that the Expert(s) process—through data mining—and analyze the existing 
quantitative data extracted from the SisPuentes system with quantitative techniques 
designed to elicit primary data from a wide range of beneficiaries, and enterprises.
In addition, the Study Team should incorporate, to the extent feasible, analysis for any 
differences between effects on men and women or male and female participation.
Desk review of relevant documents
As minimum, the consultant(s) shall review the following documents:
● Final Performance Evaluation: Bridges to Employment in El Salvador
16
● Country Development Cooperation Strategy 2013-2020 and 2020-2025 (?)
● Journey to Self-Reliance Strategy and Learning Agenda
● USAID Private Sector Engagement Policy
It is expected that qualitative data from the Final Performance Evaluation might be used to 
triangulate and analyze the results from this study, appropriately. 
Meetings 
An initial team-planning/kick off meeting will be held between USAID staff and the Expert(s) 
before the submission of a Study Plan so that USAID El Salvador staff can clarify any questions 
from the Expert(s), expectations, and guidelines. The expected results of this meeting are to: 
● Clarify each team member’s role and responsibilities
● Confirm the anticipated timeline and deliverables
● Discuss data analysis and methodologies
Additional meetings may be held as deemed necessary by USAID staff and/or the Expert(s).
E. Deliverables
It is estimated that no more than 30 working days of services from the starting date of the study 
will be required to complete a high-quality analysis as required under this SOW. During that time 
frame, the Study Team shall submit the following deliverables:
1. A Study Plan, in Word Gill Sans font15 size 12, to be completed by the Study Team after 
the Team Planning Meeting, no later than 5 calendar days after the starting day of the 
Study. The Study Team is encouraged to propose analysis methods in this Study Plan that 
Expert(s) considers can yield stimulating, robust evidence in answering each of the 
research questions.
USAID will receive the Study Plan via electronic mail and review it to provide comments 
no later than 5 working days after receiving the document. The Study Plan will provide 
details of how the various deliverables, tasks, and activities will be undertaken. It must 
include at least:
● Timeline and/or Milestone Plan, including tentative starting time for data analysis and 
duration of each activity conducted under the study;
● Study Team composition, roles, and responsibilities; 
The Mission will review and approve the final Study Plan at the start of the consultancy.
2. A Preliminary Findings Briefing for USAID staff, and other stakeholders that 
USAID/El Salvador consider necessary on the preliminary findings, results, and data-based 
recommendations and conclusions for data-driven decision making identified by the 
Expert(s) immediately after finalizing the data analysis phase and before starting the draft 
report. According to the audience, the briefing may be conducted in English or Spanish 
using USAID approved slide templates. Only the Team Leader needs to be present for 
this briefing; however, all the members of the Study Team may also attend. The 
15 If the Evaluation Team does not have Gill Sans font available, they must use any other approved font as 
per the USAID Graphics Manual and Partners Co-Branding Guide.
17
Preliminary Findings Briefing will be used by the Expert(s) as a feedback exercise to 
prepare the Draft of the Final Report.
3. A Draft of the Final Report in Word, Gill Sans font size 12, submitted for review due 
no later than 5 calendar days after the Preliminary Findings Briefing. RPO in USAID/El 
Salvador will be responsible for coordinating the peer-review process with different 
offices within the Mission for comments. RPO will consolidate all comments and send the 
draft back to the Study Team within 7 working days. At a minimum, and in accordance 
with the USAID Evaluation Policy and ADS 201, the Final Report and its draft versions 
must include the following sections:
● Executive Summary of the purpose, background, evaluation questions, findings, 
conclusions, and recommendations;
● Study purpose and questions;
● Through description of the study design and any challenge/limitations,16 with 
emphasis on the timeliness and methods for data analysis;
● Relevant data analysis tables;
● Findings and conclusions drawn from the analysis of the findings;17
● Appendices: 
o Original SOW, annotated with any changes approved by USAID
o Review matrix of documents consulted
o Meeting notes
o Complete schedule of study activities, meetings, and interviews
o List of individuals and organizations contacted and sites visited, as appropriate
● Tables, graphs, pictures taken during site visits, maps, Reporting and Dissemination 
plan (including upload to USAID’s Development Experience Clearinghouse based on 
approval by USAID)
● Any statements of difference regarding significant unresolved differences of opinion 
by funders, implementers, and/or members of the review team; 
● All sources of information, properly identified and listed; 
● In accordance with ADS 201, the contractor will make the final sector review report 
publicly available through the Development Experience Clearinghouse within three 
months of the approval of the final report.
USAID/El Salvador expects to receive a high-quality Draft Report from the Study Team, 
USAID/El Salvador will assess the quality of it using the Evaluation Report and Review 
Template. 
4. Final Report in PDF, font Gill Sans size 12, no longer than 15 pages in its body, excluding 
the cover page; Table of Contents; List of Acronyms; and Appendices. The report should 
be written in American English.
16 The Evaluation Team must identify (a) steps taken to mitigate limitations, and (b) how/whether the 
limitations affect any particular finding, conclusions, or recommendations.
17 In moving from findings to conclusions, the analysis must be clear as to how findings are synthesized 
through different techniques such as divergence, convergence, and amalgamation; propensity; weighting; 
etc. 
18
The approved Final Report must adhere to USAID’s Evaluation Policy and ADS 201, 
Criteria to Ensure the Quality of the Study Report,18 and must be submitted in English 
and Spanish and have incorporated USAID's comments, as appropriate. 
The Final Report will be due to USAID/El Salvador 7 working days after the Study Team 
receives comments on the draft. Two high quality printed, bound copies in English and 
Spanish of the Final Report must be submitted to USAID/El Salvador within 10 calendar 
days of acceptance of the Final Report.
5. A One-Page summary of the study purpose, findings, conclusions and recommendations. 
The One-Page summary will be prepared in English and Spanish in PDF.
F. Study Team
Team Leader/Local Data Analyst Specialist
Minimum qualifications
Education: Bachelor degree in Education, Economics, Finance, Business Administration, 
Accounting, Industrial Engineering, Statistics, Data Management, or other disciplines related to 
financial inclusion, workforce development, or economic development. Master, Ph.D. or 
Doctorate degree or professional with Doctoral candidacy is a plus. Formal training in monitoring 
and evaluation is preferred. 
Language Proficiency: English Level IV and Spanish Level IV
Work Experience: At least six years of relevant prior experience in data management and 
analysis using quantitative and qualitative methods for different types of research and evaluations. 
In addition, at least six years of experience managing and analyzing socio-economic, 
sociodemographic and citizen security variables. Experience in Latin America or Central America 
is an advantage. Experience with management of a multidisciplinary team is a plus. 
Position Description: The Team Leader will be responsible for overseeing, presenting, editing, 
and coordinating all documents and activities related to this study and for ensuring the production 
and completion of quality deliverables in a professional manner, in conformance with this SOW.
One Data Specialist 
Minimum qualifications
Education: Bachelor in Education, Statistics, Social Research, Engineering, Economics, Business 
Administration, or other fields related to data analysis. Master’s degree is a plus. Formal training 
in monitoring and evaluation is preferred.
Language Proficiency: Spanish Level IV and American English Level III.
Work Experience: At least two years of experience in cleaning and processing of quantitative 
and qualitative data. Experience in data management and analysis. Knowledge of research methods 
and techniques is a plus. Familiarity with USAID’s objectives, approaches, operations, and policies, 
particularly as they relate to evaluation is a plus.
18 See ADS mandatory reference 201maa, available at 
https://www.usaid.gov/sites/default/files/documents/1870/201maa.pdf 
19
Position Description: The Expert will be responsible to clean and analyze the SisPuentes data, 
and prepare any visualization of the data, if necessary.
Level of Effort (LOE)
Position LOE
Team Lead 30 
Data Specialist 30 
20
APPENDIX 2: TABLES AND GRAPHS
2.1 Exploring the linearity of the relationship between logarithms of salaries and 
potential predictors
We examined data on 3,932 employed youth with complete information on the attended and 
course files. Figure 1 shows the distribution of monthly wages by each of the following variables: 
(1) number of approved years of formal education, (3) age at time of enrolment in BE; (3) cohort 
of admission into BE; (4) number of courses completed; (5) number of courses with improvement 
level; (6) type of employment (full-time, part-time, other); (7) type of contract (monthly, daily, 
hourly, other); (8) type of qualification (formal or informal); (9) sex; (10) area of residence (urban 
or rural); (11) region of residence (San Salvador metro area or other); and (12) economic sector. 
A degree 3 spline curve was added to each graph.
Figure 1: Youth Monthly Wages by Different Variables
 
21
Reading from left to right, the data in Figure 1 indicate that: (1) salaries do not vary by sex; (2) as 
expected, salary variability increases with education; (3) there is more variation in the salaries of 
those working as employees than among entrepreneurs; (4) salaries are higher for youth hired 
for either temporary or full-time jobs; (5) age does not seem to impact on salary differentials; 
and (6) salaries do not vary across economic sectors. There is a slightly higher variation in the 
manufacturing and information-technology sectors, and a larger concentration above the mean in 
the salaries for the energy sector; (7) salaries do not differ across different forms of payment. 
However, the salaries of youth paid on a monthly basis show more variability than other payment 
periods; (8) there seems to be a weak nonlinear relationship between salaries and number of 
courses attended. The salary declines as the number of courses increases from one to three 
when it shows a slight increase; (9) youth who either completed or recorded an improvement in 
their last course seem to have higher salaries than those who either did not complete or did not 
have an improved performance in it; (10) there is not clear relationship between the cohort a 
youth was part of and salary; (11) there was lesser variability in the salaries of youth attending a 
course included in the INSAFORP curricula than those attending other courses.
2.2 Fit diagnostics for the MARS model for the logarithm of wages
Figure 2: Model Diagnostics -MARS
22
23
APPENDIX 3: BIBLIOGRAPHY OF DOCUMENTS REVIEWED
Altonji, J.G., A.A. Smith Jr. & I. Vidangos (2013). Modeling earnings dynamics. Econometrica 81(4): 
1395-1454. 
Becker, G.S. (1993), Human Capital. A Theoretical and Empirical Analysis with Special Reference to 
Education. 3rd ed. Chicago, IL: The University of Chicago Press.
Blundell, R.L., L. Dearden, C. Meghir & B. Saniesi (1999). Human capital investment: The returns 
from education and training to the individual, the firm and the economy. Fiscal Studies 20(1): 1-
23. 
Friedman, J.H. (1991). Multivariate adaptive regression splines. The Annals of Statistics 19 (1): 1-67. 
Ishikawa, M. & D. Ryan (2002). Schooling, basic skills and economic outcomes. Economics of 
Education Review 21(3): 231-243. 
Hastie, T. J., R. J. Tibshirani, & J. H. Friedman (2009). The Elements of Statistical Learning: Data 
Mining, Inference, and Prediction. 2nd ed. New York, NY: Springer-Verlag.
Mincer, J. (1974). Schooling, Earnings, and Experience. New York, NY: Columbia University Press. 
Mincer, J. (1989). Human capital and the labor market: A review of current research. Educational 
Research 18(4), 27-34. 
USAID El Salvador, 2014, Country Development Cooperation Strategy 2013-2017 (CDCS), 
https://www.usaid.gov/sites/default/files/documents/1862/CDCS%203-3-
14%20Public%20Version%20FINAL.pdf , accessed December 2020.
USAID El Salvador, undated, Country Development Cooperation Strategy 2020-2024 (CDCS), 
https://www.usaid.gov/sites/default/files/documents/CDCS-El_Salvador_PUBLIC.pdf, accessed 
January 2021.
USAID, undated, The Journey to Self-Reliance, https://www.usaid.gov/selfreliance, accessed January 
2021. 
USAID El Salvador, undated, Fact Sheet USAID Bridges to Employment, 
https://www.usaid.gov/sites/default/files/documents/1862/Fact_Sheet-
_USAID_Bridges_to_Employment_Final.pdf , accessed January 2021.
Wooldridge, J.M. (2016). Introductory Econometrics. A Modern Approach, 6th ed. Boston, MA: 
Cengage Learning.