The Impact Of Vocational Training On Labor Market Outcomes In The .

11d ago
12 Views
2 Downloads
645.54 KB
25 Pages
Last View : 1d ago
Last Download : 4d ago
Upload by : Dahlia Ryals
Transcription

THE IMPACT OF VOCATIONALTRAINING ON LABOR MARKETOUTCOMES IN THE PHILIPPINESPaul Vandenberg and Jade LaranjoNO. 621October 2020ADB ECONOMICSWORKING PAPER SERIESASIAN DEVELOPMENT BANK

ADB Economics Working Paper SeriesThe Impact of Vocational Training on Labor Market Outcomesin the PhilippinesPaul Vandenberg and Jade LaranjoNo. 621 October 2020ASIAN DEVELOPMENT BANKPaul Vandenberg ([email protected]) is a senioreconomist and Jade Laranjo ([email protected]) is anassociate economics analyst at the Economic Researchand Regional Cooperation Department of the AsianDevelopment Bank.

Creative Commons Attribution 3.0 IGO license (CC BY 3.0 IGO) 2020 Asian Development Bank6 ADB Avenue, Mandaluyong City, 1550 Metro Manila, PhilippinesTel 63 2 8632 4444; Fax 63 2 8636 2444www.adb.orgSome rights reserved. Published in 2020.ISSN 2313-6537 (print), 2313-6545 (electronic)Publication Stock No. WPS200274-2DOI: http://dx.doi.org/10.22617/WPS200274-2The views expressed in this publication are those of the authors and do not necessarily reflect the views and policiesof the Asian Development Bank (ADB) or its Board of Governors or the governments they represent.ADB does not guarantee the accuracy of the data included in this publication and accepts no responsibility for anyconsequence of their use. The mention of specific companies or products of manufacturers does not imply that theyare endorsed or recommended by ADB in preference to others of a similar nature that are not mentioned.By making any designation of or reference to a particular territory or geographic area, or by using the term “country”in this document, ADB does not intend to make any judgments as to the legal or other status of any territory or area.This work is available under the Creative Commons Attribution 3.0 IGO license (CC BY 3.0 o/. By using the content of this publication, you agree to be boundby the terms of this license. For attribution, translations, adaptations, and permissions, please read the provisionsand terms of use at https://www.adb.org/terms-use#openaccess.This CC license does not apply to non-ADB copyright materials in this publication. If the material is attributedto another source, please contact the copyright owner or publisher of that source for permission to reproduce it.ADB cannot be held liable for any claims that arise as a result of your use of the material.Please contact [email protected] if you have questions or comments with respect to content, or if you wishto obtain copyright permission for your intended use that does not fall within these terms, or for permission to usethe ADB logo.Corrigenda to ADB publications may be found at B recognizes “Korea” as the Republic of Korea.The ADB Economics Working Paper Series presents data, information, and/or findings from ongoing research andstudies to encourage exchange of ideas and to elicit comment and feedback about development issues in Asia andthe Pacific. Since papers in this series are intended for quick and easy dissemination, the content may or may not befully edited and may later be modified for final publication.

G THEORY AND EVIDENCE2III.TECHNICAL AND VOCATIONAL EDUCATION AND TRAININGIN THE PHILIPPINES4IV.MODELS AND METHODOLOGIES5V.CHARACTERISTICS OF THE SAMPLE7VI.RESULTS10VII.CONCLUSION15REFERENCES17

TABLES1Summary Statistics of Selected Profile by Highest Education Level82Characteristics of Workers Who Completed Secondary School and Post-SecondaryTechnical and Vocational Education and Training93Share of Persons Employed in the Sample104Determinants of Wage115Number of Post-Secondary Technical and Vocational Education and Trainingby Highest Year Completed126Logistic Regression for the Propensity Score (Technical and Vocational Educationand Training versus Other Levels of Education)137Differences in Log Monthly Wage between Technical and Vocational Education andTraining and Other Levels of Education Using Propensity Score Matching138Employment Probability, Marginal Effects14

ABSTRACTThe paper analyzes the labor market outcomes of graduates of post-secondary technical andvocational education and training (TVET) in the Philippines. Based on human capital theory, ourguiding assumption is that investment in education is rewarded through higher wages and a greaterlikelihood of being employed. Using household data for 2015–2016, the results show significantlyhigher wages for TVET graduates relative to those who entered the job market with a secondary schooleducation or below. However, individuals who both trained in TVET and pursued tertiary (university)education tend to have a lower wage than those with secondary school education or below. Thiscounterintuitive result is explained by the tendency for such dual-level individuals to complete thelowest level of TVET. Another result is that TVET graduates are more likely to be employed than boththose who only studied at secondary school or below and those who studied at the tertiary level. Theresults are generated from a linear regression model that corrects for sample selection intoemployment and from the use of propensity score matching which addresses selection into TVET.Keywords: labor market, Philippines, skills training, TVETJEL codes: I26, J24

The Impact of Vocational Training on Labor Market Outcomes in the PhilippinesI.1INTRODUCTIONEducation allows people to secure employment and pursue an occupation. It provides an “in” to thelabor market and an opportunity to earn a decent wage or salary. This may not be the sole reason foreducation as it also provides personal development and the opportunity to pursue an interest.Nonetheless, for most people, employment is a key reason they expend time, effort, and money,particularly on post-compulsory education at college or university.It is possible, however, that education does not result in the intended labor market outcome.On the demand side, an individual may pursue an area of study for which employers are not hiring. Thejob seeker falls victim to what is known as a horizontal skills gap. On the supply side, the poor quality ofeducation, in terms of curriculum, facilities, or teaching, may result in very little learning taking place.Employers may perceive that graduates have limited knowledge and few skills, and their diploma ordegree is of little value. This is part of the vertical skills gap. An employer may see little difference inhiring a graduate in a specialized field from someone with basic education, as both need furthertraining by the employer. While the above scenarios would seem to go against the logic of educationfor-employment, the possibility of weak labor market outcomes is very real.1Governments and development partners have sought to increase enrollment, particularly inprimary and secondary education, under the banner of “education for all” and an effort to achieve theMillennium Development Goal of universal access to primary education. As a result of these efforts,enrollment, which is considered a quantity measure, has risen considerably in most parts of thedeveloping world. The focus has now shifted to the quality of education under a new concern thatthere is a “learning crisis” in which children are in the classroom but not learning very much. Theavailability of international test results, particularly for 15-year-olds under the Programme forInternational Student Assessment (PISA), has provided the evidence underlying the learning crisis.The lack of similar test results makes it more difficult to verify a learning crisis at the postsecondary level. However, as this level of education prepares students more directly for the labormarket, labor market outcomes may be a more appropriate indicator of the quality of education.2 Ineffect, employers become the judge. This quality check may be particularly relevant for technical andvocational education and training (TVET) as it has an equal and possibly more direct intent to educatefor the labor market than the university system. In developed countries, TVET is generally effective insecuring employment because of both the quality of instruction and the focus on skills needed in theworkplace. In developing countries, the quality of TVET may suffer from the same deficiencies thathave caused a learning crisis in primary and secondary education.This paper examines the labor market outcomes of post-secondary TVET in the Philippines.The outcome is defined as both the level of wages or salary (wage effect) and ability to secure a job(employment effect). The research can make an important contribution to the question of whethersignificant reform is needed to the TVET system. Very weak labor market outcomes might suggest aneed for major reforms, whereas decent outcomes would indicate that modest improvements may besufficient. The succeeding sections of the paper are structured as follows. The second section reviews12This study focuses on the supply side, but certainly the demand by employers for people with education and skills will be adeterminant of whether trained graduates secure employment and are offered good wages.However, if the market is inefficient at matching skills supply with demand, then good quality graduates may not find goodpaying jobs that utilize their skills.

2ADB Economics Working Paper Series No. 621some of the literature on the topic, notably the empirical work. The third section outlines the Philippineeducation system with an emphasis on TVET. The fourth section presents the model and empiricalmethods used, and the fifth provides the education and labor force characteristics of the sample. Thesixth section presents and discusses results. A final section concludes.II.EXISTING THEORY AND EVIDENCEThe concept of human capital was developed more than half a century ago by Schultz (1961), Becker(1964), and Mincer (1974) and remains a powerful conceptual framework to understand the linkbetween education and work. Human capital is the aggregate of knowledge, skills, abilities, andaptitudes that an individual possesses and can be used to contribute to productive work. Greaterhuman capital contributes to increased and better output, in terms of efficiency and the quality ofwork, and thus can command a higher wage. Individuals can accumulate human capital througheducation, training, and experience. Because human capital can be costly to accumulate, there is arelationship between the spending (i.e., investment) in education and the return through the wagespaid to a worker in employment. In this way, knowledge is viewed as “capital” just like physical capital(plant and machinery) or financial capital, and thus links investment and return.TVET can be categorized into four main types according to its location within the educationand employment system: (i) secondary (high) school vocational education; (ii) post-secondarytechnical and vocational education; (iii) short-course training as part of active labor market policies(ALMP) targeting youth and the unemployed; and, (iv) employer-sponsored training, also known as inservice or on-the-job training.3 The length and intensity can differ across types and, for assessing theimpact, the control group of untrained individuals can be different. For example, students whocomplete secondary TVET are compared against their peers in the secondary school academic stream,whereas post-secondary TVET graduates are compared against secondary school graduates who donot proceed to further education. Creating an appropriate control group is one of the keymethodological challenges in assessing impact of TVET and has prompted the use of experimentaldesigns and randomized control trials, where possible (ADB 2015). Impact is normally assessed interms of two labor market outcomes, notably the wage effect and the employment effect. The basicidea is that those with TVET receive a higher wage and are more likely to be employed than those(control group) who did not pursue TVET.Bettinger et al. (2010) assess the differences between those who received funding vouchersfor private vocational education from those who did not. Those who received the vouchers, whichwere allocated by lottery, were more likely to stay in vocational school (and not transfer to academicprograms). They were also more likely to complete secondary school and score better on final examsthan those who did not receive a voucher. The study did not specifically address whether vocational3The four-part categorization is based on the authors’ knowledge of the different types of TVET. Studies on the first threemethods are cited in the rest of the section. Evaluation of the fourth type (on-the-job training) is difficult and there arefew robust studies. Note that the term ‘on-the-job training’ can have different meanings. Strictly speaking, it is trainingdone while working, usually supervised by a mentor. Employers may, however, take workers away from their work (theproduction line, the store front, etc.) and provide training in another part of the facilities or send workers to train with atraining provider offsite.

The Impact of Vocational Training on Labor Market Outcomes in the Philippines3education is better or worse than pursuing the academic streams but did conclude that a privatevocational education may generate better academic results than public academic programs.Malamud and Pop-Eleches (2010) exploited a natural experiment in Romania in which 2 yearsof general education were added and vocational education was shortened. The study found nodifference in the wage effect and the employment effect between students affected by the changecompared to those who attended secondary school prior to the change. Field et al. (2019) estimate thewage and employment differences of those who studied in competitive (oversubscribed) Mongoliansecondary TVET programs relative to those who studied in other vocational programs, and some whopursued the academic stream. In a randomized controlled experiment, those admitted to thecompetitive schools were selected by lottery from a larger set of applicants who met the admissioncriteria. Students (male and female) admitted to the competitive schools were more likely to gainemployment than those not selected. Women from the competitive schools received higher wagesthan those who were not selected. However, men selected received lower wages, although the resultwas not significant.Olfindo (2018) studied the impact of vocational education on wages in the private sector,using labor force survey data from the Philippines for 2015. About 80% of those who received TVETdid so at the post-secondary level. TVET graduates earned higher wages than those who onlycompleted secondary school. However, for students who studied both TVET and at the tertiary level(university), wages were lower than for those who undertook only tertiary studies. The author suggeststhere may be a “penalty” for adding TVET to university education. Propensity score matching (PSM)generated similar results, although statistical significance was not achieved, and balancing propertieswere not met under certain specifications. For example, those with TVET earned higher wages thanthose with only secondary school education, and while the results were statistically significant, a test ofbalancing properties was not satisfied. The employment effect was not tested. Choi (2016) alsoassessed the impact of TVET on wages in the Philippines using data from the labor force survey of2014 and found similar significant effects for TVET graduates relative to those who completed onlysecondary school. However, the estimates did not control for economic background, parents’education, or other factors that might account for differences in ability.Lee, Han, and Song (2019) studied the impact of nonformal job-related education (consideredvocational education) on wages and employment in the Republic of Korea. Such education wascorrelated with higher wages and a greater likelihood of employment relative to upper secondaryschool graduates. The education variable was significant across a range of model specifications.Several key studies have assessed short-course vocational programs that target unemployedyouth. Attanasio, Kugler, and Meghir (2011) and Attanasio et al. (2015) studied a program in Columbiathat offered 3 months of classroom training followed by 3 months of on-the-job training. More than ayear after completion, women were 7% more likely to be in paid employment and earned 20% morethan those who were not trained. For men, differences were not statistically significant. Card et al.(2011) used a randomized control trial to test a similar program in the Dominican Republic. About 780youth from low-income households were given 350 hours of training, followed by an internship at aprivate business. Subsequent employment earnings were about 10% higher than those of a controlgroup. However, differences in the employment effect between the two groups were not statisticallysignificant.Hicks et al. (2016) evaluated the impact of vocational education training vouchers onrandomly selected youth in Kenya. 42% of the individuals did not complete the course, despite beingprovided with adequate funding. Under various model specifications, the wage and employment

4ADB Economics Working Paper Series No. 621effects were not significant. McKenzie (2017) reviewed 12 studies of eight countries—two of whichwere in Asia: Turkey and India—and found positive, although small wage and employment effects forthose who were trained. Hirshleifer et al. (2016) studied the impact of vocational training on theunemployed in Turkey. About 30 hours of training were provided per week over a 3-month period.Trainees subsequently earned 5.6% more than similar youths in a control group and had a 2% higherrate of employment. The results, however, were not statistically significant. Kluve et al. (2016)reviewed 74 studies of youth programs that included skills as a key component and found that just overa third (37%) showed a significant positive employment effect. Furthermore, the researchersconcluded that “much of the difference in performance [among youth employment programs] seemsto be related to design and implementation factors, as well as the characteristics of the country and thebeneficiaries” (Kluve et al. 2016).This short review of empirical studies suggests that TVET generally has positive wage andemployment effects. However, many of the results are not statistically significant or the significance isnot obtained for both the wage and the employment effect. Short-course and longer-course TVETprograms can be very different and should be considered separately.III.TECHNICAL AND VOCATIONAL EDUCATION AND TRAININGIN THE PHILIPPINESPrior to 2011, the Philippine education system consisted of 6 years of primary (elementary) schoolfollowed by 4 years of secondary (high) school, with students graduating at the age of 15 or 16. It wasthe sole system of education in Asia with only 10 years of combined primary–secondary education andone of only three in the world. A major reform was undertaken to lengthen the duration of schooling.Kindergarten was added in 2011 and then two further grades in 2016 and 2017. The current system has6 years of elementary school, 4 years of junior secondary school and 2 years of senior secondaryschool. Students graduate at age 17 or 18. The reform brought the length of schooling in line with K-12systems common in other countries.Understanding the prereform system is important because the data used for this study wasobtained from households during 2015–2016. Under the old system, there was no vocational stream atsecondary school.4 Students completed Grade 10 and then had three options: attend a TVET programat a college or related institution; enroll at university or other tertiary institution; or enter the jobmarket. At that time and currently, the TVET system is managed by the Technical Education and SkillsDevelopment Authority, which accredits programs and organizes the system of skills certification oftrainees. After training, students can sit an examination to obtain a National Certificate (NC), which isgranted at levels I, II, III and IV, with the latter being the most advanced. Vocational training can alsolead to a Certificate of Competency, for those who did not complete secondary school. There are alsoa variety of TVET opportunities that are not part of the NC system but lead to qualificationsrecognized by the industry and government. TVET is considered nontertiary education, whereasuniversity and related higher-level education are classified as tertiary. It is also possible for an individualto pursue both TVET and tertiary education. We mention this case of combined education because ofthe interesting and often counter-intuitive results that it throws up later in the paper.4In the current system, a vocational track is offered in senior secondary school, leading to NCs I and II.

The Impact of Vocational Training on Labor Market Outcomes in the PhilippinesIV.5MODELS AND METHODOLOGIESThe objective is to test the correlation between TVET and outcomes in the labor market. There aretwo outcomes of interest: the wage effect and employment effect (i.e., the likelihood of beingemployed). We begin by specifying a standard Mincerian wage model as follows:𝑂 𝛽 𝛽 𝐸 𝛽 𝑋 𝑢(1)where 𝑂 is either the wage or the employment effect, 𝐸 is educational attainment, and 𝑋 is workexperience. Following standard notation, the 𝛽s are the variable coefficients, 𝜇 is the error term, andthe subscript 𝑖 represents the individuals. Education can be a continuous variable for years ofeducation or, as used here, a discrete variable for the type or level of education. We have six types:(i) secondary school or below,5 (ii) TVET, (iii) partial tertiary education without obtaining a diploma ordegree, (iv) completed tertiary, (v) both TVET and partial tertiary, and (vi) both TVET and completedtertiary. Education is included in the model as dummy variables. Experience is an individual’s currentage minus the age at which the individual completed education. As such, our equation expands to:𝑂 𝛽 𝛽 𝑉 𝛽 𝑃 𝛽 𝑇 𝛽 𝐼 𝛽 𝐶 𝛽 𝑋 𝑢(2)where 𝑉 is TVET, 𝑃 is partial tertiary, 𝑇 is completed tertiary, 𝐼 is TVET and partial (incomplete)tertiary, and 𝐶 is TVET and completed tertiary. The 𝑖 subscripts are dropped for simplicity ofpresentation. Each of the education variables takes a value of 1 for that individual type of educationand 0 otherwise. The base type is secondary school or below.Such a specification may generate biased results due to omitted variables and result ininaccurate correlations between the education variables and the wage. Omitted variables wouldinclude other characteristics, besides education and experience, that affect the wage a worker receives.The other characteristics can include intelligence (or ability) and effort. A worker may be paid a higherwage because of her educational qualification but also because she is perceived by an employer asbeing more intelligent or as someone willing to work harder. If we exclude these variables, the effectsare incorrectly captured by the (larger) size of the education coefficients. Conversely, if omittedfactors are negatively correlated with education, the education variable would have a downward bias.6Variables for intelligence and effort are not normally available. For ability, it is best to use testscores, such as secondary school examination results for both those who entered the workforcedirectly after school and those who continued to pursue TVET or tertiary education. Unfortunately, wedo not have test results in the dataset. As a proxy, the educational attainment of the individual’smother is used. Fortunately, the survey did ask individuals whether they work hard to complete tasksand the responses are used as the variable to gauge effort. As such, the specification in equation (2) isexpanded as follows:𝑂 𝛽 𝛽 𝑉 𝛽 𝑃 𝛽 𝑇 𝛽 𝐼 𝛽 𝐶 𝛽 𝑋 𝛽 𝐴 𝛽 𝐹 𝑢56(3)Secondary school or below: denotes those whose highest educational attainment was (i) completed or did not completesecondary school, (ii) completed or did not complete primary (elementary) school, and (iii) did not attend school at all. Ineffect, it is any individual who did not attend TVET or tertiary education.This might occur if TVET acts as an absorber of those who are not able to secure employment at a decent wageimmediately after completing school.

6ADB Economics Working Paper Series No. 621where 𝐴 represents ability and 𝐹 represents effort. Gender and control variables for sector and regionare also included.As noted, the data used for wage analysis is a subsample of the full sample and therefore is notrandom. Persons who pursued TVET or another type of education but were not working at the time ofthe survey did not have a wage to report. This problem of nonrandomness was identified by Heckman(1976, 1979). As such, we use a two-step Heckman correction procedure. The first step involves theestimation of the probability of employment (being employed) using a probit model. The result is usedto compute the Inverse Mills Ratio which is then included as a regressor in the ordinary least squares(OLS) wage equation (second step) to control for possible bias due to nonrandomness.In addition to dealing with nonrandom selection into employment, we have also sought toaddress nonrandom selection (participation) in TVET. As such, the second technique used for thewage effect is PSM (Rosenbaum and Rubin 1983). PSM is a common technique in education andemployment studies to create comparable control and treatment groups (Graham and Kurlaender2011). For this study, individuals with TVET (treatment group) are matched to individuals with similarcharacteristics except that they did not pursue TVET (control group). The procedure requires first thata propensity score is assigned to each individual by estimating the probability of obtaining TVET usingthe logit model. Then the scores of control and treatment groups are matched using a kernel matchingmethod and the difference between groups is estimated. Variables for experience and gender areincluded as well as proxies for ability and effort, and control variables for industry. Considering thevariations in the covariates between groups, we applied the procedure separately to the threeeducation groups of the individuals: completed secondary, partial tertiary, and completed tertiary.Lastly, for each education group, we stratified the individuals into blocks. Then a test for balancingproperty was conducted for each block to determine whether the control and treatment groups weresufficiently similar to provide an accurate basis for comparison. The balancing property is satisfied ifRubin’s B, which is the absolute standardized percentage difference of the means of the linear index ofthe propensity score in the treated and (matched) control group, is less than 25%. It is also satisfied ifRubin’s R, which is the ratio of the treated to the (matched) control variances of the propensity scoreindex, is between 0.5 and 2.0 (Rubin 2001).For the employment effect, the analysis uses a multinomial logit model as follows:𝑃𝑟𝑜𝑏 𝑂 𝑗 𝑍 (4)where 𝑂 is the dependent variable employment with three states 𝑗 which are: employed; unemployed(not working but seeking work); and not participating in the labor market (not working, not wanting towork). 𝑍 are the same independent variables as in the wage effect analysis but including variables onage, having spouse, ability relative to father’s education, and socioeconomic status.

The Impact of Vocational Training on Labor Market Outcomes in the PhilippinesV.7CHARACTERISTICS OF THE SAMPLEThe data is from the STEP Skills Measurement Household Survey 2015–2016 for the Philippines. Itprovides information on education, employment, wages, and other characteristics for 3,000individuals between the ages of 15 and 64. The survey applied a four-stage sampling design. In the firststage, the primary sampling units which are composed of barangay segments in urban regions wereselected through implicit stratification and systematic sampling. Within each selected primarysampling units, secondary sampling units, which are the dwellings, were systematically picked out. Andthen in each selected dwelling, a household was randomly chosen with equal probability. Finally, anindividual aged 15–64 years old was randomly selected with equal probability from each selectedhousehold to complete the survey questionnaire (Pierre et al. 2014).The data provide details on education that include not only the highest qualification obtained(certificate or degree) but also the years of study, which is important for those who pursued but didnot complete a tertiary degree. For TVET, there is a breakdown by NC level. In addition, there is acategory of Other TVET that does not lead to an NC but is recognized by the government and/orindustry. A substantial portion of those who completed TVET are in this category. All those whoundertook TVET in our dataset completed secondary school. The breakdown by NC level and theinclusion of a variable for effort are the chief advantages of the dataset over others, notable the laborforce survey.The full sample of 3,000 is used in the analysis of the employment effect. A subsample is usedfor the analysis of the wage effect as follows. Only those who are employed can provide data on wagesand therefore those who were not employed are excluded. The self-employed were also excludedbecause it is difficult to determine a “wage” that is distinct from the income of the business. Also, theself-employed tend to introduce outliers, with very high earnings in some cases which are generatedfrom entrepreneurial talent rather than skills acquired through education. As a result, the subsamplefor the wage effect includes a total of 980 individuals.

The Impact of Vocational Training on Labor Market Outcomes in the Philippines 1 I. INTRODUCTION Education allows people to secure employment and pursue an occupation. . technical and vocational education; (iii) short-course training as part of active labor market policies (ALMP) targeting youth and the unemployed; and, (iv) employer-sponsored .