«SKILLS MISMATCH IN LATVIAN MANUFACTURING SECTOR Authors: Ilze Zumente and Kārlis Putriņš ISSN 1691-4643 ISBN 978-9984-842-28-8 November 2011 Riga ...»
A more recent paper by Lindley and McIntosh (2010) adds to the existing literature by measuring the permanence of the overeducation and find that approximately 50 % of workers being over-educated in the 1991 were still over-educated in 2005 and only 25% had improved their situation in terms of a better educational match suggesting that the effect of overeducation seems to be permanent (Lindley & McIntosh, 2010).
All in all, overeducation seems to lead to decreased job satisfaction and relatively lower wages in form of wage penalty (Vaisey, 2003).
Undereducation, on contrary, is reported to lead to lower productivity of the firm and thus again to lower remuneration. Groenevald and Hartog (2003) argue that undereducation in the same way as overeducation leads to wage penalties and disturbed career development (Groenevald & Hartog, 2003). The view is supported by McIntosh and Steedman in the research about low skill workers (McIntosh & Steedman, 2000).
Ilze Zumente and Kārlis Putriņš 11 A closer to home study on Estonia done by the European Central Bank for the period from 1997 to 2003 finds large wage penalties associated with educational mismatch as estimated by the Estonian Labor force survey. Authors estimate a regression using log of hourly wage rate as the dependent variable and explanatory variables for overeducation, gender and age and achieve robust results for various model specifications (Lamo & Messina, 2010).
Study about Latvian labor market conducted in 2006 seeks to conclude whether the education programs provided in Latvia correspond to the needs expressed by the labor market.
The study is broad and includes all the major industries of the market. The conclusion is drawn that practical skills are lacking in the most of the cases while the skills that are demanded the most are motivation, professionalism and communication abilities. The level of acquires skills is mostly higher than required by employers, which leans in the direction of the phenomena of overeducation (Sloka, 2007). A similar research by Juris Krūmiņš states that there exist serious problems in the vocational school graduates preparation for the labor market demands- the latest technologies are not used sufficiently; the specialization programs are too narrow and the graduates are mostly psychologically unprepared for the labor discipline (Krūmiņš, 2007).
Skills mismatch impact on wages Education and skills mismatches are often reported to cause wage gaps in form of “wage penalties” for the employees not working in the job appropriate to their education and skills. A serious amount of literature has been dedicated to this topic as the direct contribution from skills mismatch on the economy might be approximated in this manner.
Allen and van der Velden (2001) using data for 11 European countries and Japan conclude that education mismatches is a good predictor of wages, while skills mismatches tend to have a smaller direct impact on the wage rate. When taking into account both effects, both remain significant while the largest impact is created by the educational mismatches. Approval to the assignment theory is thus provided by the authors (Allen & van der Velden, 2001). The results are consistent with Falter (2009), who examines 7 European countries in three measures of skills: prose literacy, document reading literacy and quantitative literacy. The negative skills mismatch impact on wages is present in the analysis (Falter, 2009). Further approval for the theory is provided by Hoppe et al. (2003), who examine the German labor market from 1984Hoppe, Muyken, & Rieder, 2002) and Muysken et al.(2002) for USA labor market from 1986 to 1996 ( (Muysken, Weissbrich, & von Restorff, 2002).
Ilze Zumente and Kārlis Putriņš 12 A contrary insight is provided by Di Petrio et al. (2003), who by examining the Italian graduates find little evidence to the assignment theory and very weak wage effect following the skills mismatch (Pietro & Urwin, 2006). Weak effect of educational and skills mismatch is approved also by Korpi et al. (2006), who in the research for Sweden for the time period from 1974-2000 find little approval to the skills mismatch effect on the wage rates and its growth over time (Korpi & Tahlin, 2006).
A paper examining concrete skills and also estimating the economic effect on wages has been developed by Sgobbi and Suleman (2009) for the Portuguese banking sector. Authors use a survey where banking sector employees are evaluated in industry specific skills and later divided according to the performance in the skills in order to distinguish between various skills (mis)match types. A significant impact of skills mismatch on wages is documented (Sgobbi & Suleman, 2009).
A study about Latvian Labor market conducted in 2006, which tries to document the various factors affecting wages, pay the most attention to the two broad factors impacting the wages of concrete employees- individual characteristics including demographic information and education and firm characteristics that currently employs the person. The paper documents strong positive effect of education, larger enterprises and language skills as proxied by ethnicity (Zepa & Hazans, 2006). No effect of skills mismatch is measured in the paper.
All in all, although the majority of the academic research has documented negative effects on the wage rate following skills and education mismatch, there exists also literature, which finds no support for the hypothesis. It can also be argued that the evidences differ across countries as the contradictory views are not representing the same regions.
Skills mismatch in the manufacturing sector Although manufacturing sector can be seen as one of the most important ones in terms of the large export potential and contribution to the GDP, there is little evidence about the organizational problems in terms of skills mismatch particularly in this sector.
David Howell (1999) examines the real weekly wages of the production workers in the 1980s. He finds sharp decline since 1970 in the wages of low-skill workers that are not connected to technology invention, but with cost reduction and people willingness to work. By concentrating particularly to the manufacturing sector, he finds less mismatch between skills demanded and skills supplied, but more between skills demanded and wages paid leading to Ilze Zumente and Kārlis Putriņš 13 implications that education and training system should be improved, not the skills (Howell, 1999).A similar time period is investigated by Berman et al. (1994), who use the Annual Manufacturing survey in USA for 1979 to 1989 in order to find a shift towards skilled labor force in the manufacturing sector (Berman, Bound, & Griliches, 1994).
More recent results are presented by Bjørnstad (2000), who has undertaken an econometric analysis based on Norwegian manufacturing sector. Author divides the sample in 5 educational groups according to the level of education and estimates an analysis including producer price index, labor productivity, unemployment rate measured in logarithmic scale and education-specific unemployment binary variables. His analysis shows that skills mismatch has increased and is predicted to be permanent because of low focus on skills mismatch in the wage setting mechanisms (Bjørnstad, 2000).
A unique approach is created by Peters (2000), who develops a skills-mismatch index and applies it to the manufacturing sector in Missouri, USA. Occupational classes followed by estimated educational attainment is summarized in order to gain a skill level needed for a concrete occupation (ranging from 1 to 3). Mismatch index is afterwards expressed as the square sum of percentage of population having the concrete level of skills reduced by the percentage of workers with the concrete skill level in the industry. No specific skills are mentioned in the methodology. The worst skills mismatch is found to occur in industries like transportation and computer equipment production proving that particularly these industries that incorporate lots of new technologies also present the highest share of skills mismatch (Peters, 2000). The measure by Peters is later also employed by the International Monetary Fund measuring the overall skills mismatch in the U.S. after the recession (IMF, 2010).
To sum up, there is not sufficiently developed literature concerning the skills mismatch problem in the manufacturing sector. By taking into account the sector specifics in terms of high technology development and frequent possible changes, the sector is very likely to undergo unemployment problems caused by skills mismatch. This paper therefore aims to fill the gap in the literature and provide an up-to-date data and analysis.
Ilze Zumente and Kārlis Putriņš 14
Questionnaire The quantitative study of the issue has been done via survey distributed to workers employed in the lowest levels in the manufacturing sector in Latvia. The surveys have been done anonymous in order to avoid overconfidence and gain possibly true and unbiased results. The survey was translated and distributed also in Latvian and Russian in order to be understandable among the workers employed in Latvia. The English version of the questionnaire can be found in the appendices 1 and 2.
The first part of the survey is based on the methodological questions developed by Allen and van der Velden (2001) and also used by Green and McIntosh (2007) and includes following statements based on which type of skills mismatch may be concludedQ1: “In my current job I have enough opportunities to use the knowledge and skills that I have”, Q2: “In my current job I apply the skills and knowledge that I have” Q3: “I would perform better if I had additional skills or knowledge” Q4: “I would like to enhance my knowledge and master new skills” (Allen & van der Velden, 2001).
The agreements to the statements are planned to be self-assessed and evaluated on the 5 points Likert scale. The types of skills mismatch or skills match then could be obtained via cross sectional analysis of the answers provided and measuring underutilization of skills and skills deficit summarized in the table below.
Skills surplus occurs when there is no need for additional skills and the use of the previous knowledge is low. Skills shortage situation is present if there is a need for additional skills although much of the previous knowledge is currently used. Employees have wrong skills if they cannot use the previous knowledge and need additional skills to cope with the job tasks.
Finally, skills match is said to be present if an employee does not need additional skills and has low skill underutilization in terms of usable previous knowledge (Allen & van der Velden, 2001). Based on the answers to the previously stated questions number 2 and 3, the mismatch type can be concluded. Answers to questions number 1 and 4 will be later used to additionally conclude trends in the skill usage opportunities in the industry. The answers in range between 1 and 3 are considered to be disagreeing to the statement, while 4 and 5 point evaluation represents Ilze Zumente and Kārlis Putriņš 15 agreeing to the presented statement. Such a point distribution is based on previous studies in the field (Allen & Vries, 2004).
The second part of the survey concentrates on specific skills required in the manufacturing sector. A self-assessment of concrete skills has been asked from the respondents in terms of how good one considers himself at the concrete skill as measured by 5 point scale.
Although it might be argued that self assessment might result in bias results because of the overconfidence of the respondents, the approach has been successfully used in previous academic work for instance by Allen and van der Velden (2001). Furthermore, as surveys have been anonymous, people should have no incentive to lie or overestimate their true skills. A similar list of the skills included in the survey is also used in a research by Sgobbi & Suleman (2009). The skill list includes skills like Latvian and Russian language, communication, cooperation, teamwork, responsibility, technical knowledge, standards and procedures, tasks planning, time management, learning skills, adaptability and problem solving. The skill set provided in the questionnaire has been discussed in the interviews with industry professionals as well as partly adjusted to the previous research done by Sloka (2007).
The third and final part of the survey includes general demographic questions like age and gender followed by more concrete questions specifying the industry, education level, marital status and number of children, experience in the labor market, average working time and the wage rate. The demographical questions of the third part in the survey have been used to constitute the basis for the econometric analysis planned to measure the economic impact of skills mismatch.
Before each of the main survey distribution session, another survey was given to the employer or a supervisor in the company in order to detect which skills are the most demanded Ilze Zumente and Kārlis Putriņš 16 by the employer’s side. Afterwards, a comparison and analysis was undertaken in order to provide statistics about the most commonly demanded skills and the skills developed by the employees. Employer survey can be found in appendix 2.
Regression analysis As argued before skills mismatch phenomenon, might result also in direct wage effects, which on aggregate may lead to lower economic returns in general for people suffering from one of the skills mismatch types.
In order to measure this impact from skills mismatch on the economic returns, a model developed by Duncan and Hoffman in 1981 and later used by Sgobbi and Suleman (2009) was used. The model estimates an OLS regression in order to capture the explanatory effect of various skills mismatch types on the wage rate (Duncan & Hoffman, 1980). The model as
specified by the previous academics was estimated as follows:
Ilze Zumente and Kārlis Putriņš 17 Strategy evaluation The last goal of the bachelor thesis was to analyze and evaluate the strategies employed by Latvian Government to lessen the skills mismatch. As it is argued that in the Latvian case the mismatch may stem from the unprofessional and old-fashion vocational school system, in depth interviews were conducted in order to find out the different opinions about the situation.
Vocational schools that focus on same industries as covered by the questionnaire were chosen in order to have greater explanatory potential.