Impact of External Job Mobility and Occupational Job Mobility on Earnings

Purpose: The purpose of the study is to examine the relationship between external job mobility and occupational job mobility on earnings among engineers in Malaysia. Design/methodology/approach: Using curricular vitae data from a job agency, this paper tracks job mobility through job histories and examine how it affects earnings. Findings: Results obtained from regression analysis indicate that higher external job mobility will contribute to higher earnings, but occupational mobility will have adverse effect on earnings. Research limitations/implications: Limitation of the study is that the results are extrapolated from a self-report dataset. Practical implications: Nonetheless, the results give important implications to the Malaysian job market on how firm-specific skills and occupational specific skills are rewarded among engineers who actively seek for alternative employment online, and a guide to job applicants in career planning. Originality/value: The findings has also revealed important variables to be included in explaining high skill labor earnings in the context of Malaysian engineers, it serves as an important reference for future in modeling earnings.


Introduction
The role of job mobility in human capital development literature yields mixed results. Job mobility has received considerable attention in studies related to salary attainment on individual career progression (Ng, Sorensen, Eby & Feldman, 2007). Some studies have suggested that individuals who have undergone inter-firm mobility experienced higher career and income progressions as compared to their counterparts who did not (Lam, Ng & Feldman, 2012;Brett & Stroh, 1997;Ghosh, 2007;Lam & Dreher, 2004). Talent engineers in Malaysia remained an issue to be explored. Therefore, this leads the study to hypothesize that: Hypothesis 2: Occupational mobility is significantly related to earnings.
Generally, external job mobility indicates reduced firm specific human capital, and occupational mobility indicates reduced occupational specific human capital. Hence, Hypothesis 3 aims to test the different effects of external and occupational mobility on earning Hypothesis 3: External job mobility and occupational mobility affect earnings differently.

Social Demographic Factors and Earnings
Other than work related attributes, factors that are not directly related to productivity sometimes do affect an individual's earnings. In particular, social demographic factors like age, gender, education and ethnic group are closely related to earnings (Vardi, 1980;Carless & Arnup, 2011;Mincer, 1974).
As mentioned in the previous section, firms are less likely to invest in older workers compared to younger workers since their remaining time of contribution to the firm reduces with age. In addition, employees face reduced financial and non-financial incentives and increasing mobility costs as one aged. Therefore, older workers are likely to be less mobile in the job market (Groot & Verberne, 1997). However, it does not mean that older workers have less opportunity to have higher earnings compared to their younger counterparts, as working experience remain essential in determining earnings. It should be noted that it is vital to distinguish between age and working experience as the two factors do not always draw the same implications in different occupations. An older worker does not necessary have sufficient experience in a particular field if occupational mobility is high throughout the career path. Therefore, a firm's decision in recruiting and investing in experienced human capital is based on working experience in a specific field rather than a worker's age. Senior workers who are lacking of firm or occupational specific skills and knowledge are less in demand (Klevmarken & Quigley, 1976).
Gender plays a vital role in determining success and earnings in the engineering profession as it is often perceived that engineering is a masculine profession, thus, the jobs are deemed more suited for males (Hatmaker, 2013;Powell, Dainty & Bagilhole, 2012;González-Ramos & Bosch, 2013). Aside from underrepresentation of female in engineering profession, females are often found or perceived to have weaker performance and higher drop-out rate in engineering programmes in higher education institutions (Arastoopour, Chesler & Shaffer, 2014;Felder, Felder, Mauney, Hamrin & Dietz, 1995). With these backgrounds, the paper will examine whether gender play an important role in determining Malaysian engineers' earnings.

Sample
The primary criterion of sample selection is the occupation, where only engineers will be selected for the study. More specifically, the selection is confined to Malaysian engineers who are actively seeking for alternative employment through online job portals. Education is a controlled variable where selected samples must attain at least a tertiary education.
All data were updated as of year 2014. Several online recruitment agencies were contacted but they were unable to participate, including Jobstreet.com., which is one of the pioneer and largest online recruitment agencies in Malaysia. JobsCentral Malaysia is the only company which is able to participate and provide the data needed for this study. In terms of registered jobseeker, JobsCentral Malaysia has more than Hence, all personal information remained confidential and they are not accessible by researchers.

Procedures
Only updated information required for the study are collected. Job title and duration for each were recorded to examine working experience and job mobility. Other demographic details were also recorded for further analysis.

Analytical Model
The paper attempts to analyse the relationship between job mobility and earnings, in addition to the other variables established in the Mincer's model. Hierarchical multiple regression analysis is used to examine the relationship between the study variables. The analytical model is shown below, variables are chosen based on human capital theory and literature review: Age and working experience are measured in years, external job mobility and occupational mobility are measured in the number of job changes, and finally gender is measured as a dummy variable where female = 0 and male = 1.

Definition of Terms and Data Management
Job mobility is defined as external job mobility which involves changing organisation (Lam et al., 2012;Ng et al., 2007). It is measured by the number of job changes throughout an individual's career. Due to limitation of data, it will not specify whether the change is on voluntary basis.
Defining an individual's occupation as an engineer is based on two criteria, i.e. the education background should be engineering at a bachelor degree and/or above, and have at least one job experience in their career history worked as an engineer. The term engineer is defined according to the Malaysia Standard

Classification of Occupations (MASCO) 2008.
Occupational mobility is defined by jobs that are not classified within the mentioned terminology. For example, if person A worked as an engineer in his first job but switched to be a banker in his second job, it will be counted as one occupational change.
The data collected contain certain difficulties such as ambiguous job title which makes it difficult to define certain professions. This research managed such constraints as below: a) Research assistants, trainees, laboratory workers and interns are not classified as occupational mobility from engineering professions.
b) Generic titles such as "manager" or "supervisor" are classified as occupational mobility from previous profession. Working experience is defined as the years of active working experience after tertiary education, thus it excludes years of working experience before an individual is qualified as an engineering. There are instances where a relatively aged individual may have relatively less years of working experience (e.g. aged 43 but had only 3 years working experience) compared to their counterparts. Although this might be explained by various reasons, such as taking a break from work, nevertheless the specific reasons of less working experience relative to age is not the focus of this study.
Earning is defined by the latest pre-taxed monthly salary paid to the samples, measured in Malaysian Ringgit (MYR).

Results
This section provides an overview of the sample characteristics and mobility trend. The second part reveals the results of the regression model and testing of the hypotheses mentioned above.

Descriptive Statistics
In

Relationships between Job Mobility and Earnings of Malaysian Engineers
Prior to running the regression analysis, it is suspected that multicollinearity could be an issue since age and working experience are highly correlated in nature. Correlation analysis and collinearity diagnosis were conducted and it is found that the Variance Inflation Factor (VIF) is less than 10, within the acceptable level for regression analysis, therefore both age and working experience will be included in the regression model (see Appendix).
To test the hypotheses mentioned in section 2, hierarchical multiple regression is used to examine the impact of job mobility on earnings as shown in Table 3. Model 1 examines effects of control variables Lam, Ng and Feldman (2012), higher external job mobility may not necessary contribute to higher earnings, unless it is accompany by working experience. In other words, job-hopping without accumulation of working experience may not be beneficial. In order to test this hypothesis, an interaction term consists of working experience and external job mobility is included Model 3 for analysis. The results indicate similar results as Model 2, and the interaction term is insignificant.

Discussion
Modifying from Mincer (1984) human capital model, but controlled for the education factor, this study examines the impact of age, working experience, external job mobility and occupational mobility on earnings of engineers in Malaysia. The findings indicate that external job mobility is positively related to earnings (Lam et al., 2012;Murrell et al., 1996;Topel & Ward, 1992) while occupational mobility is negatively related to earnings (Gius, 2014;Neal, 1995). This implies that, generally employers in Malaysia are willing to give more rewards to engineers who have working experience in more than one organisation. This could be due to engineering skills are either transferrable or could be assimilated into new organisations easily. Thus, firm specific skills may not be the determining factor in earnings for engineers who are actively seeking for alternative employment in this study. Nevertheless, occupational mobility is negatively related to earnings, it indicates that occupational specific skills are vital in deciding an engineer's earnings. This is an important finding as it emphasises the importance for engineers to stay in the engineering profession in order to earn a higher salary.
Working experience remains the primary factor contributing to earnings in this study, followed by external job mobility. One of the interesting findings in Table 3 is that age appears to be insignificant factor in the equation. It shows firms prefer to compensate those with working experience over age although the two factors could be interrelated. This concurs with Mincer (1984) that firms would invest in younger workers compared to older workers given they have similar working experience. It further confirmed the need to separate age and working experience as different factors contributing to earnings (Klevmarken & Quigley, 1976).
Insignificant results of interaction term (working experience x external job mobility) in Model 3 suggested that external job mobility may not need to rely on working experience to have an effect on earnings. It implies that compensation based on job mobility happens in all career stages, regardless of the accumulated years of experience.

Conclusion and Limitation
External job mobility is found to be an important factor contributing to earnings in Malaysia's engineering job market. Nevertheless, working experience remains the primary factor contributing to earnings. Graduates from engineering schools are advised not to move out from the engineering profession as it could potentially give negative signals to employers and in turn affects earnings.
The findings indicate that external job mobility led to higher earnings. Engineers who wish to earn more should be more mobile in the job market to learn different skills from different organisation and to enrich their profiles. For employers, it is crucial for them to develop strategic human resource management to handle more mobile talents.
A major limitation of this study is data limitation. As job histories are self-report in nature, there is limitation on the authenticity of the record. It is possible that the information indicated in the CVs is used to maximise the chances of employment opportunities and histories that could induce negative perceptions by employers are selectively removed by job applicants. Nevertheless, the CVs require applicants to name the organisations in the job histories and referees contact details are required by the job search website for employer's reference. Although it is still possible for the applicants to remove certain negatively perceived short-term employment histories, the likelihood for the job applicants to forge their job histories is relatively low because it is relatively easy for the potential employer to check its authenticity.
Another limitation is that this study has not taken ethnic group into account as Malaysia is known to have ethnic oriented social economic differences in her historical background. This is because there are little job applicants indicated their ethnic group in the data collected, more ethnic based compensatory analysis is needed in future Malaysia's job market research. In collinearity diagnostics, we can see that there might be a potential issue with age and working experience. However, it is vital to include both factors as they represent different aspects of job attributes according to Klevmarken and Quigley (1976), therefore the model includes both factors in the model. -896-

A2. Normality of data
The histogram has shown the dependent variable across the sample is normally distributed, and thus suggesting the condition of normality has been met.

A3. Statistical independence of error
The standardised residual versus standardized predicted value plot is used to test the statistical independence of error. The plots are relatively scattered relatively symmetrical around zero, and thus meeting the regression condition.