Angela Ivančič Angela Ivančič INVESTMENT IN EDUCATION AND TRAINING AND SHIFTS BETWEEN JOBS IN THE SLOVENIAN LABOUR MARKET Abstract Job mobility is commonly dealt with as a function of individual resources and characteristics as well as structural factors . Labour market structures determine mobility opportunities, whereas individ- ual resources and characteristics enable individuals to make use of these opportunities . Educational attainment is regarded as an indi- vidual resource which is likely to open firms' doors, while its impact on internal job mobility is less clear. Some people continue to invest in their human capital later on in their working lives, be it to improve their level of education or to obtain job/occupation specific qualifica- tions. This article deals with the importance of educational attainment and the effects of completed education during employment, as well as participation in non formal further training on job mobility processes in the Slovenian labour market . The overall job transition rate, intra- firm job transition rate and inter-firm job transition rate are taken into consideration. Retrospective data on job histories covering the period 1974-1994 are analysed, which enables us to point out differences caused by changes in labour market structures in the period of tran- sition to a market economy . The findings show that people with higher levels of educational attainment are more likely to change jobs, but also that labour force experience and non formal further training seem to make up for miss- ing educational certificates in overall job shifts and inter-firm job shift, whereas original occupation has the same function in intra-firm job shifts processes. Additionally the function of education completed during ones working life as job mobility resource is much less clear than expected . Neither labour force experience nor further training function as a retention mechanisms i .e ., they do not chain people to existing jobs ; organised non formal training supports job shifts. The transition peri- od, significantly increases intra firm job shifts, however it does not seem to alter the significance of human capital variables and original occupations on job transition rates . Key words: job shifts, job mobility, career development, labour mar- ket outcomes, mobility resources 186 DR, VoI.XII(1996)22-23 INVESTMENT IN EDUCATION AND TRAINING AND SHIFTS BETWEEN JOBS IN THE SLOVENIAN LABOUR MARKET INTRODUCTION A great deal of effort has been invested in empirically estimating the rela- tionship between education and labour market outcomes . Yet researchers have rarely directly examined the matching process by which individuals reach their positions . The perspective of "new structuralism" has recently emerged in the social stratification literature as a means of investigating the interplay between labour market structures and the process of stratifica- tion . The major contribution has been the focus on the demand side of the labour market, which represents the opportunity structure of mobility processes. Most contemporary job mobility theories view mobility as a func- tion of both labour market structures and individual characteristics . A number of studies (Soerensen and Tuma 1981 ; Tuma 1985; Mayer and Caroll 1987 ; Blossfeld 1986; Mach, Mayer and Pohoski 1994) have employed a dynamic approach to analysing employment histories in order to investigate the mechanisms that govern labour market outcomes . This article tries to determine the impact of investment in human capital on job moves in the Slovenian labour market under labour market structures characteristic of the socialist system and under changes in these structures initiated by the transition to a market economy . Investment in educational attainment is taken into account, as is investment in specific occupational qualifications in the form of work experience and non-formal further train- ing courses . THEORIES ON MATCHING PEOPLE TO JOBS Three main theoretical models on job allocation have been developed con- cerning the role of individual resources in the job-matching process . In the 1960s and early 1970s, the status attainment model was introduced in social mobility studies (Blau and Duncan 1967, Sewell and Hauser 1975) . The model is based on the assumption that positions in the labour market are accessible to everyone who has the necessary resources . The model has been criticised for its one-sided emphasis of the relative importance of edu- cation and social origin and other individual characteristics (sex and race), and its neglecting mobility opportunities determined by labour market structures (Soerensen 1977) . Furthermore, neoclassical economists have developed a powerful tool, the theory of human capital explaining labour market attainment (Becker 1975, Mincer 1974) . The human capital theory postulates that changes in labour market attainment reflect differences in productivity, whereby productivity is a result of investment in human capi- tal . Educational attainment and skills thus lead to the improvement of occupational positions . With both considered relatively constant in the process of attainment, on-the-job training and occupational experience are regarded as a major source of changes in labour market outcomes . Accord- ing to Mincer (1974) training is not equally distributed through life . People invest in their human capital as long as they expect the returns on invest- ment to exceed the costs . Therefore training takes place primarily in the first few years of employment . This explains the declining prospects of changing job later in working life . DR, Vol .XII(1996)22-23 187 Angela Ivančič Tuma (1985) argues that, considering the human capital theory, job moves can occur only where the worker and the employer are not in equi- librium: where there is equilibrium job change does not take place since neither the worker nor the employer can improve their position . But the human capital thesis does not say anything about the causes and frequen- cy of disequilibrium. According to Soerensen and Tuma (1981), labour mar- kets differ fundamentally with respect to their mechanisms for change in attainment . A perfect competition in the labour market is only possible if employment is flexible . But, as stated by Tuma (1985), a criticism of the neo-classical theory implies that employment is inflexible . Soerensen (1983) found an important source of inflexibility in the nature of employment rela- tionships. He has developed the concept of "open" and "closed" employment relationships . In "open" employment relationships access to jobs is under the control of employers, while in "closed" relationships it is controlled by the workers .' A job can be occupied by a new incumbent only when the current incumbent retires, decides to leave (for a better job), or dies . The vacancy competition model applies to closed employment relationships . The pivotal mobility mechanism in .this model is the creation of vacant positions and not changes in individual productivity, thus, in general education and occupational skills . A job move can occur only when there is a vacancy in the labour market . Personal resources, such as education, social background and ability are crucial in the vacancy competition model because they determine which candidate gets access to a vacancy . Blossfeld (1986) asserts that the vacancy competition model has intro- duced labour-market structure into the analysis of occupational mobility, but only with the assumption that the extension and shrinking of employ- ment on all hierarchical levels produces a constant rate of vacancies . It is thus a model of promotion within a given inequality structure . Yet the structure of the labour market is not constant over time . He claims that labour market outcomes do also depend on specific historical conditions shaping opportunities at the time of entering the labour market, and on labour market structures at the actual historical time . Looking at the job-matching models described above, the Slovenian labour market under socialism can be designated as a labour market where mobility regimes were governed by closed employment relationships with virtually perfect control by workers over access to jobs . Formally, candi- dates from the external labour market competed for vacancies under the same conditions as those from internal labour markets, and special bodies elected from workers in the firm were authorised to carry out the selection process . Qualification certificates represented the main selection criterion in the allocation process, in which not just educational credentials but also qualification certificates obtained through special assessment procedures in working organisations were taken into account. This suggests that formal recognition of qualifications gained through work experience and various forms of non-formal training might have diminished the role of educational credentials in the job-matching process . 188 DR, Vol .XII(1996)22-23 INVESTMENT IN EDUCATION AND TRAINING AND SHIFTS BETWEEN JOBS IN THE SLOVENIAN LABOUR MARKET With the transition to a market economy the most important resources of workers' control over access to jobs have been abolished, and greater flexi- bility in the functioning of the labour market has been assured . Since one of the goals connected with structural changes in the labour market is the adjustment of the qualifications of the workforce according to these changes, it seems reasonable to expect the value of formal education in the labour market to change and the human capital thesis to start applying in the job-matching processes . DATA, VARIABLES, METHOD The analysis is based on data from the "Quality of Life in Slovenia" sur- vey, a nationally representative data set of 1,807 individuals of 18 years and over. The data covers the period 1974-1994 . Data on the working lives of respondents was collected retrospectively . It provides information on every position held in the labour market by the respondents over the 20 year period .2 Positions held both in and outside the labour market were first recon- structed. Only those who had held at least one position in the labour mar- ket (either employment, self-employment, or unemployment) in the observed period were selected . Taking into account this criterion, a sub- sample of 1,389 respondents was selected . Altogether 3,263 episodes in and outside the labour market were generated, 2,302 of these being employment episodes (jobs) . Each employment episode is described by: - the month and year of the beginning of the episode, - the month and year of the end of the episode, Also regarding employment episodes, information on occupational posi- tion is given along with information on whether a person had come to the position from outside or whether she/he was transferred within the firm . On the basis of this data it was possible to make a distinction between inter-firm and intra-firm job shifts . In order to make it possible to analyse the role of investment in education and training during a working life, educational careers during working lives were also reconstructed . This was done separately for formal education and for participation in non-formal training courses . Reconstruction of marital histories and births of children was done in the same manner . For the purpose of this analysis the following are specified as dependent variables : a) the overall job transition rate ; b) the intra-firm job transition rate; and c) the inter-firm job transition rate .3 As for independent variables, these can be divided into time-constant and time-dependent variables . Time-dependent variables such as age and labour force experience were actualised at the beginning of each episode, whereas variables such as mar- ital state, number of children, education, and participation in non-formal training courses4 during each employment episode were included by episode splitting (see Blossfeld, Hamerle and Mayer 1989 ; Rohwer 1994 ; Blossfeld and Rohwer 1995) . In this way 8,152 sub-episodes were obtained . DR, Vol.XII(1996)22-23 189 Angela Ivančič OCCUPATIONAL CATEGORIES Occupational categories were constructed from data on individual occu- pational positions recorded on the basis of respondents' self-assessment . The job descriptions so obtained were coded with 3-digit codes from the International Standard Classification of Occupations (ILO 1990) . These 3- digit codes where then aggregated into broader occupational groups accord- ing to qualification hierarchy and occupational activities . An attempt was made to create occupational groups as homogeneous as possible with regard to levels of qualifications and occupational activities .5 The following broader occupational groups were obtained : a) managerial positions (top and medium executive positions in finance and politics; administrative (managerial) positions ; managerial positions in consumer and social services) ; b) professional occupations ; c) semi-professional occupations - technical and natural fields ; d) semi-professional occupations - other ("service") fields ; e) administrative, clerical and commercial occupations ; I) skilled service occupations ; k) skilled manual occupations ; 1) unskilled and semi-skilled manual and service occupations . LIST OF INDEPENDENT VARIABLES INCLUDED IN THE ANALYSIS : a) Educational attainment : - 1 - 3 year occupational training at secondary school level, dummy vari- able ; - 4- and 5-year secondary education, dummy variable ; - college diploma, dummy variable ; - university/post-university diploma, dummy variable ; (the reference category is represented by the compulsory schooling) . b) Individual characteristics : - Age in months ; - Male, dummy variable (female is the reference category) ; - Married, dummy variable ; (non-married - single, divorced, widowed - is the reference category) ; - Number of children younger than 16 years . c) Education completed during employment :6 - The period after the first educational episode at post-secondary level, dummy variable ; - The period after the first educational episode at secondary level, dummy variable ; - The period after the second educational episode at post-secondary level, dummy variable ; - The period after the second educational episode at secondary level, dummy variable ; 190 DR, Vol.XII(1996)22-23 INVESTMENT IN EDUCATION AND TRAINING AND SHIFTS BETWEEN JOBS IN THE SLOVENIAN LABOUR MARKET (the period prior to each completed educational episode is the reference category) . d) Special occupational resources : Occupational attainment : Managerial position, dummy variable ; Professional occupations, dummy variable ; Semi-professional occupations - technical and natural fields, dummy variable ; Semi-professional occupations - other fields, dummy variable ; Administrative, clerical and commercial occupations, dummy variable ; Skilled service occupations, dummy variable ; Skilled manual occupations, dummy variable ; (unskilled and semi-skilled manual and service occupations are the refer- ence category), . Occupational prestige measured by Treiman's International Occupation- al Prestige Scale ; Acquisition of special occupational skills : General labour force experience in months? ; Participation in non-formal training courses :8 The period after the 1st job-related training course, dummy variable ; The period after the 1st non-job-related training course, dummy vari- able ; The period after the 2nd job-related training course, dummy variable ; The period after the 2nd non-job-related training course, dummy vari- able ; The period after the 3rd job-related training course, dummy variable ; The period after the 3rd non-job-related training course, dummy vari- able ; The period after the 4th job-related training course, dummy variable ; The period after the 4th non-job-related training course, dummy vari- able ; (the period prior to the end of each training course is the reference cate- gory) . e) Inter-firm job shift, dummy variable ; (intra-firm job shift is the reference category) . f) Cohorts and Periods - Cohort 1946-55, dummy variable ; - Cohort 1956-65, dummy variable ; - Cohort >1965, dummy variable ; (Cohort < 1946 is the reference category) . - The period from 1988 onward9 dummy variable ; (the period prior to 1988 is the reference category) . For the analysis of effects of independent variables on transition rates, Event History Analysis has been used (Hamerle, Mayer 1989 ; Blossfeld, Rohwer 1995) . The analysis was performed by the TDA computer package (Rohwer 1994) . DR, VoI.XII(1996)22-23 191 Angela Ivančič The duration of an episode is represented in the statistical model by a non-negative random variable T. The density and distribution function are denoted by f(t) and F(t) respectively. The distribution function can be speci- fied as F(t) = Pr(T_t) The hazard rate, r(t) is the central variable in the model ; it can be inter- preted as the unobserved transition rate from a job r(t)=f(t)S(t) or a more precise definition r(t) = lim P(t 5 T < t+AtIT >_ t) At -* OA t At As such the hazard rate is defined as the limit as A t approaches zero of the probability of an event (i .e . a job shift) occurring in the interval of time between t and t + A t, conditional on survival to time t and relative to that interval . Since transition rates to specified destination states have been modelled, we used competing risk models . A transition specific ( i .e. the distinction between intra-firm and inter-firm job shifts) hazard rate my be defined as follows rj(t I x) = lim P(t<_ T_ t,x) At -~OAt At where J denotes a state variable which may take on values out of the set of destination states A state specific rate is therefore the instanta- neous probability of an event of type j occurring at time t, conditional on survival to time t . RESULTS The analysis starts with a description of the duration of a given job . For this purpose the product limit estimation (Kaplan-Meier method) of survival function has been applied (Rohwer 1994; Blossfeld, Rohwer 1995) . According to the descriptive statistics, the average length of time a job is held is about 9 years . In graph 1, the survival function in a given job is plotted10 ; the curve shows that the survival function decreases faster in the first years of holding a job, thereafter it leaves it levels off . The decrease is faster again after some 25 years, and then it approaches zero . The pre- sented dynamics are in accordance with general pattern characteristic else- where. Termination of a job at the beginning of its duration is commonly 192 DR, Vol .XII(1996)22-23 INVESTMENT IN EDUCATION AND TRAINING AND SHIFTS BETWEEN JOBS IN THE SLOVENIAN LABOUR MARKET Graph 1 : Duration of jobs t c o - Cw it 0 60 120 180 240 300 1 360 420 480 Time in months explained using the human capital thesis (Mincer 1974; Becker 1975) and the dual labour market theory (Doeringer and Piore 1971) . At the beginning of employment in a job the investment in job specific skills is low and there- fore the costs of leaving the job are lower for both, the employee and the employer. With a longer stay in a job, specific occupational skills are acquired which discourage moving to another job, occupation or employer . There may also be other benefits after spending some time in a job which discourage job shifts, such as fringe benefits connected with seniority, pro- motion prospects and the like . In the context of surviving in a job in the Slovenian labour market, the probation period of newly hired employees should be mentioned . Within this period the employee is free to leave a job at any time without prior notice, and the employer has the right to dismiss a worker if during this period it turns out that he/she does not meet the requirements of the firm . Young people who enter employment for the first time after completing education have to pass a period of internship . Only after successfully completing internship they are allocated to a job which is commensurate with their qualifications . The assignment may be with the same employer or with another employer. Regarding the tendency for greater flexibility of the labour force in the period of transition to a market economy, it is likely that a probationary period, temporary internships and temporary employment are more often used in order to make it easier for numerical adjustment of the work force . DR, Vol.XII(1996)22-23 193 Angela lvančič ESTIMATED BASELINE TRANSITION RATES Maximum likelihood models have been used in order to estimate transi- tion rates . Piecewise constant models were applied to account for duration dependence (Blossfeld, Rohwer 1995) . Table 1 : ESTIMATED BASELINE TRANSITION RATES Duration of overall intra-firm inter-firm holding a job job shifts job shifts job shifts 0 - 1 year -5.5146" -6.5624 * -5 .3647* * 1 - 5 years -5.3628" -6.3290" -5.5488* 5 - 8 years -5 .6869* * -6.5117** -5.7445" 8 - 10 years -5 .8227* ` -6 .5224* * -6.3148** 10 - 15 years -5 .7867`" -6 .3261 * * -6 .3438* * 15 years and over -6.3174** -6.6490** -7.0659" The duration of jobs was divided into six time periods : 0 - 1 year, 1 - 5 years, 5 - 8 years, 8- 10 years, 10 - 15 years, 15 years and over. The esti- mated baseline transition rates indicate that in the duration of time spent in a job there are not very notable differences ; this holds true for the overall rate of transition to another job and the rate of intra-firm job shifts . The risk for the overall job shifts is slightly higher in the period of 1 to 5 years of holding the job and somewhat lower in the period from 15 years onward, while the highest risk of being transferred within the firm is characteristic of a periods from 1 to 5 years and from 10 to 15 years of holding a job . The likelihood of moving to another job within the firm is thus also somewhat higher in the later phase of holding a job. These later internal job shifts might indicate the promotion of a worker as well as protection against los- ing employment or worsening working conditions. On the contrary, the baseline inter-firm job transition rate indicates that a higher likelihood of moving to a job in another firm is characteristic of the first year of holding a job, while in later periods it decreases continuously . Thus the longer a per- son stays in the same job the less likely he/she is to move to another employer. EFFECTS OF EDUCATIONAL ATTAINMENT ON JOB TRANSITION RATES The attained level of education is considered the most important individ- ual job mobility resource, regardless of whether it is dealt with from the viewpoint of the human capital theory, screening thesis, signalling thesis or credentialists' theory . The evidence is consistent in that better educational attainment enables individuals to get better positions in the labour market . Most authors stress the importance of education in entering a firm (Bills 1988; Soerensen and Tuma 1981), but there is also some evidence which confirms that education is also relevant in the case of intra-firm mobility, albeit with different effects (Di Prete and Krecker 1991) . Our results con- firm that the level of education attained significantly and positively affects 194 DR, Vol.XII(1996)22-23 INVESTMENT IN EDUCATION AND TRAINING AND SHIFTS BETWEEN JOBS IN THE SLOVENIAN LABOUR MARKET the job-to-job transition rate : the higher the attained level of education the stronger the effect on the transition rate compared to compulsory educa- tion. This holds equally true for the overall job transition rate (model 1 in table 2), for the internal job transition rate (model 1 in table 3 .1), and for the inter-firm job transition rate (model 1 in table 3 .2) .11 Such results cor- respond with the principle of the human capital theory that more invest- ment in education raises the marketability of the worker . Taking into account the functioning of the labour market under socialism, it seems appropriate to look for an additional explanation of the obtained results also at the opportunity structures . According to some findings regarding labour market outcomes in countries with different educational systems, educational opportunities and specific structures of educational systems are as important in terms of labour market mobility regimes as are an indi- vidual's attributes. Allmendinger (1989) and Koenig and Mueller (1986) claim that in countries where educational systems are closely linked to the occupational structure and provide highly standardised and certified edu- cational outcomes, people without the required training certificates are lim- ited in their access to jobs in the relevant occupations. Given that in the Slovenian labour market occupational positions at each qualification level strongly correlate with particular level and type of education, the number of vacant positions with particular qualification demands creates mobility opportunities for individual educational categories . Official statistical data for the observed period indicate that the proportion of those employed in the traditional industrial sector decreased, while the share of those employed in the service sector increased . Also, the greatest educational dis- crepancies are found at the lowest and highest qualification levels (Mohorčič Špolar, Ivančič 1996) . At the lowest qualification level there is over-employment, while there is a shortage of the employed with higher level qualifications . As a result, those with higher levels of education have more opportunities for moving to another job . However, as stated by Tuma (1985), according to the human capital theory job shifts occur only if and when the employee and the employer are in disequilibrium . Considering the wage system with small wage differentials that was characteristic of the socialist period, the higher probability of job shifts by better educated workers can in addition be explained by their feeling under-rewarded, which makes them search for jobs with higher rewards . The fact that the attained level of education is important for inter- as well as intra-firm job shifts (model 1 in tables 3 .1, 3.2)12 can be linked to the absence of internal labour markets with job ladders that would enable the internal promotion of workers on the basis of special occupational skills and seniority . EFFECTS OF AGE, GENDER AND FAMILY SITUATION ON TRANSITION RATES Individual characteristics such as age, sex, marital status and number of dependent children are treated as "individual constraints" to job mobility processes (Soerensen 1975) ; they affect the transition rate regardless of individual resources . Results presented in model 2 in table 2 are in favour DR, Vol.XII(1996)22-23 195 Angela lvančič of such an assumption . The effect of age on the overall job transition rate is significant and negative, which confirms the assumption that the capacity to change job decreases with age . This is in accordance with the job-match- ing thesis (Granovetter 1986, Jovanovic 1979) and with the human capital thesis . Workers later on in life are less ready to move to other jobs 13, and employers also refuse to hire older workers since they consider them less trainable and their working capacity reduced . Accordingly they do not expect returns on investment to justify the hiring costs . Such an assump- tion is additionally confirmed by the finding that age does not significantly affect the rate of intra-firm job shifts, whereas it seems to be an obstacle in the case of inter-firm job shifts (model 2 in tables 3 .1 and 3.2) . Furthermore, results indicating the effect of gender on overall job transi- tion rate are consistent with those in other countries (Caroll and Mayer 1986, DiPrete and Soule 1988) : being a male significantly and positively affects the probability of changing job (model 2 in table 2) . The lower female job-to-job transition rate can be explained by the traditional view of labour market participation of women, according to which the women's career is primarily within the family . Accordingly, women's employment is still often treated as supplementary even though we are talking about full-time employment. Socialist regimes were not able either to guarantee that legally established equal rights for men and women would necessarily be applied in everyday life . Additionally, women themselves often do not aspire to career development . The gender-specific division of family roles forces them to sacrifice their occupational careers to the well-being of the family . As a result women often prefer to have stable jobs without additional responsi- bilities . Benefits, such as the organisation of working time to suit family obligations, can compensate for more prestigious labour market positions . It is expected that differences in investment in human capital cause differ- ences in women's labour market careers, but this is not a question addressed by this analysis . It is typical for sex to affect both the intra-firm and the inter-firm job transition rates significantly (model 2 in table 3), the effect being somewhat stronger in the case of intra-firm job shifts . Married people also display higher job stability than unmarried people . Such a result can be connected with the additional responsibilities that people take on when they get married . In addition, marriage more often takes place in the already stabilised life period when one settles in a job . Unlike marriage, the number of children younger than 16 years does not significantly affect the likelihood of moving to another job, either intra-firm or inter-firm . Controlling for personal characteristics does not alter the significance of effects of educational attainment on job shifts, again this holds true for overall job transition rate as well as intra-firm and inter-firm job transition rate . 196 DR, Vol.XII(1996)22-23 INVESTMENT IN EDUCATION AND TRAINING AND SHIFTS BETWEEN JOBS IN THE SLOVENIAN LABOUR MARKET EFFECTS OF COMPLETING EDUCATION DURING WORKING LIFE ON RATES OF JOB SHIFTS Not all people finish their educational careers at the time of their entering employment. Some return to education later on in their working lives, be it to make up for missing opportunities during their youth or to advance their educational level . Formal education is explicitly dealt with as an investment in human capital which is easily transferable between firms and industries, it is thus presumed that the acquisition of higher levels of education improves labour market attainment . Since educational credentials formally play an important role in the Slovenian labour market, transitions to jobs at higher qualification levels are expected to occur via educational channels . Consequently it is assumed that educational achievements during employment are followed by job shifts. However, empirical evidence points to a somewhat different conclu- sion . The effects of two finished educational episodes at secondary and post-secondary level have been estimated in models 3 in tables 2, 3 .1 and 3 .2 . According to the results, only the period after the second episode at secondary level significantly and positively affects the job-to-job transition rate 14 , while other completed episodes do not show significant effects . Given these results it appears that investment in formal education later on in one's working life is not as significant an individual's job mobility resource as it was expected to be . This can be clarified by job-matching practices developed in the past . Although access to jobs was formally limit- ed by educational credentials, both the shortage of suitable qualifications and the existing wage system supported the allocation of jobs to workers without the required educational certificates . Some of these workers later acquired the required credentials . Additionally, the educational require- ments of particular jobs change over time and their incumbents may be required to improve their attained educational level . It is thus more likely that workers go back to school in order to fulfil the formal requirements of their job than to advance their occupational career . Regarding our findings, this seems especially true for those who finish post-secondary education .15 Educational upgrading later on in one's working life, both at the secondary and at the post-secondary level, thus seems to have a compensatory rather than a cumulative effect in job mobility processes ; that is, people first occu- py jobs with higher educational requirements which they only belatedly sat- isfy. Are there any differences between the effects of completed formal educa- tional episodes while in employment on transition rates in the case of inter- firm job shifts and those in intra-firm job shifts? From models 3 in tables 3 .1 and 3.2 it is evident that the periods after educational episodes do not have significant effects on the rates of intra-firm job shifts, whereas the second educational episode at secondary level significantly increases the likelihood of inter-firm job shifts . DR, Vol.XII(1996)22-23 197 Angela Ivančič Based on these findings the conclusion can be drawn that advancing the level of formal education later on in one's working life most likely represents an investment only in so far as it helps an employee keep his/her occupa- tional position, since it less often facilitates job shifts . EFFECTS OF OCCUPATIONAL RESOURCES ON TRANSITION RATES Occupational resources in this analysis are as follows : original occupa- tional position, occupational prestige, labour force experience, and qualifi- cations acquired in non-formal further education and training . Attained occupational position and occupational prestige represent one's occupa- tional achievement, whereas labour force experience and non-formal fur- ther training are dealt with as special occupational qualifications (skills) . OCCUPATIONAL ACHIEVEMENT The results suggest that only two original occupational groups signifi- cantly affect the overall job transition rate : administrative and clerical occu- pations, and skilled service occupations ; in both of these occupational groups there is an increased likelihood of transition to another job (model 4 in table 2) . Because of the considerable decrease in the proportion of unskilled and semi-skilled workers in the last 20 years, the significant effect of this occupational group on transition rates was expected, yet the results do not correspond with this expectation . It may be assumed that discrepancies in occupational skills prevent workers from moving to other occupations, instead they leave and either remain unemployed or move to states outside the labour market (early retirement), while young people with more advanced qualifications occupy new jobs . Furthermore, it is also pos- sible that the strong correlation between individual occupations and par- ticular educational levels cancels out occupational effects on transition rates. The significant effects of administrative, clerical and commercial occupations and skilled service occupations on the job shift rate may indi- cate that these occupations are educationally less homogeneous and are not bound to particular industry, therefore they afford more opportunities for crossing occupational and educational boundaries . However, more detailed scrutiny would be needed for a satisfactory explanation . Results presented in model 4 in tables 3 .1 and 3.2 show that none of the original occupational groups significantly affects the intra-firm job transi- tion rate, whereas in skilled service occupations the likelihood of inter-firm job shifts is substantially increased. In other words, occupational groups do not differ in the likelihood of moving to another job within firms, while only those employed in skilled service occupations are more likely to move to other firms than unskilled and semi-skilled manual and servant occupa- tions . After controlling for original occupational positions the significance of effects of educational attainment on the intra-firm transition rate alters : only university education remains significant . This could suggest that occu- pational attainment offsets the effect of some levels of education in internal 198 DR, Vol.XII(1996)22-23 INVESTMENT IN EDUCATION AND TRAINING AND SHIFTS BETWEEN JOBS IN THE SLOVENIAN LABOUR MARKET mobility processes, while in inter-firm mobility processes educational attainment is more significant than attained occupational position . On the other hand the finding that occupational prestige has a significant and negative effect on the inter-firm job transition rate and is non-significant in intra-firm mobility leads to the conclusion that those who occupy less prestigious jobs are more likely to move between firms, whereas in the case of intra-firm mobility occupational prestige is not relevant (models 5 in tables 3.1 and 3.2) . LABOUR FORCE EXPERIENCE AND PARTICIPATION IN NON- FORMAL FURTHER TRAINING In the human capital theory (Thurow 1975, Becker 1964) and dual labour market theory (Doeringer and Piore 1971) investment in training is positively associated with job security and the internal promotion of work- ers . The extent of labour force experience is usually a measure of the acquisition of job-specific human capital . The longer workers stay in a job the more job and firm-specific human capital they acquire, and this chains them to a particular job and to a particular firm . Accordingly, a negative effect of labour force experience on the inter-firm job shift rate is expected, while it is assumed that labour force experience facilitates internal job shifts. Yet as Wholey (1990) points out, the length of time an employee spends in a job could indicate a lot of things, for instance that a person has reached his/her maximum level of competence, which indicates that he/she has lost the necessary human resources to start in a new job . Our results point to the non-significance of labour force experience on job tran- sition rates, this applying equally to job shifts in general, intra-firm job shifts and to inter-firm job shifts . However, occupational skills are not only acquired through on-the-job training; people often attend organised training courses in the external edu- cational market. It is usually claimed that firms are more willing to invest in job-related training, while individuals themselves usually invest in more general qualifications which are valued by the external labour market . For the purpose of this analysis non-formal further training is divided in job-related training and non-jobrelated training . It is assumed that job- related training constrains external job mobility, while training which is not related to a job does not affect job shifts . The estimated effects (model 6 in table 2) suggest that the first training has a significant effect on the overall job shift rate . Given that people most often participate in educational and training events at a relatively young age (Ivančič 1995) and that this is when job shifts most often occur, it is understandable that when searching for another more suitable job addi- tionally acquired qualifications are utilised . Also, training is often acquired outside firms in the education market ; it is thus not tailor-made and skills obtained are easily transferable between firms and industries . Moreover, it is appearent that non-formal further training does not negatively affect job shifts, rather it has either significant and positive effect or non-significant effect. Further inspection of the estimated effects of non-formal further DR, Vol.XII(1996)22-23 199 Angela Ivančič training on intra-firm and inter-firm job transition rates (model 6 in tables 3.1 and 3.2) also indicates that the intra-firm job transition rate is not sig- nificantly affected by the first participation in training, it is further partici- pation that actually matters, and both job-related courses and non-job- related courses are significant . On the other hand in the case of inter-firm mobility the participation in non-job-related courses seems more important than that in job-related courses . The results obtained thus support the assumption that the acquisition of more general skills increases the opportunities for external job mobility, whereas job-orientated training is more likely to increase the likelihood of internal job mobility . At the same time these findings reveal that further training does not function as having a "holding" power in mobility process- es; it does not negatively influence the risk of moving to another job . It was assumed that labour force experience and further training repre- sent an additional mobility resource complementary to the attained level of education. Yet, looking at model 6 in tables 2, 3 .1 and 3.2 it is evident that after entering these variables in the mobility model the significance of indi- vidual educational levels on transition rates changes . In the case of the overall job transition rate and inter-firm job transition rate only college education remains significant, while the significance of educational levels for intra-firm job shift rate disappears . This suggests that qualifications acquired through work experience and non-formal further training balance, in the job-matching process, formal qualifications obtained in the educa- tional system, and this decreases the significance of educational credentials in job mobility processes . Such a conclusion is also supported by findings regarding career development in a working organisation (Čibron 1989) . EFFECTS OF COHORTS AND THE TRANSITION PERIOD ON RATES OF JOB SHIFTS The period covered by this study, 1974 - 1994, is characterised by deep economic, social and political changes . In the second half of the 1980s processes of restructuring of the labour market started which have funda- mentally affected the behaviour of workers in the labour market . We have attempted to capture the effects of these changes on mobility processes in the labour market by estimating cohort effects and period effects . A cohort design is used that consists of 10-year birth cohorts from 1936 to 1965, whereas those born prior to 1936 are aggregated into the same birth cohort, and likewise those born after 1965 . Since the birth cohort prior to 1936 and the 1935 - 45 birth cohort showed very similar survival functions they have been aggregated into one cohort. As for period effects, the period following 1987 has been included in the analysis . Starting with the cohort effects, the results obtained indicate that com- pared to the cohort born prior to 1946, all three successive birth cohorts significantly, positively and strongly affect the likelihood of transition to another job . The younger the birth cohort the greater the increase in the transition rate . Cohort effects similar to those on the overall job transition rate are also characteristic of intra-firm job shifts and inter-firm job shifts 200 DR, Vol.XII(1996)22-23 INVESTMENT IN EDUCATION AND TRAINING AND SHIFTS BETWEEN JOBS IN THE SLOVENIAN LABOUR MARKET (model 7 in tables 3 .1 and 3.2 ) . How can these differences be explained? The key explanation can most likely be found in the changed labour market conditions which the youngest birth cohort experienced at the time of enter- ing the labour market . Reduction of jobs has greatly restricted the chances for permanent employment . It is likely that temporary employment and internships account for a large proportion of the job shifts of this cohort . On the other hand, each cohort experienced different educational opportu- nities which determined the course of its labour market career. It may be assumed that younger cohorts have adapted to advanced qualifications that much better fit new job requirements than those of older cohorts ; they are therefore expected to have more chances to get access to new jobs created by advanced technology. Unlike cohorts, the transition period does not significantly affect the over- all job- transition rate and inter-firm job shifts, but its effect on the intra- firm job transition rate is statistically significant and positive . 16 As stated by DiPrete (1993) restructuring of enterprises and industries usually pro- duces a cascade of lateral and downward internal shifts in response to the elimination of jobs . Thus, our finding that the transition period significantly increases the chance of internal mobility might indicate the internal "reor- ganisation effect" of structural changes . Furthermore, it is likely that the restructuring of the labour market in the period of transition has in the first place increased the rate of job shifts of younger cohorts, whereas the value of investment in formal education and non-formal further training in the mobility processes does not seem to have changed substantially . DR, Vol.XII(1996)22-23 201 Angela Ivančič Table 2 : EFFECTS ON THE OVERALL JOB TRANSITION RATE, PIECEWISE-CONSTANT MODELS (COMPETING RISKS) M 0 D E L S 1 2 3 4 5 6 7 8 Period 0-12 months -5 .9286**-5.1544** -5 .1385** -5 .1765**-4 .7975**-5.3763** -4 .8959** -7 .2885** Period 12-60 months -5 .7501 * * -4 .9044 * * -4 .9025** -4 .9383**-4 .5532**-5.1570** -4 .6728** -7 .0094** Period 60-96 months -5 .8447**-4.9178** -4 .9276** -4 .9580**-4 .5700**-5.1849** -4 .7200** -6 .9802'* Period 96-120 months -6 .2353**-5.2616** -5 .2787** -5 .3102** -4 .9169**-5.5503** -5 .0912**-7.2672** Period 120 - 180 months -6 .1128**-5.1200** -5 .1475** -5 .1759**-4 .7821**-5 .4234** -4 .9742** -7 .0500** Period >=180 months -6 .6198**-5.6326** -5 .6424** -5 .6680**-5 .2698**-5.9042** -5 .4843** -7 .1892** Level of education Vocational training 0 .2948** 0.2131** 0.1895* 0.1219 0.1705* 0.0669 0.0515 -0.0702 4-,5-year secondary education 0 .4428** 0.4489** 0.4118** 0 .3168** 0 .4059** 0.2365 0.2340 0.0802 College diploma 0.7460** 0.8098** 0.8044** 0 .7576** 0 .9030** 0 .6374** 0.6689** 0.4893 University/post-university diploma 1 .0574** 1 .1977** 1 .2003** 1 .1020** 1 .2772** 0 .7881* 0.7973* 0 .4708 Individual characteristics Age -0 .0026** -0 .0026** -0 .0028** -0.0028** 0 .0004 -0.0005 0.0028 Men 0.1650** 0.1557** 0 .2170** 0 .2192** 0 .2230** 0.2183** 0.1733 Married -0 .4923** -0 .4158** -0 .4175**-0 .4038**-0 .3847** -0 .3438** -0 .1002 No . of children < 16 years 0 .0509 0 .0467 0.0506 0 .0447 0 .0481 0 .0808 -0 .0005 Finished schooling while in employment Period after the 1st educational episode at postsec.level -0.1426 -0.1494 -0 .1646 -0.1889 -0.1498 -0.1252 Period after the 1st educational episode at second. level 0.2008 0.1623 0.1493 0.1579 0.1752 0.0617 Period after the 2nd educational episode at postsec . level 0.0909 0.0986 0.0281 0.0168 -0.0043 -0.1625 Period after the 2nd educational episode at second . level 0 .7442** 0.7485** 0 .7369** 0.7110** 0 .6732** 0.4528** Occupational resources Occupational attainment Managerial position 0.2303 0.4787** 0.5047** 0 .4979** 0.5519** Professional occupations 0.1358 0.4458* 0.4296* 0.4688* 0.6194** Semiprofessional occupations - technical 0.0705 0.2592 0.2640 0.2395 0.2087 Semiprofessional occupations - other 0.0468 0.2155 0.2040 0 .2042 0 .2451 Administr. and clerical occupations 0 .3061** 0 .4842** 0.4809** 0 .4607** 0.4796** Skilled servant occupations 0 .3474** 0 .3482** 0.3795** 0 .3811** 0.3724** Skilled manual occupations 0 .0204 0 .0936 0 .0909 0 .0895 0 .0656 Occupational prestige -0 .0144**-0.0149** -0 .0148** -0 .0130** Labour Juice experience and non-formal training General labour force experience -0 .0033 -0.0031 -0 .0028 Period after the 1 st jobrelated course 0.6088** 0 .5968** 0.5834** Period after the 1st non-jobrelated course 1 .4538** 1 .5525** 1 .9823** Period after the 2nd jobrelated course 0.4597** 0.4117** 0.2728 Period after the 2nd non-jobrelated course 0.5475** 0.4954* 0.3129 Period after the 3rd jobrelated course 0 .2956* 0.2866* 0.3630* * Period after the 3rd non-jobrelated course 0 .0708 0.1511 0.0951 Period after the 4th jobrelated course -0.6067 -0.5446 -0.5746 Period after the 4th non-jobrelated course -0 .0197 0.0269 -0.3870 Inter-firm job shifts -0 .4007** -0 .1830** Married*male 0.0667 0 .1746 Cohort/Period Cohort 1946 - 1955 1 .1643** Cohort 1956 - 1965 1 .4657** Cohort > 1965 1 .9195** Period >1987 -0 .0967 Notes : Number of episodes = 8,152 ; Number of events = 898 ; **p<0 .05 ; *p<010 ; 202 DR, Vol.XII(1996)22-23 INVESTMENT IN EDUCATION AND TRAINING AND SHIFTS BETWEEN JOBS IN THE SLOVENIAN LABOUR MARKET Table 3 .1 : ESTIMATED EFFECTS ON THE INTRA-FIRM JOB TRANSITION RATE, PIECEWISE CONSTANT MODELS (COMPETING RISKS) M 0 D E L S 1 2 3 4 5 6 7 Period 0-12 months -7 .0071**-6.6538** -6 .6310** -6 .5334**-6.2514** 6 .6929** -9 .4427** Period 12-60 months -6 .7883**-6.2887** -6 .2814** -6 .1857**-5 .9016**-6.3689** -8 .8947** Period 60 - 96 months -6 .8999**-6 .3081 ** -6 .3234** -6 .2292** -5 .9435**-6.4180** -8 .6666** Period 96-120 months -6 .9631 **-6 .3118** -6 .3340** -6 .2445**-5.9583**-6 .4473** -8 .5857** Period 120-180 months -6 .6862**-6.0065** -6 .0317** -5 .9403**-5 .6542**-6.1500** -8 .1304- Period >=180 months -6 .9819**-6.3102** -6 .3153** -6 .2152**-5 .9222**-6.3862** -7 .8567- Level of education Vocational training 0 .3588** 0.2909* 0 .2665* 0.2397 0.2769* 0.1856 0 .1033 4-5-year secondary education 0 .4973** 0.4974** 0.4582** 0.2467 0.3150 0.1690 0 .0964 College diploma 0 .7647** 0.7934** 0.7625** 0.5169* 0.6288** 0.4261 0 .5560 University/post-university diploma 1 .0597** 1 .0908** 1 .0608** 0.8672** 0.9884** 0 .7065 0 .7741 Individual characteristics Age -0 .0009 -0 .0009 -0 .0013* -0 .0013* 0 .0010 0 .0019 Men 0.3416** 0.3373** 0.35997**0.3565** 0 .3411** 0 .3890** Married -0 .6992** -0 .6835** -0 .6803**-0 .6671**-0.6329** -0 .1175 No . of children < 16 years -0 .0416 -0 .0492 0 .0477 -0.0505 -0 .0528 -0 .1103 Finished schoolings while in employment Period after the 1st educational period at postsec.level 0.3266 0.2880 0.2856 0.1937 0 .3011 Period after the 1st educational period at second . level 0.4675 0.4120 0.4049 0.4088 0 .2739 Period after the 2nd educational period at postsec. level 0.1344 0.1764 0.1401 0.0161 -0.2107 Period after the 2nd educational period at second . level 0 .4728 0 .4976 0.4869 0 .3467 0 .0035 Occupational resources Occupational attainment Managerial position 0 .4685 0.6501 * 0 .6159* 0 .6024 Professional occupations 0 .1591 0.3898 0 .1964 0 .3406 Semiprofessional occupations - technical 0.4644* 0.5928** 0 .5817** 0 .4370* Semiprofessional occupations - other 0.0713 0.1843 0 .1715 0 .1556 Administrative and clerical occupations 0.3788* 0.4985** 0 .4793** 0 .4074* Skilled servant occupations -0.2190 -0.2235 -0 .1796 -0 .2525 Skilled manual occupations -0 .0701 -0.0197 -0 .0346 -0 .0678 Occupational prestige -0.0103 -0 .0096 -0 .0063 Labour force experience and non-formal training General labour force experience -0 .0024 0 .0017 Period after the 1st jobrelated course -0.0877 0 .0339 Period after the 1st non-jobrelated course -0 .1380 0.6187 Period after the 2nd jobrelated course 0.6298** 0 .4264* Period after the 2nd non-jobrelated course 1 .0675** 0 .5223 Period after the 3rd jobrelated course 0.6276** 0 .7016** Period after the 3rd non-jobrelated course -1 .7018* -1 .6594 Period after the 4th jobrelated course -8.8545 -8.8500 Period after the 4th non-jobrelated course -8 .1684 -8 .6276 Cohort/Period Cohort 1946 - 1955 1 .5117** Cohort 1956 - 1965 2.0772** Cohort > 1965 2.7464** Period >1987 0.7445** Notes : Number of episodes = 5721 ; Number of events = 331 ; **p<0 .05 ; *p<0 .10 DR, Vol.XII(1996)22-23 203 Angela lvančič Table 3 .2: ESTIMATED EFFECTS ON THE RATE OF INTER-FIRM JOB TRANSITIONS, PIECEWISE CONSTANT MODELS (COMPETING RISKS) M 0 D E L S 1 2 3 4 5 6 7 Period 0-12 months -5 .8769**-4.6750** -4 .6617** -4 .7577** -4.4546**-5 .2179** -7.7631** Period 12-60 months -6 .0333**-4.7234** -4 .7247** -4 .8177**-4 .5118**-5 .3064** -7.7469- Period 60-96 months -6 .1957**-4 .7521 ** -4 .7667** -4 .8527** -4.5432**-5 .3504** -7 .6416** Period 96-120 months -6 .7486**-5.2394** -5 .2634** -5 .3444** -5.0324**-5 .8629** -8.0680** Period 120-180 months -6 .7645**-5.2367** -5 .2657** -5 .3434**-5 .0308**-5 .8644** -7.9161 - Period >=180 months -7 .4537**-5.9474**-5.9552**-6 .0319**-5 .7131**-6 .5296** -8 .14 .05** Level of education Vocational training 0 .4889** 0.3662** 0,3458** 0 .2659** 0 .3053** 0.1736 0.0418 4-,5-year secondary education 0 .5179** 0.5255** 0.4949** 0 .4559** 0 .5273** 0.3163 0 .1666 College diploma 0 .8250** 0.9438** 0.9237** 0 .9395** 1 .0471** 0.6817** 0 .5784* University/post-university diploma 1 .1748** 1 .4090** 1 .4101** 1 .3995** 1 .5443** 0.8446 0.5920 Individual characteristics Age -0 .0042** -0 .0042** -0 .0043** -0 .0042** 0.0000 0 .0027 Men 0.2535** 0.2458** 0.2933** 0 .2930** 0.1992** 0 .3338** Married -0 .6171** -0 .5952** -0 .5994**-0 .5879**-0 .5800** -0 .0941 No . of children < 16 years 0 .0483 0 .0508 0 .0559 0 .0518 0.0544 -0 .0072 Finished schooling while in employment Period after the 1st educational episode at postsec .level -0.4717 -0 .4491 -0.4594 -0 .4503 -0,422 Period after the 1st educational episode at second . level -0 .0607 -0 .1053 -0 .1174 -0.1004 -0.1828 Period after the 2nd educational episode at postsec . level 0 .3528 -0.3587 0.3044 0.3558 0 .1429 Period after the 2nd educational episode at second . level 0 .8400** 0.8313** 0 .8132** 0.8299** 0 .5601 ** Occupational resources Occupational attainment Managerial position 0.1103 0.2988 0 .3571 0 .3642 Professional and scientific occup . 0 .0582 0 .2954 0.3832 0 .5303* Semiprofessional occupations-technical -0 .0763 0 .0757 0.1131 0 .0108 Semiprofessional occupations - other 0 .0886 0 .2261 0.2280 0 .2082 Administr. and clerical occupations 0 .1531 0 .2932 0.2900 0 .1928 Skilled servant occupations 0.5182** 0 .5112** 0.5452** 0 .4758** Skilled manual occupations 0.1262 0.1834 0.1875 0 .1558 Occupational prestige -0 .0116**-0.0125** -0 .0098** Labour force experience and non-formal training General labour force experience -0.0044 -0 .0028 Period after the 1 st jobrelated course 0.9303** 0 .8946** Period after the 1st non-jobrelated course 1.8512** 2 .3254** Period after the 2nd jobrelated course 0 .2995 0 .1543 Period after the 2nd non-jobrelated course 0.5889* 0 .2686 Period after the 3rd jobrelated course 0.0102 0 .0504 Period after the 3rd non-jobrelated course 0.5171 * 0 .4647* Period after the 4th jobrelated course 0.4196 0 .5107 Period after the 4th non-jobrelated course 1.5879** 1 .1410 Cohort/Period Cohort 1946 - 1955 1 .2894** Cohort 1956 - 1965 1 .8499** Cohort > 1965 2.3346** Period >1987 0 .2595 Notes : Number of episodes = 5721 ; Number of events = 600; **p<0 .05 ; *p<0 .10 ; 204 DR, Vol .XII(1996)22-23 INVESTMENT IN EDUCATION AND TRAINING AND SHIFTS BETWEEN JOBS IN THE SLOVENIAN LABOUR MARKET CONCLUSION Mobility is commonly dealt with as a function of individual resources and characteristics as well as structural factors . Labour market structures determine mobility opportunities, while individual resources and character- istics enable individuals to make use of these opportunities . This analysis has primarily focused on investigating the effects of human capital vari- ables on overall job shifts, intra-firm job shifts and inter-firm job shifts ; the level of education attained, completed educational episodes during employ- ment, labour force experience, and participation in non-formal further training represent these variables . Since the data covers a period of impor- tant structural changes, the effects of these changes are also estimated . It has been found that survival over time in a given job creates a shape which is consistent with general expectations : somewhat faster decrease in the beginning, which than slows down, and faster decreases again only at the end of a working life . It also seems that only the inter-firm job shift rate shows somewhat stronger duration dependence - a higher initial transition rate which then decreases constantly . Men are more likely than women to move between jobs, while age above all constrains inter-firm job shifts . Our findings regarding the effects of educational attainment seem to be in accordance with the predictions of the human capital theory : higher levels of education are more likely to move to another job . This holds true for overall job shifts, internal shifts and inter-firm shifts . However, the findings that labour force experience and non-formal training cancel out the signifi- cance of effects of formal educational attainment suggest that job/occupa- tion specific qualifications are more likely than general qualifications to facilitate job mobility . Accordingly, occupational qualifications gained out- side the educational system do not represent an occupational resource cumulative to educational credentials, as claimed by the human capital thesis, but more likely function in the job-matching process as a compensa- tion for missing formal education. It can be argued that labour market institutions and arrangements characteristic of the socialist period have to a large extend determined such a position of educational credentials in job mobility processes . The feeling of being under-rewarded encourages better educated workers to search for jobs with rewards more commensurate with their investment in education . Furthermore, our findings indicate that com- pleting formal education during employment functions as a compensatory factor for incumbents of jobs without required educational certificates rather than an individual mobility resource . In closed employment relation- ships a job shift is possible only if a vacancy occurs ; improvement of formal education is thus not automatically followed by a transition to another job, but only if such a job is vacant . As far as investment in formal education later on in one's working life functions as a mobility resource it facilitates inter-firm job transitions . At the same time it holds true that non-formal further training does not bind workers to their jobs ; it supports job mobili- ty, whereby non job related training seems to be more important for inter- firm shifts while both, jobrelated and non-job related training facilitate intra-firm shifts . DR, Vol.XII(1996)22-23 205 Angela Ivančič A strong cohort effect on job shifts was also discovered . Younger cohorts display significantly higher rates of moving between jobs, this applying to all types of shift . As for the transition period it can be argued that in addi- tion to producing a chain of internal job shifts as a consequence of the restructuring of working organisations, its most obvious effect is the increased job mobility of the youngest cohort . The analysis has not discov- ered any significant changes in the role of investment in human resources in job shifts in the period of transition. However, firmer conclusions regard- ing gains and losses of individual educational categories in teh labour mar- ket would only be possible with careful scrutiny of the directions of job shifts and of whether they were voluntary or involuntary . NOTES 1 Workers gain control of access to jobs through a variety of means : specific quali- fications for a job, educational credentials, collective action, etc . A closed employ- ment relationship is typical of primary labour markets, whereas secondary ones are governed by an open employment relationship. Thus human capital theory is more likely to apply to secondary labour markets (Soerensen and Tuma 1981) . 2 A more detailed description of the data collected by the "Quality of Life Survey" is given in Drobnič 1995 . 3 In addition to the overall job transition rate, effects on transition rates to self- employment, to unemployment, to retirement and to (unspecified) states outside the labour market were also estimated in the same models, but are not presented in this article . 4 Only training courses of at least 30 hours were included . 5 Some inconsistencies can be found in the way occupations were classified : a) different managerial levels are aggregated into the same group ; as a consequence the group is not homogeneousregarding qualifications, additionally it also differs in the degree of autonomy and responsibility; b) semi-professional occupations are split into two groups since it is assumed that they are highly gender-biased . The same may also be justified in the case of profes- sionals, yet because of the small number in this category, division was not per- formed ; c) the number of semi-skilled and unskilled service occupations was very small ; these occupations were added to semi-skilled and unskilled manual occupations . In this way qualification homogeneity was preserved, but at the same time the homo- geneity of occupational activities was violated . 6 A maximum of four educational episodes had been completed by one person, yet the 3rd and the 4th schooling showed only few numbers, and were therefore dropped from the analysis . 7 A proxy for general labour force experience has been constructed . To the legal school leaving age (15 years of age, this age also representing the transition to sec- ondary education) the average time needed to finish a particular level of education was added . It was assumed that the first entry into employment occurred immedi- ately after completing education, and that the employment career was not interrupt- ed by non-employment spells . 8 Originally seven subsequent participation in training courses had been record- ed, but only the first four participations showed high enough number to be included in the analysis . 9 Since the interviews were conducted in May, June and July 1994, this period includes the years 1988-1993 and the first half of the 1994 . 10 All employment episodes are included, regardless of their destination states . 206 DR, Vol .XII(1996)22-23 INVESTMENT IN EDUCATION AND TRAINING AND SHIFTS BETWEEN JOBS IN THE SLOVENIAN LABOUR MARKET 11 In the case of overall job moves, compared with primary education having a university or a post-university degree increases the risk of leaving the original job for another job by about 188%, college education by about 111%, 4- and 5- year sec- ondary education by 56 % . The lowest increase applies to vocational secondary edu- cation, about 34 %. 12 In the case of intra-firm job shifts, vocational education increases the transi- tion rate by about 43%, 4- and 5 year secondary education by about 64 % , a college diploma by about 115 % and university and post-university diplomas by about 188 %. For inter-firm shifts the corresponding figures are as follows : 63%; 68 %; 128 %, and 224 % . P 13 According to the job-matching thesis (Jovanovic 1979, Granovetter 1986) young people at the beginning of their employment career shop around for a job which best matches their expectations and abilities, while later on they settle in a job . 14 The period after the completed second educational episode at secondary level increases the transition rate by about 110 % . 15 According to statistical data, the largest share of workers with post-secondary education is employed in service sector, especially in social services (education, health, research) ; reforms of these services often bring about a requirement for high- er levels of education . 16 According to the results the period of transition increases the risk of internal job shifts by about 110 % . REFERENCES : Allmendinger Juta (1989), Educational systems and Labour Market Outcomes . European Sociological Review, Vol.5, No .3, Dec ., pp . 231-250 Becker, G. (1975), Human Capital : A Theoretical and Empirical Analysis with Spe- cial Reference to Education, 2nd ed ., Chicago, University of Chicago Press Blau, P. and Duncan, 0 ., D. (1967), The American Occupational Structure, New York, Wiley Bills David B. (1988), Educational credentials and Promotion : Does Schooling do More Than Get You in the Door? Sociology of Education, Vol . 61, No. 1,Jan ., pp . 52 - 60 Blossfeld, H ., P ., (1986), Career Opportunities in the Federal Republic of Germany : A dynamic approach to the study of life-course, cohort and period effects, European Sociological Review, 2, pp .208-225 Blossfeld, Hans, P. (1987) : Entry into the Labour Market and Occupational Career in Federal Republic of Germany. International Journal of Sociology, Vol . 17, pp . 86 - 115 . Blossfeld, H ., P ., Hamerle A. and Mayer, K., U . (1989) : Event History Analysis, Hillsdale, NJ : Lawrence Erlbaum Blossfeld, H ., P ., Rohwer, G ., (1995) : Techniques of Event History Modelling: New Approaches to Causal Analysis, Hillsdale, NJ : Lawrence Erlbaum Caroll G . R. & Mayer K . U. (1986), Job-shift Patterns in the FRG: the Effects of social class, industrial sector and organizational size, American Sociological Review, Vol. 51, pp. 323 -341 . Čibron, A. (1990), Delovna kariera v organizaciji Magistrsko delo, Univerza Edvar- da Kardelja v Ljubljani, Fakulteta za sociologijo, politične vede in novinarstvo, Ljubl- jana; Prevalje DiPrete, T., A. (1993), Industrial Restructuring and the Mobility Response of Ameri- can Workers in the 1980s . American Sociological Review, Vol . 58, No. 1, pp . 74-96 DiPrete, T., Krecker, Margaret, L. (1991), Occupational Linkages and Job Mobility within and across Organizations, Ducke U . Durham. Research in Social Stratification and Mobility 10, pp . 91 - 131 . DR, Vol .XII(1996)22-23 207 Angela Ivančič DiPrete, T ., A. and Soule, W ., T . (1988), Gender and Promotion in Segmented Job Ladder Systems, American Sociological Review, Vol . 53, pp. 26-40 Doeringer, P., B., Piore, M ., J . (1971), Internal Labour Markets and Manpower Analysis. Lexington, Massachusetts, Heath . Drobnič, S. (1995), Metodološki vidiki proučevanja kvalitete življenja . V: Svetlik., I. (ed .), Kakovost življenja v Sloveniji, TiP, Fakulteta za družbene vede, Ljubljana Granovetter, M. (1986), Labour Mobility, Internal Markets, and Job Matching: A Comparison of the Sociological and Economic Approaches, Research in Social Strati- fication and Mobility, Vol .5, pp .3-39 International Labour Organization (1990), The International Standard Classifica- tion of Occupations, Geneva . Ivančič, A . (1995), Vseživljenjsko izobraževanje in socialna neenakost . In : Svetlik, I . (ed .), Kakovost življenja v Sloveniji, TiP, Fakulteta za družbene vede, Ljubljana Jovanovic, B . (1979), Job Matching and the Theory of Turnover, Journal of Political Economy 87, No . 5, Part I, pp . 972-990 Kerckhoff, A ., C . (1990), Educational Pathways to Early Career Mobility in GB . Research in Social Stratification and Mobility, Vol .5, pp . 131-157 . Koenig, W. & Mueller, W. (1986), Educational Systems and Labour Market as Determinants of Work Life Mobility in France and West Germany: a Comparison of Men's Career Mobility . European Sociological Review, Vol.2, No .2, Sept., pp . 73 - 96 Mach, Mayer and Pohoski (1994), Job changes in the Federal Republic of Germany and Poland:a longitudinal assessment of the impact of welfare capitalist and state socialist labour market segmentation . European Sociological Review, Vol . 10, No .1 Mayer, K.,U . and Carol], G., R. (1987), Jobs and classes : structural constraints on career mobility, European Sociological Review, 3, pp . 14 - 38 Mincer J . (1974), Schooling, Experience and Earnings, New York, National Bureau of Economic Research . Mohorčič Špolar, V., lvančič, A . (1996) : Potreb po izobraževanju odraslih, študije in raziskave 5, Andragoški center Republike Slovenije, Ljubljana Rohwer, G. (1994), TDA working papers, Bremen, University of Bremen Sewell, W., H . and Hauser, R., M. (1975), Education, Occupation and Earnings, New York, Academic Press Soerensen, Aa .,B . (1983), Processes of Allocation to Open and Closed Positions in Social Structure . Zeitschrieft fuer Soziologie, Jg . 12, Heft 3, pp . 203-224 Soerensen Aa.,B . (1977): The Structure of Inequality and the Process of Attain- ment. American Sociological Review, Vol . 42 ., pp . 965 - 978 Soerensen Aage (1975), Models of Social Mobility . Social Science Research, Vol .4, pp . 65 - 92 Soerensen Aa., B. & Tuma, N ., Brandon (1981), Labour Market Structure and Job Mobility, Social Stratification and Mobility 1, pp . 67 - 94 . Thurow, L ., C . (1975), Generating Inequality, New York, Basic Books Tuma, N., B. (1985), Effects of Labour Market Structure on Job Shift Patterns . In : Heckman, J . and B ., Singer (Eds .), Longitudinal Analysis of Labour Market Data, Cambridge University Press, Cambridge Wholey, D .(1990), The Effects of Formal and Informal Training on Tenure and Mobility in Manufacturing Firms, The Sociological Quarterly, Vol . 31, No . 1, pp. 37- 57 208 DR, Vol .XII(1996)22-23