Longitudinal Changes of Establishment Level Co-variables

Next, I will turn to presenting descriptive information of the structure of Finnish labour markets in relation to starting and ending employment tenures and their variance between different establishment types. The main concern here is to examine how the industrial branch of the establishment, annual accounts

turnover, changes in turnover, number of personnel, and ownership type of the establishment differentiates the picture that lies behind the starting of employment annually and the length of employment tenures, in general. These variables are not in this descriptive part related to each other. However, short observations are made about the relationships between variable categories.

Industrial Branch

In Finland, the largest industrial branches from the perspective of started employment are basic industry and basic services; roughly two-thirds of all employment starts in one of these branches. However, the share of these branches is diminishing and the share of new services, such as the R&D sector as well as consulting and planning services, has been growing since the 2000s. The tendency is that basic industry employment is more likely to end and basic services employment is less likely to end during the first year of employment.

Despite the commonly held view that service workers have short-term employment tenures, there seems continuously to be a lower probability of employment ending compared to basic industry. However, in the new services branch, there is a steadily growing probability of employment ending, when compared to basic industry. Furthermore, it seems that in the high technology industry, there is the lowest probability of employment ending as compared to other industries during the entire observation period (Figure 23; Appendix 9).

Previously discussed, we noticed that females tend to have longer tenures than males. This observation is probably interlinked with the concentration of a female labour force in the service sector. Thus, these branch differences might radically change once the gender influence is held constant. In Western labour markets, it has also been seen that those working in hotels and restaurants have the shortest tenures, and that these differences are rather stable across countries and over time (Ayer and Cazes 2000). However, in Finland, it seems that in basic services (including hotels, restaurants and grocery stores), the tenures are slightly longer than in basic industry (Figure 24).

Annual Capital Turnover

The structure of new employment stays quite stable over the observation period in relation to annual capital turnover of establishments. Most employment starts (approximately 70%) in firms that have over € 500,000 annual turnover. Actually, the largest share of new employment is in establishments that have € 500,000 – 5 million in annual turnover. Also, the probability of employment ending during the first year seems to be quite constant over years in different capital turnover establishments (Figure 25; Appendix 10).

When comparing the probabilities of employment ending between different annual capital turnover establishments from 1991–2007, we notice that middle level turnover companies constantly have the strongest probability of employment ending than compared to the lowest turnover groups. This difference grew until 2006 (Figure 26).

Number of Personnel

The structure of new employment remains quite stable over the observation period, and also in relation to the size of establishments. Around 70% of all new employment starts in establishments that have less than 50 persons.

Approximately 20% start in establishments that have 50–199 employees.

Furthermore, the structure of ended employment during the first year of tenures stays rather constant and it does not seem to differentiate between sizes of establishment (Figure 27; Appendix 11).

When comparing establishment sizes to each other, the probability of tenures ending has been lowest in large establishments during the 1990s. However, since 2003, this difference has been decreasing, and in the last years of the observation period, the probability is actually smallest in the small establishments (Figure 28). It is generally hypothesised that tenures should be longer in large establishments because of a better functioning internal labour market. However, empirical results of this phenomenon are mixed throughout the Western world;

in some data the influence varies, in others the hypothesis is abandoned, or sometimes, it gains support (Boockmann and Steffes 2010; Auer and Cazes 2000;

Bellman, Bender and Hornsteiner 2000; Burgess, Pacelli and Rees 1997). In Finland, it seems that there are changes in regard to this hypothesis.

Establishment Economic Situation

Most employment begins in economically stable establishments (50–60% of started employment) and economically growing establishments started to employ larger labour forces once the 1990s recession was over. However, their share of new employment once again diminished after the growth period, and the share of employment started to grow in economically fluctuating establishments at the same time, since 1995. Coming to 2007, the share of started employment was even higher in fluctuating than in stable establishments. This might indicate larger structural change in Finnish industry, and that growing establishments started fluctuating after the first years of recovery. This effect is controlled for in the multivariable Cox proportional hazards model described in next paragraph (Figure 29; Appendix 12).

The probability of ending employment has been highest in diminishing and in fluctuating establishments in Finland. It has been lowest within growing establishments. Furthermore, this is quite predictable and the expected result is rather stable over time and there are only few changes during the observation period. However, growing establishments could have an even clearer lengthening effect on tenures (Figure 30).

Foreign Ownership of Establishments

Foreign ownership is gradually gaining a foothold in Finnish labour markets. At the end of the observation period, 11.5% of new employment was situated in establishments that were foreign. The probability of ending employment during the first year does not seem to be strongly divided between the ownership of establishments. When comparing the overall probability of employment ending between Finnish and foreign owned establishments, we find that Finnish owned establishments have a higher probability of employment ending. However, this difference is decreasing gradually as the share of foreign owned establishments grows within the Finnish labour market (Figure 31; Appendix 13; Figure 32).

4.1.3 Longitudinal Changes of Labour Market Level Co-variables