1.1 Labour Supply

1.2 Employment trends in Austria

1.3 Wages and productivity, competitiveness of the Austrian economy

1.4 Unemployment and its structure

1.5 Flow dynamics in the Austrian Labour Market

2.1 Stability Features of the Fundamental Economic Policy Settings

2.2 The Cyclical Volatility of Output, Employment, Labour Supply and Unemployment

2.3 Real Wage Flexibility, Insider-Outsider Behaviour

2.4 Trend increases in structural unemployment

3.1 Wage bargaining, wage flexibility and differentiation

3.2 Industrial Relations in Austria

3.3 The tax and contribution system and the role of labour taxes

3.4 Social security system in general and unemployment benefit system in particular

3.5 Active labour market policy in Austria

3.6 Education and training

3.7 Product market competition, employment barriers and the business environment

3.8 The National Action Plan

Figure 1.6 Early retirement by type of retirement

Source: WIFO, Main Association of Social Insurance Institutions.

1.2 Employment trends in Austria

In an EU comparison Austria's employment rate holds an excellent position, coming in third after Denmark and Sweden. Austria's situation is even better if the comparison figures are adjusted for part-time employment by using full-time equivalent employment rates (see figure 1.7). Denmark continues to have the highest employment figures, followed by Austria. Also significantly above the EU average according to this indicator are Portugal and Sweden. The employment situation can also be differentiated by sex. While Austria has the highest rate of employment among men, the ranking regarding women is Denmark, Sweden and Finland, followed by Portugal and Austria.

Figure 1.7 Employment rates (in full-time equivalents) 1996

Although the Austrian labour market situation is very favourable by international standards, some gradual deterioration has occurred since the 1970's. After the first oil crisis the dynamic employment growth of the early 1970's had been lost. Employment growth in the period 1975 to 1981 was slow and more or less in line with the growth of employment in the whole EU area. The weaker demand on the labour market did not translate into rising unemployment, however, due to the reduction of foreign labour and to the decline in employment among older workers.

Figure 1.8 Growth in total employment in Austria and EU-15

Source: OECD ECO.

The adjustment problems caused by the second oil price shock and the change to the hard currency policy lead to an unfavourable development on the Austrian labour market. In the early 1980's employment dropped significantly for the first time, hitting men the hardest. Employment among women registered a smaller decline and rose slightly until 1988. The number of registered unemployed had tripled, with men showing a stronger increase than women. The drop in employment was especially sharp in manufacturing. Employment growth in Austria was lagging behind the EU average (see figure 1.8).

The poor global economic trends in the 1980s brought various problems to the fore, first and foremost the accumulated state debts, the related interest load and the aggravated structural problems (see Rothschild 1989). The increasing internationalisation of the Austrian economy and the problematic fiscal stance prevented the Austrian government from an expansionary fiscal policy to promote employment. The major crisis of state-owned industry also came in this period. This mainly involved enterprises in the basic goods and heavy industries. The policy of keeping existing jobs in the state-owned industry impeded the necessary structural change. Although attempts were made to restructure these industries and develop new markets for them, such efforts met with little success. When futures transactions posted major losses in the mid-1980s, the incumbent government resolved to radically reform state-owned industry. The reforms also introduced the possibility of privatisation and the state's ownership in these enterprises was reduced sharply. Most of the large companies that operated on many markets and had numerous business locations were broken down into smaller units in accordance with international practices. State enterprises were soon turned into businesses that were more than capable of surviving on the market. However, employment in the former state-owned enterprises is considerably lower as it was in the 1970s.

The period between 1988 and 1992 is characterised by a strong economic upturn and a new phase of dynamic employment growth. Austria could profit above average from the opening up of the Central and Eastern European economies and from the German Reunification. In these years employment grew with an annual rate of 1 ¾ %. The share of foreign workers in total wage and salary earners increased from 5.2 % to 9.0 %. Despite considerable growth in employment, the number of unemployed hardly declined. Although unemployment remained low compared to the EU average, unemployment nevertheless became a central problem for economic policy. After 1992 the employment level more or less stagnated, however, at a high level. Adjustment problems caused by Austria's joining the EU, the economic situation in Germany and the fiscal consolidation dampened labour demand.

Table 1.2: Growth of Total Employment, 1980 - 1996

* self-employed are included.
Source: own calculations, OECD Economic Outlook.

All in all, total employment in Austria rose by 143.000 persons since 1980.⁽⁵⁾ The growth rate of employment was smaller than in other comparable EU countries (see table 1.2). The best employment performance of the selected countries in this time period was shown by the Netherlands with 1 Million new jobs, all were created in the private sector of the economy. Employment growth was also stronger in Ireland and Denmark as in Austria.

A number of diverging trends were behind the development of total employment in Austria. Only 50.000 new jobs for wage and salary earners had been created by the private sector. The employment growth was dominated by the government sector, which created 208.000 Jobs. This contrasts with developments in other European countries where employment in the government sector has lost relative weight (see EC 1998). The number of self-employed persons declined considerably. Responsible for this trend is the strong out-migration from the agricultural sector; self-employment declined by 124.000. This huge loss could by no means be compensated by the small increase of self-employment in the industry and service sector of 9.000 persons. The following figure shows the end result of these trends. The share of government employment increased by 5 percentage points since 1980. Accordingly relative employment in the private sector and especially the share of self employment in agriculture has been shrinking.

Figure 1.9 Employment Shares 1980 and 1996

Table 1.3: Employment intensity of growth

Source: OECD Economic Outlook, own calculations.

From 1970 to 1996, the US economy grew at a faster pace than the European Union. Austria's growth was well above the EU average and was actually more on a par with American growth. Comparing these figures to the employment trends provides revealing information on the employment intensity of growth. Even given the output growth, the USA exhibits by far the greatest increase in employment. Employment rose on average by 0.6 percentage points for every one percentage point increase in output. Austria's employment intensity is significantly lower⁽⁶⁾, but still above the EU average. The hypothesis that the employment intensity of economic growth is declining as a result of technological progress certainly does not apply to Austria in light of the data provided. In comparison to the 1970s and particularly to the 1980s, the employment intensity of growth even increased between 1990 and 1996.

In the period 1980 to 1996 the employment intensity of growth differs considerably across the Netherlands, Denmark, Ireland and Austria. The employment intensity in the Netherlands is almost four times as high as in Austria or in Ireland, and two times stronger than in Denmark. These results seem to indicate that the impressive employment growth process in Ireland is probably mainly the result of catching up with the rest economically, whereas the rise in employment in the Netherlands resulted in part from expanding part-time employment⁽⁷⁾.

Figure 1.10 displays the structural change in employment between 1982 and 1996 in Austria. The share of agriculture declined by 2 ¾ percentage points and relative employment in the manufacturing sector has shrunk by 8 ¾ percentage points, whereas employment in the service sector has considerably gone up. This sectoral employment pattern is in line with other OECD countries.

Figure 1.10 Sectoral breakdown of employment in Austria

Source: OECD, LFS.

In comparison to the EU average, Austria has a relatively traditional economic structure, the below-average share of employees in the service sector is compensated by the higher share employed in agriculture (see figure 1.11). This impression is reinforced if the comparison is made with the Netherlands, Denmark, Sweden, France or the UK. Austria's share of persons employed in agriculture and forestry is much higher. In contrast, comparatively few jobs have been created in the service sector in Austria. The relative employment rates are 5 to 10 percentage points behind the other countries in spite of the significance of the tourism industry in Austria.

Figure 1.11 Employment by sector 1996

Source: EU Report on Employment 1997.

Table 1.4 allows a closer look at the sectoral decomposition of wage and salary earners in Austria. The Austrian share of employees in the industry sector is well above Denmark or especially the Netherlands. Furthermore, the construction sector in Austria is large in comparison to the other European countries. Conversely, the Austrian share of wage and salary earners in the service sector is below average. In this respect one has to consider the importance of the tourism sector for the Austrian economy. Therefore, the more traditional service industries such as trade, hotels and restaurants account for a relatively high share. However, employment in the more dynamic service industries like business services is still very low in Austria.

According to data from the EU labour market survey roughly 14.5 % of total employment in Austria are self-employed (including family members that help out in the family business). The structure in Austria corresponds to the EU average. The highest rate of self-employment is to be found in Mediterranean countries such as Greece, Italy, Spain and Portugal. At the other end of the scale are Denmark, Luxembourg and Sweden. An important factor influencing the share of self-employed persons is the sectoral structure of production in a country. Countries with a relatively large share of agriculture also have higher rates of self-employed persons.

Table 1.4: Wage and salary earners by sector in 1996 (in % of total)

Source: EC 1998.

For this reason the figure below differentiates between self-employed persons in agriculture and forestry, and in other industries. The resulting picture is quite different. While Austria holds second place after Ireland regarding self-employed persons in agriculture, only Denmark ranks lower than Austria in the non-agricultural industries. Ireland and the UK exceed Austria's level by good measure, while France and Sweden have more or less the same share of self-employment in non-agricultural industries as Austria. However, the comparison with Ireland and the Netherlands indicates some potential for business start-ups.

Part-time work is gaining ground within the EU. Since 1985, the share of part-time jobs rose from 10.8 % to 16.4 %. A similar development can be observed in Austria (see table 1.5). Between 1983 and 1994 the share of part-time employment in total employment increased by one third. Part-time employment is dominated by women. Over time the part-time employment of women increased slowly but steadily. Whereas in 1974 only 4.6 % of all working women⁽⁸⁾have been working between 14 to 36 hours, the share of women working 12-35 hours increased to 23.6 % in 1996 (Mühlberger 1998:104). However, part-time employment of male is still on a low level.

Figure 1.12 Self-employment by sector, 1996

Source: EU Report on Employment 1997, OECD, own calculations.

According to the Austrian labour market survey 15 % of all employed persons worked part-time in 1996.⁽⁹⁾ In absolute numbers 538,000 persons worked less than 35 hours a week. In international comparison Austria was in the middle range showing a part time share of 28 % among women, which is much higher than the share of men at 4 %. The share in Austria corresponds roughly to EU averages (total of part-time jobs 16.5 %; men 5.5 %; women 31.5 %). There are clear regional disparities within the EU. While the percentages in the Mediterranean countries of Greece, Italy, Spain and Portugal as well as Luxembourg are below 10 %, in the Netherlands 38 % of employed persons do not hold full-time jobs. Part-time employment rates in the UK, Sweden and Denmark are between 20 % and 25 %.

Table 1.5: Development of part-time employment in Austria

*1974 instead of 1973.

Source: Bastelaer et al. 1997, Mühlberger 1998.

Roughly 10 % of part-time employees in Austria hold jobs that may be considered almost full-time jobs (31 to 35 hours a week), 37 % work between 21 and 30 hours a week and the majority of part-time employees, i.e., 42 %, work between 11 and 20 hours a week. Only 11 % were employed for less than 10 hours. Employment for less than 10 hours a week is much more common in Denmark, the Netherlands (28 %) and the UK (25 %) than in Austria. The average number of weekly working hours for part-time employees in Austria is 22 hours, which is 2 hours more than the EU average.

The labour market survey carried out in 1995 explains the reasons for part-time work (Bartunek 1997). Of the 510,000 part-time employees in Austria, 37,000 (29,000 women) did not work part-time voluntarily, but were not able to find full-time jobs. Some 90,000 (77,000 women) did not want full-time jobs and 36,000 persons were only able to take on part-time employment because of their low level of education and/or vocational qualification. The majority, 335,000 (297,000 women) persons, claimed "other" reasons, which in many cases refer to the family situation, such as childcare or care of other family members (see Bartunek 1997).

A closer look at the part-time jobs offered shows quite clearly that most are low-quality jobs: most jobs offered are for cleaning personnel and restaurant helpers, the most jobs sought are in office and clerical work. Women who are not able to work full-time for personal or family reasons are often forced to change workplace or must accept less qualified work. In 1992 60 % of the females working part-time are semi-skilled or unskilled workers or white collar-workers performing unskilled work (Mühlberger 1998).

A rising tendency is shown for the number of employed persons working for pay below the level of significant employment (ATS 3,830), who up until 1997 were not covered by the obligatory insurance scheme. However, more than half on the these workers are covered by additional insurance schemes. Between 1994 and 1997 marginal employment increased from 117.209 to 164.445 persons, 72 % of which were jobs held by women. Roughly two-thirds of all such workers were employed in the sectors of wholesale and retail trade, business services, accommodations, manufacturing and health and social work. Since the flexibilisation of shopping hours marginal work in retail trade has surged. Furthermore, there exist some empirical hints that students are the main part of the male marginalised workers; with respect to women re-entry after inactivity dominates (see Mühlberger 1998).

Another type of contingent employment in Austria is hiring-out of labour ("Leiharbeit"). The number of temporary workers grew from 7.955 in 1989 to 17.980 in 1997, with increasing pace since 1994. However, the share of temporary workers in total wage and salary earners is still very low (0.9 %). 87 % of temporary workers are male and more than half are employed in manufacturing (see Mühlberger 1998).

1.3 Wages and productivity, competitiveness of the Austrian economy

The following graph illustrates the trend in real wages per active employee⁽¹⁰⁾(at producer prices) and labour productivity for Austria. It clearly shows that wage policy tends to take its lead in the medium-term from price and productivity trends. Over the business cycle, wages and productivity do repeatedly, but only briefly, diverge from one another, but in the medium term the real wage gaps that occurred were always closed again. In interpreting this data, however, one should naturally bear in mind that both labour productivity and real wages are endogenous variables in the economy and are mutually dependent on one another.

Figure 1.13 Real wages and labour productivity (1976-1996=100)

* real wages per active employee and GDP per active employee.

Source: IHS.

Changes in labour costs and productivity differentials have a strong influence on a country's competitiveness. In 1996 a man hour in Austrian industry cost ATS 270.10, placing Austria among the countries with the highest labour costs. The only two other countries that currently have distinctly higher labour costs are western Germany and Switzerland; labour costs in Belgium, Denmark and Norway are only slightly higher than in Austria. In the long run, Austria's cost position has deteriorated significantly. Data from Cologne's Institut der deutschen Wirtschaft shows that 12 countries had higher labour costs than Austria in 1980.

Figure 1.14 Total labour costs and unit labour costs
Austria versus average of trade partners

Source: WIFO.

Real wages in industry grew by 30 % in Austria in the period from 1980 to 1996; Japan, western Germany and the UK recorded similar trends in real wages. In Denmark real wages rose at only half the rate and the Netherlands, Sweden and Switzerland only achieved one-third of Austria's real wage growth. But it was only thanks to the above-average increase in labour productivity that Austria was able to raise its real wages to this extent. Among the OECD countries used in comparison, Ireland was the only one to achieve stronger productivity gains in the 1980s. In the first half of the 1990s, the improvement in Austria's productivity even accelerated and only Finland and Ireland posted higher efficiency gains (Guger 1997). The rapid rise in productivity in the 1990s was, however, combined with a marked drop in employment: In the first half of the 1990s the level of employment in industry dropped by about 3 %.

But in order to evaluate an economy's competitiveness, it is not labour costs per se that are the relevant factor, but rather unit wage costs, which are calculated as the ratio of costs per man hour to the hourly rate of productivity. In the 1980s Austrian industry considerably improved its competitive position relative to its Western trade partners. Although the growth in productivity of Austrian industry even sped up in the first half of the 1990s, unit wage costs deteriorated relative to those of its trade partners due to the appreciation of the schilling and the marked slowdown in unit wage cost increases in competitor countries (Guger 1997). Due to the continuous increases in domestic productivity and appreciation of the currencies of some of Austria's most important trade partners, Austria has considerably improved its relative unit wage cost position since 1996, so much so that it has now returned to the unit wage cost position it held over its trade partners in 1990.

Table 1.6: International comparison of labour costs, hourly productivity and unit wage costs in manufacturing

(1) The average of Austria's trade partners is weighted pursuant to the WIFO exchange rate index and labour costs are weighted according to the number employed in industry.

Source: Guger 1997.

Political stability, EU membership, proximity to the expanding markets of Eastern Europe, a pro-business tax system, favourable environmental standards and in particular the social stability and the high quality of the work force are the main merits of the business location Austria. However, there are also important challenges. Austria has proven itself to be especially competitive in products with medium value-added intensity and on nearby markets. The productivity gains came primarily in the industrial sector of the economy. However, there are clear indications that the cost and productivity ratios in heretofore protected areas of the economy are substantially less favourable (Kramer 1997). Austria's exports focus on medium-range technologies in resource and labour-intensive segments. However, they are lacking in modern technologies. This may be attributed in part to the below-average research and development activities. Only 1.5 % of GDP were spent for R&D in 1996. Although Austria's share of technology and research has, indeed, improved in the long term, it does not rank among the industrial elite (Kramer 1997). Other weak points, aside from giving high-tech short shrift, are its concentration on neighbouring markets instead of global markets and the lack of innovative business clusters. Furthermore, Aiginger and Peneder (1997) argue that it is necessary to modernise the regulation system for entrepreneurial activities in various aspects, e.g. rules for working time, efficiency of public administration, access to entrepreneurial activities.

1.4 Unemployment and its structure

Unemployment in Austria has remained significantly below the levels of other EU countries over the last two decades. Using the harmonised definition of the European Statistical Office, which is based on the labour force concept and uses data from the labour force survey, yields an unemployment rate of around 4.4 %.⁽¹¹⁾ With the exception of Luxembourg (3.7 %) Austria has the lowest rate within the European Union. In accordance with the majority of the other EU countries, unemployment is slightly higher for women (5.3 %) than for men (3.6 %).

Figure 1.15 Unemployment rates in the European Union

Source: EU Report on Employment 1997.

A major problem in Europe is youth unemployment. Roughly every fifth European citizen between the ages of 15 and 24 was unemployed in 1996. Exorbitant levels of youth unemployment were posted in Spain, Finland and Italy, and above average rates in France, Belgium and Sweden. However, with 6.0 %, Austria's rate of unemployment among 15 to 24-year-old youths is by far the lowest in the European Union. Only Luxembourg, Greece and Germany have single-digit figures to show. One of the reasons for the good performance of Germany and Austria is the apprenticeship system, which facilitates school to work transition considerably. While youth unemployment is low in Austria, the situation for the group of older workers may be considered relatively more problematic. Of course, the unemployment rate of the 55 to 64-year-olds is 4.6 % and much lower than the EU average of 10.6 % (OECD 1997a). A country comparison places Austria at the bottom of the lower third of all EU countries, with only small differences between the countries in this group.

Almost every second unemployed person in the EU is out of work for over a year. In contrast, in Austria only every fourth is long-term unemployed. Together with the Scandinavian countries of Sweden and Denmark, Austria shows the best performance regarding long-term unemployment. What is less comforting, however, is the direction that this indicator is taking: concurrently with the other EU countries, a clear upward tendency in long-term unemployment has to be recorded.

Despite the low unemployment by international standards, unemployment has risen considerably since 1980 (see figure 1.16 and the discussion of employment dynamics).⁽¹²⁾ International comparisons reveal that in Austria unemployment rises in business downturns have not been very sharp. Therefore, Austria's favourable position is not the result of considerable declines in times of strong business activities.

Despite the current business upturn in Austria, there is no tendency for a noticeable decline in the unemployment rate. This contrast sharply with developments in Denmark or the Netherlands in the recent past. The stickiness of unemployment can be explained at least partly by the strong cyclical elasticity of the labour force in Austria. Furthermore, some changes in institutional factors, e.g. the shortening of maternity leave from 2 to 1 ½ years, the reduced possibilities for early retirement and closely connected the abolishment of the "Sonderunterstützung", tend to increase "measured" unemployment. However, there are also segmentation processes in the unemployment pool to observe. For example, the number of unemployed, which are hard-to-place, mainly women, has further increased (37 % of total unemployment in June 1998).

The national data, compiled by the Main Association of Social Insurance Institutions and the Public Employment Service (AMS), allows a more detailed analysis of registered unemployment. Due to the availability of detailed data we concentrate on 1996.⁽¹³⁾ On average in 1996, 230,000 persons were registered as unemployed with AMS. The registered rate of unemployment was 7.0 % (men 6.9 % and women 7.3 %) of the dependent labour force .

Figure 1.16 Unemployment in Austria and EU

Source: OECD ECO, IHS; the definition of the unemployment rate shown here follows the commonly used definition concept used by the OECD.

The dynamic of the Austrian labour market is illustrated by the fact that roughly 700,000 persons were affected by unemployment in 1996. In a five-year period every fourth worker received unemployment benefits or unemployment assistance; in a 20-year time period even every third worker (Synthesis 1997). The considerable turnover in the Austrian labour market is also reflected in the dynamics of the movement in and out of unemployment. In 1996, roughly 700,000 of both inflows and outflows were recorded for unemployment (approx. 20 % of the dependently employed labour force). Remarkable is the parallel development of inflows and outflows over the business cycles. While anticyclical movements of inflows into unemployment correspond to a priori expectations, it may seem a bit surprising that the outflow rate also rises in business cycle downturns. This means that higher unemployment in periods of slumps goes hand-in-hand with a higher rate of replacements within the system of employment. This mechanism thus intensifies the trend toward a pattern in total unemployment, in which the share of persons with relatively low re-employment probabilities is on the rise.

The risk of becoming unemployed and level of education clearly correlate negatively. The majority of persons affected by unemployment do not have education beyond the compulsory school level or apprenticeship training. The share of workers belonging to these two categories among the unemployed was 85 %. Roughly 11.5 % of workers with only compulsory education are unemployed. The figure is 6 % for those who have completed apprenticeship training. For more highly qualified workers the rate of unemployment is only half as high. Every third unemployed adult is hard-to-place. In 45 % of cases, (of which 90 % are women) the reasons given for the difficulty in job placement are limited mobility regarding working hours and place of work.

The distribution of unemployment by age group shows markedly below-average rates for the groups at the tails of the age distribution (15 to 18-year-olds, 60 years and older), slightly above/below average rates for 19 to 29-year-olds and 30 to 49-year-olds, respectively, and unemployment rates far above average for the age group of 50 to 59-year-olds. While the probability of becoming unemployed decreases with age, the average duration of unemployment spells rises with age. 22 % of unemployed youths accounted for 17 % of the volume of total unemployment. In the case of older workers the ratio is the other way around: 13 % of those affected were 50 and older, which makes up approximately 17 % of the total burden.

The average duration of an unemployment spell was 121 day in 1996: for women 131 day or 16 days more than for men. Approximately half of the unemployed were registered for a period up to 3 months, every fifth between 3 and 6 months. This view of averages, however, should not obscure the fact that the total duration of unemployment is very unevenly distributed among the persons affected. The upper 20 % of the unemployed with the largest duration (9th and 10th deciles) account for almost half of the burden of unemployment. Their share in the volume of total unemployment was 47.6 %, while the lowest quintile held a share of 3 %, which is hardly noticeable (AMS 1997). Furthermore, it should be noted that only (completed) unemployment spells have been discussed so far. But since persons also experience several spells of unemployment within a year (20 % of unemployment leavers had 2; 2.5 % had 3 or more), the cumulated duration of unemployment was 139 days for men and 167 days for women.

Table 1.7: Registered unemployment in Austria, 1996

¹ Agriculture and forestry, construction and tourism

² Unemployed persons with episodes lasting at least 180 days.

Source: AMS.

The problem of long-term unemployment has also become significantly worse in Austria since the mid 1980s. In 1980 every twelfth unemployed adult was unemployed for longer than six months, by 1996 every fourth. When the common international duration definition of one year is applied, the number of people affected by long-term unemployment in 1996 was 72,786. These cases are strongly concentrated in the retail and metalworking sectors (a total of 38 %), and in construction and tourism (19 %). With rising age, the share of persons affected by long-term unemployment goes up markedly from 1.0 % (15 to 18-year-olds) to 10.1 % (30 to 39-year-olds) reaching 24.8 % and 29.2 % respectively for the age groups of the 50 to 54-year-olds and 55 to 59-year-olds (AMS 1997).

Table 1.8: Unemployment by skill 1987, 1990, 1996

* in percent of dependent labour force; dependent employees are distributed by education according skill distribution of microcensus; 1 Secondary technical and vocational schools (3-years course) 2 Secondary general education school / secondary commercial school (5-years course).

Source: own calculations.

Table 1.8 presents evidence on trends in unemployment for unskilled relative to skilled workers; education is used as indicator for skills. Unemployment rates have increased for all skill groups over the period 1987-1996. Total unemployment rose by 65.000 persons, of which ¾ are low-educated (compulsory schooling or apprenticeship) workers. Unemployment rates of high-skilled workers has been going up as well, but remaining at a relatively low level. The difference between the unemployment rates of low- versus high-educated workers increased considerably. The relative risk of becoming unemployed for a worker with compulsory schooling is almost five times as high than for a worker with a university degree. In sum, the labour market changes of low-skilled workers deteriorated over time. The relative risk of unemployment for low-skilled workers has gone up in Austria. Note, however, that the unemployment risk for workers with compulsory schooling is still two times higher than for people with completed apprenticeship.

Austria is a country with one of the highest rates of seasonal unemployment in Europe. In a rough estimate, about one-third of all unemployment spells per year and almost one-fourth of total unemployment is due to seasonal fluctuations in employment. Unemployment insurance is a form of a implicit subsidy for those industries that use wage compensation payments paid during unemployment to finance periodic and predictable seasonal fluctuations. Seasonal unemployment produces substantial monetary costs. Estimates for 1993 calculate that direct costs (unemployment benefits and unemployment assistance) amounted to 3.5 billion ATS. If the indirect costs (social security contributions not paid and payroll taxes) are also taken into account, this adds another ATS 4 billion (Brandel et al. 1994b).

Closely connected and in part identical with seasonal unemployment is the problem of so-called "temporary layoffs" (Pichelmann and Riedel 1992), i.e. unemployment spells terminated by reemployment with the former employer. In 1985 approximately one-third of all unemployed persons were affected by this type of unemployment (Fischer and Pichelmann 1991). New studies indicate that this percentage has probably risen again. In Styria, for example, almost every second person leaving unemployment into employment went back to the former employer in the years between 1994 and 1995 (Zilian 1998). A similar picture exists in other Länder as well⁽¹⁴⁾. Zilian expands the concept of seasonal unemployment and has coined the term "flexibility unemployment". Re-employment with the former employer includes traditional seasonal workers, workers in those industries that have started developing seasonal structures in order to be able to counteract the costs of capacity utilisation fluctuations, and last of all, those unemployed persons who were dismissed and then rehired for the purpose of avoiding personnel expenses caused by idle labour capacity.

1.5 Flow dynamics in the Austrian Labour Market

The rate of job reallocation that occurs between existing enterprises in Austria is considerably. For example, at the beginning of the 1990s, roughly 5 % of existing jobs disappeared⁽¹⁵⁾, but this loss was more than made up for by jobs created in other enterprises. While in the period from 1991 through 1992, employment gains were achieved, the job destruction rate accelerated during the 1992-93 slump by approximately 1.5 percentage points and gross job losses were around 7 % of existing jobs. In the course of the economic recovery in 1994 the job creation rate rose again to 6 %, while the job destruction rate remained at its high level and even rose slightly.

Table 1.9: Job turnover in continuing firms (in % of total employment)

Source: OECD Employment Outlook 1996, for Austria IHS.

The rate of job reallocation in Austria corresponds roughly to the rates in comparable European countries. The present data somewhat underestimates the job turnover rate, because newly founded enterprises or closures have not been taken into account.

Table 1.10: Job turnover by industry and enterprise size in 1990-91

Source: IHS.

According to the above data, job turnover has a high degree of significance in relation to the changes in net employment. At a constant overall level of employment, an estimated 100,000 jobs are lost in Austria every year in enterprises that reduce in size or close down. The reallocation of jobs has concentrated on movements within the individual industries, while sectoral shifts in employment patterns only account for roughly one third of the annual job turnover rate.

Table 1.11: Distribution of job creation/job destruction by firms, 1990-91

Source: IHS.

Considerable shifts between shrinking and expanding businesses are also taking place within the individual industries. Tourism has the highest turnover rate: the job turnover rate is 20 percent per year. In addition, the construction industry and the wholesale and retail sector have high turnover rates; the same applies to the remaining service industries with the exception of local public bodies. Manufacturing shows a job creation rate well below average, which also pushes the job turnover rate down. Furthermore, there is also a clear link between the size of an enterprise and the job turnover rate. The larger the enterprise, the lower the job turnover rate in percentage of persons employed. However, the conclusion cannot be drawn that primarily smaller enterprises are responsible for the largest net job gains, because not only the job creation rates but also the job destruction rates are far above average in this case. In addition, roughly 15 % of jobs created were in the group of enterprises with more than 100 employees in spite of the low job creation rate of 3 to 4 percent. However, during the recession of 1992-93 almost one-fifth of jobs in these enterprises disappeared.

An interesting empirical aspect is the fact that the creation of new jobs and the elimination of jobs tends to concentrate on a few enterprises. Roughly half of new jobs are created in 10 % of the expanding enterprises, while only 10 % of shrinking enterprises are the cause for more than 50 % of jobs eliminated. It should also be noted that these statistics do not take account of start-ups or closures.

Much higher than the job turnover rate is, of course, the rate of turnover in employment as persons move between existing jobs. Taking account of reported hiring and firing on short notice, seasonal fluctuations, summer jobs, etc., the number of employment contracts commenced or terminated per year in Austria is around 1 million. Every third employment contract is a new one established in the course of a year (Synthesis 1997). Within the past 20 years, over 6 million persons were employed at least once, of which only 38 % held jobs for at least 10 years. On average, every employed person has held 3.3 jobs lasting on average roughly 2.5 years (Synthesis 1997).⁽¹⁶⁾ Rising dynamics has been observed over the past years in this context. While the average number of newly commenced employment contracts per year was roughly 830,000 between 1976 and 1990, the number climbed to one million in the following years (Synthesis 1997).

Table 1.12: Job tenure on enterprise level by sex (duration in years until cut-off day in May 1994)

* Only persons over the age of 25 have been included.

Source: IHS.

Table 1.12 presents further evidence on the (incomplete) duration of employment spells and a few other indicators on stability and duration of employment in Austria. In May 1994, the median duration of employment per employee (referring to employment until cut-off day in May 1994) under contract was slightly longer than two years; the lowest fifth of employees had been under contract for less than five months, the upper fifth of employees for longer than 6.4 years.

The probability of having been employed for longer periods of time obviously rises with age. But even among those employees over 40 years of age, more than 20 % had held jobs lasting less than one year. The variation in the length of employment contracts for men and women has been negligible to date, at least in the aggregates. However, the impact of industry structures and age remains to be investigated.

Table 1.13: Job tenure by status, age group and industry (median values: duration in years until May 1994)

Source: IHS.

What is remarkable is that almost half of the employees surveyed had had more than one employer within the past five years and more than half of the employees had experienced two or more different job episodes. According to this indicator, the stability of the employment status tends to rise with age, but, for example, even among the 40 to 55-year-olds only 57 % had held only one job within the past five years.

Tabl