Mismatch unemployment in Austria: The role of regional labour markets for skills

Böheim, René; Christl, Michael Working Paper Mismatch unemployment in Austria: The role of regional labour markets for skills JRC Working Papers Series on Labour, Education and Technology, No. 2021/08 Provided in Cooperation with: Joint Research Centre (JRC), European Commission Suggested Citation: Böheim, René; Christl, Michael (2021) : Mismatch unemployment in Austria: The role of regional labour markets for skills, JRC Working Papers Series on Labour, Education and Technology, No. 2021/08, European Commission, Joint Research Centre (JRC), Seville This Version is available at: https://hdl.handle.net/10419/236540 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by/4.0/ Centre JRC Technical Report Mismatch unemployment in Austria The role of regional labour markets for skills JRC Working Papers Series on Labour, Education and Technology 2021/08 René Böheim and Michael Christl This Working Paper is part of a Working paper series on Labour, Education and Technology by the Joint Research Centre (JRC). The JRC is the European Commission's science and knowledge service. It aims to provide evidence-based scientific support to the European policymaking process. The scientific output expressed does not imply a policy position of the European Commission. Neither the European Commission nor any person acting on behalf of the Commission is responsible for the use that might be made of this publication. Contact Information Name: Michael Christl Address: European Commission, Joint Research Centre (Seville, Spain) Email: [email protected] Tel.: EU Science Hub https://ec.europa.eu/jrc JRC124896 © European Union, 2021 The reuse policy of the European Commission is implemented by the Commission Decision 2011/833/EU of 12 December 2011 on the reuse of Commission documents (OJ L 330, 14.12.2011, p. 39). Except otherwise noted, the reuse of this document is authorised under the Creative Commons Attribution 4.0 International (CC BY 4.0) licence (https://creativecommons.org/licenses/by/4.0/). This means that reuse is allowed provided appropriate credit is given and any changes are indicated. For any use or reproduction of photos or other material that is not owned by the EU, permission must be sought directly from the copyright holders. All content © European Union, 2021 How to cite this report: Böheim, R. and Christl, M., Mismatch unemployment in Austria: The role of regional labour markets for skills, JRC Working Papers on Labour, Education and Technology 2021/08, Seville: European Commission, 2021, JRC124896. Mismatch unemployment in Austria The role of regional labour markets for skills René Böheim (Johannes Kepler University Linz) and Michael Christl (European Commission, Joint Research Centre) Abstract During the last decade, the Austrian labour market experienced a substantial outward shift of the Beveridge curve - the relationship between the unemployment and the job vacancy rate. Using detailed administrative data on vacancies and registered unemployed by region and skill level, we test which factors caused this shift. We find that the Beveridge curve shifted primarily because mismatch increased substantially. Looking on the regional and skill dimension of mismatch unemployment, we find a substantial increase of mismatch unemployment for manual routine tasks as well as for the region of Vienna. Keywords: Beveridge curve, unemployment, matching efficiency Authors: René Böheim (Johannes Kepler University Linz; IZA Bonn; CESifo Munich; WIFO Austria, Vienna; and GLO.) and Michael Christl (European Commission, Joint Research Centre) Acknowledgments: We are thankful for helpful comments of Alexander Ahammer and Sergio Torrejon Perez. Additionally, we are indebted to Veronika Murauer (AMS Austria) for her patient support with the preparation of the data. Joint Research Centre reference number: JRC124896 Contents 1 Introduction.................................................................................................................................................................................... 1 2 Theoretical Background........................................................................................................................................................... 2 3 Data and Calibration................................................................................................................................................................. 3 4 Results............................................................................................................................................................................................... 7 4.1 Results based on aggregated data for Austria................................................................................................. 7 4.2 Results by federal state ................................................................................................................................................. 9 4.3 Results by skill level......................................................................................................................................................... 10 4.4 Results by skill level and federal state.................................................................................................................. 11 5 Conclusion....................................................................................................................................................................................... 12 References .............................................................................................................................................................................................. 14 List of figures........................................................................................................................................................................................ 16 List of tables.......................................................................................................................................................................................... 17 Annex......................................................................................................................................................................................................... 17 Mismatch unemployment in Austria 1 Introduction The Austrian unemployment rate increased from about 4 percent at the beginning of 2000 to 5.6 percent in 2005 and, after the Great Recession, it increased to about 6 percent by 2015. The increasing unemployment rate and a substantial increase in the vacancy rate led to a marked outward shift of the Beveridge curve. Schiman (2018) argues in a macro-model framework that the Austrian Beveridge curve shifted due to a labour supply shock caused by the opening of the labour market to several Eastern European countries after 2008. However, Christl et al. (2016) and later Christl (2020) using detailed data labour-market transition argue that the shift was caused primarily by an increase in labour market mismatch. Following Veracierto (2011) and Şahin et al. (2014), we test whether or not the outward shift of the Beveridge curve in Austria was caused by mismatch unemployment. Mismatch unemployment is defined as the unemployment that can be attributed to changes in the matching efficiency observed on the labour market. We first analyze aggregate data on the national level. We use unemployment data from the Austrian unemployment office (AMS) by skill level and labour market district level. We combine these data with information from the Austrian Mikrozensus, which includes detailed information on employment by skill levels and by regions. We subsequently provide analyses at different levels of disaggregation, by regions and skill levels, to provide more detailed evidence for the shift of the Beveridge curve. As shown by Autor et al. (2006), Goos and Manning (2007), Goos et al. (2010), and Autor and Dorn (2013), the employment share of occupations in the middle of the skill distribution declined rapidly in the US and Europe while at the same time the upper and lower skill occupation share has increased substantially, however, this general phenomena can differ across countries due to different institutional settings, socio-demographic dynamics or migrations (see, e.g., Oesch and Rodríguez Menés (2011)). The literature on automation stresses that these jobs often consist of routine tasks that are relatively easy to automatize and which thus are disappearing due to reduced demand (Autor et al., 2003; Michaels et al., 2014). However, the literature on the impact of overall employment effects of automation suggests a rather small impact. Acemoglu et al. (2021), for example, show for the US that while artificial intelligence replaces human workers at different types of tasks, there is currently no aggregate effect on the labour market. When looking on the impact of robots on overall employment, the literature also suggests changes in the task content of jobs rather than a strong reduction of employment1. These changes in labour demand lead to substantial challenges in most developed countries. While the demand for certain skills may change quickly, supply side reactions are typically slow as the adjustment of workers requires more time for re-skilling or re-training. Such developments may lead to substantial mismatch and stress the importance of identifying reasons for labour market mismatch and appropriate policy responses. Our results show that the outward shift of the Austrian Beveridge curve was primarily caused by a substantial increase of mismatch unemployment for manual routine tasks. We find that mismatch unemployment for manual routine tasks increased from about 2 percent to almost 8 percent between 2013 and 2016. This implies that under constant matching of workers and vacancies on the labour market, the mismatch unemployment rate for manual routine tasks, and therefore also the unemployment rate for manual-routine tasks would be 6pp lower. Mismatch unemployment for interactive non-routine tasks also increased, from about 1 to 3 percent. In contrast, we find that mismatch unemployment increased only moderately for other skill groups. Our analysis also highlights regional differences in the increase of mismatch unemployment. We find that Vienna has the greatest overall increase in mismatch unemployment from about 1 percent in 2013 to more than 3 percent in 2016. Overall, however, the results do not suggest that insufficient regional mobility, due to e.g., house ownership (Farber, 2012), is the reason for increased mismatch. 1See, e.g., Klenert et al. (2020), Dauth et al. (2017) or Barbieri et al. (2019). 1 Mismatch unemployment in Austria 2 Theoretical Background We use the model of Veracierto (2011) where each firm offers jobs. Jobs remain vacant or become filled by a worker’s acceptance of the offer. Workers are either employed, unemployed or inactive. Employed workers separate from their jobs with a probability λEU t. For simplicity, our notation does not distinguish between skills and regions, which are additional dimensions we consider below. The matching between unemployed workers and vacant jobs is modelled with a standard matching function, where the number of new matches Mtis a function of the matching efficiency (At), the number of unemployed workers (Ut), and vacant jobs (Vt): Mt=AtUα tV(1−α) t,(1) where α,0< α < 1, imposes constant returns to scale (Petrongolo and Pissarides, 2001). Workers move between three states, unemployment (U), employment (E), and inactivity (I). Hazard rates, λIJ t, describe the transitions from labour market status Ito labour market status Jat time t. In other words, λtis the share of workers who move from Ito Jat time t,NIJ t, over the number of workers who were in Iat time t−1,NI t−1. E.g., λIJ =NIJ t/NI t−1. The movement of workers across labour market states is described by the following set of equations: Ut+1 =Ut+λEU t∗Et+λIU t∗It−λUI t∗Uα t−AtUα tV(1−α) t,(2) Et+1 =Et+AtUα tV(1−α) t+λIE t∗It−(λEU t+λEI t)∗Et,(3) It+1 =It+λEI t∗Et+λUI t∗Ut−λUI t∗Ut−(λIE t+λIU t)∗It.(4) The steady state unemployment is given by: uss t=st st+ft ,(5) where the separation rate is st=λEU t+ (λEI t∗λIU t)/(1 −λII t)and the job finding rate is ft= λUE t+ (λUI t∗λIE t)/(1 −λII t). We then define mismatch unemployment umm tas the difference between the steady state unemployment rate, uss t, and the counterfactual unemployment rate, u∗ t, that would have been the outcome of stable matching function: umm t=uss t−u∗ t=st st+λUE t+λUIE t − st st+λ∗UE t+λUIE t .(6) where λUIE t=λU I t∗λIE t 1−λII t . In order to calibrate the model, we calculate the parameter αof the matching function. We follow Barlevy (2011) and Veracierto (2011) and assume constant transition rates in the period before the Beveridge curve shift2. We assume a constant matching productivity Aover the observed period before the shift. Choosing the month with the strongest and the month with the weakest labour market tightness, separately by region and skill level, allows us to calculate the αparameter (Veracierto, 2011). We set Ato the average labour market tightness of the month with the strongest and the month with the weakest labour market tightness, separately for each combination of region and skill level. Following this approach, we calculate the αand use these estimates to calculate the matching efficiency parameter At3. We obtain hypothetical vacancy rates for the period after 2014, when we observe the shift in the Beveridge curve, by setting Atfor this period to the average level of the period before 2014, conditional on the observed unemployment rate (Veracierto, 2011). 2As shown in Figure 3, this assumption seems to be reasonable also for the data we are using. 3For a general discussion on estimating matching efficiencies, see e.g., Crawley et al. (2021). 2 Mismatch unemployment in Austria We calculate these parameters for all degrees of disaggregation (region, skill level, and their interaction). Following Barnichon and Figura (2010), we identify the source of Beveridge curve shifts. Shifts can be caused by several factors: supply-side factors, demand-side factors or a change in the efficiency of matches on the labour market. The shift of the Austrian Beveridge curve at the national level and the associated increase in unemployment after 2014 stems mainly from a change in matching efficiency, while other factors play only a minor role (Christl, 2020). We focus in particular on the changes at disaggregated levels to obtain more detailed information about the roots of the increasing labour market mismatch. 3 Data and Calibration We use data from the Austrian public employment services (PES) from 2004 to 2016, which provide detailed information on the skill levels of the unemployed and the required tasks of posted vacancies (AMS Österreich, 2020). Following Spitz-Oener (2006), we group 119 specific occupations (ISCO-08) into five categories, manual routine tasks, manual non-routine tasks, analytical non-routine tasks, interactive non-routine tasks, and cognitive routine tasks. The detailed list of how occupations are classified is given in Table 6. The data are quarterly data from 2004:Q1 until 2016:Q4 for five skill categories, aggregated to the nine federal states. We use the Austrian Labour Force Survey (LFS, ‘Arbeitskräfteerhebung’) to estimate the job finding rate and employment levels by federal state and skill level4. The LFS uses the same occupational classification (ISCO-08, at three-digit level) as the Austrian PES. Before 2011, the ISCO-88 classification was used and we convert both classifications to five skill categories5, following Bock-Schappelwein et al. (2017). The Austrian LFS has a rotating panel structure which allows us to follow workers for five consecutive quarters. This allows us to estimate job finding rates by skill category and by region. Table 1 shows the distribution of unemployment and employment across federal states in Austria. About 19.2 percent of employed persons and 31.9 (36.0) percent of unemployed persons were in Vienna. In the table, we report the region’s share of unemployed persons based on both the number of registered unemployed observed by the PES and the number of the unemployed observed in the LFS which uses the ILO’s definition of unemployment. In general, the unemployment shares are fairly similar in both sources, although the unemployment rates typically differ substantially due to the different definition of unemployment. 4See, e.g., Statistik Austria (2020) and Moser (2010). 5Table 6 in the Appendix shows the exact categories used for each skill group. 3 Mismatch unemployment in Austria Figure 7: Mismatch unemployment, by region. 0 .01 .02 .03 .04 mismatch unemployment (rate) 2004q1 2006q1 2008q1 2010q1 2012q1 2014q1 2016q1 time Burgenland Lower Austria Vienna Carinthia Styria Upper Austria Salzburg Tyrol Vorarlberg Source: Own calculation based on data from Statistik Austria (2020) and AMS Österreich (2020). Notes: Mismatch unemployment is defined as the difference between the unemployment rate under a stable matching productivity and the steady-state unemployment rate. 4.3 Results by skill level Labour markets differ in their supply of and in their demand for different skills. We see substantial regional differences in the unemployment rates and vacancy rates by skill category. Job finding rates may also differ substantially. Figure 15 highlights the development of the mismatch indicator over time by skill category. We see that the mismatch increased in particular for manual routine tasks and to some extent also for cognitive routine tasks and analytical non-routine tasks after 2014. These differences correspond to shifts of the estimated Beveridge curves, where the shift is especially pronounced for manual routine tasks. Different skill categories have evolved differently over the recent years. In particular, we observe a substantial increase in the unemployment rate and a stable vacancy rate of manual routine tasks, where unemployment is typically higher than for other skill types. This suggests increased labour market polarization which is caused by increased skill-mismatch for manual routine tasks. In contrast, we find stable unemployment rates, and a substantial increase of the vacancy rate, for cognitive routine tasks. We interpret this as evidence for a shortage of this specific skill type where few workers are available to fill vacancies. We plot the resulting mismatch unemployment rates in Figure 8. Although mismatch unemployment for manual routine tasks was greater than for other skill categories before 2011, it increased substantially after 2014, from about 2 percent to almost 8 percent in 2016. While we observe an increase of mismatch unemployment after 2014 also for other skill categories, the increase for manual routine tasks is much more pronounced. It appears that the increase in mismatch unemployment for interactive non-routine tasks started already by 2010, after which it continually increased. 10 Mismatch unemployment in Austria Figure 8: Mismatch unemployment, by skill level. 0 .02 .04 .06 .08 mismatch unemployment 2004q1 2006q1 2008q1 2010q1 2012q1 2014q1 2016q1 time Analytical non-routine tasks Interactive non-routine tasks Cognitive routine tasks Manual routine tasks Manual non-routine tasks Source: Own calculation based on data from Statistik Austria (2020) and AMS Österreich (2020). Notes: Mismatch unemployment is defined as the difference between the unemployment rate under a stable matching productivity and the steady-state unemployment rate. 4.4 Results by skill level and federal state If we assume that each skill type has a distinct labour market in each region, we may repeat the analysis for the resulting 45 different labour markets. The interpretation of the results requires caution as, at least for neighboring regions or similar skill types, some markets are clearly connected. In addition, some of these labour markets are small, which leads to substantial uncertainty because of the sample size of the Labour Force Survey (LFS). In Figure 18 we plot the Beveridge curves for analytical non-routine tasks for each of the nine regions. We do not find shifts of these Beveridge curves, with the exception of Upper Austria and Salzburg. We conclude from this evidence that the mismatch for analytical non-routine task is a minor problem in the Austrian labour market. In contrast, the Beveridge curves for interactive non-routine tasks, plotted in Figure 19, exhibit considerable shifts in all federal states. It is striking that, with the exception of Vienna and Carinthia, the shifts are mainly caused by an increase in the vacancy rates. This suggests increased demand for interactive non-routine tasks, especially in Upper Austria, Salzburg and Vorarlberg. The Beveridge curves for cognitive routine tasks, Figure 20, reveal shifts only in Styria, Upper Austria, and Salzburg. The shifts appear to be driven more by supply side factors as unemployment rates are relatively more stable than vacancy rates. The Beveridge curves for manual routine tasks, Figure 21, shift outwards in almost all regions, with the exception of Upper Austria. These shifts, in contrast to the shifts for cognitive routine tasks, are caused by an increase in the unemployment rates rather than by greater vacancy rates. This suggests that the demand for manual routine tasks has been declining over time, with the implication that it will be difficult for unemployed workers with manual routine skills to find employment. The Beveridge curves for manual non-routine tasks, plotted in Figure 22, are fairly stable and there are only minor outward shifts in few regions. In contrast to manual routine tasks, we do not find substantial changes in the matching efficiency for manual non-routine tasks. We plot the estimated mismatch unemployment by region and skill-type in Figure 9. The plots 11 Mismatch unemployment in Austria reveal substantial differences by skill level and region. In particular, mismatch unemployment increased in all regions, with the expception of Vorarlberg. Mismatch unemployment increased most noticeably in Vienna, where we estimate an increase for manual routine tasks and interactive non-routine tasks. While the increase in mismatch unemployment is most pronounced in Vienna, we estimate increased mismatch unemployment for analytical non-routine tasks also in the other regions, however, at more moderate levels. Figure 9: Mismatch unemployment, by region and skill level. 0 .05 .1 .15 .2 mismatch unemployment 2004q1 2007q3 2011q1 2014q3 2018q1 time Burgenland 0 .05 .1 .15 .2 mismatch unemployment 2004q1 2007q3 2011q1 2014q3 2018q1 time Lower Austria 0 .05 .1 .15 .2 mismatch unemployment 2004q1 2007q3 2011q1 2014q3 2018q1 time Vienna 0 .05 .1 .15 .2 mismatch unemployment 2004q1 2007q3 2011q1 2014q3 2018q1 time Carinthia 0 .05 .1 .15 .2 mismatch unemployment 2004q1 2007q3 2011q1 2014q3 2018q1 time Styria 0 .05 .1 .15 .2 mismatch unemployment 2004q1 2007q3 2011q1 2014q3 2018q1 time Upper Austria 0 .05 .1 .15 .2 mismatch unemployment 2004q1 2007q3 2011q1 2014q3 2018q1 time Salzburg 0 .05 .1 .15 .2 mismatch unemployment 2004q1 2007q3 2011q1 2014q3 2018q1 time Tyrol 0 .05 .1 .15 .2 mismatch unemployment 2004q1 2007q3 2011q1 2014q3 2018q1 time Vorarlberg Analytical non-routine tasks Interactive non-routine tasks Cognitive routine tasks Manual routine tasks Manual non-routine tasks Source: Own calculation based on data from Statistik Austria (2020) and AMS Österreich (2020). Notes: Mismatch unemployment is defined as the difference between the unemployment rate under a stable matching productivity and the steady-state unemployment rate. The general increase in mismatch unemployment for interactive non-routine tasks we have seen before seems to be especially driven by the development in Vienna, where the increase is especially strong with almost 10 percent mismatch unemployment in 2016. For the rest of the skill levels, we do not see a strong increase in mismatch unemployment, even though there are smaller upward movements visible in manual non-routine tasks in Salzburg, Upper Austria, Lower Austria, and Tyrol at the end of our observation period. 5 Conclusion We analyze the Austrain Beveridge curve shift that happened after 2014. We use detailed vacancy data, on both skill and regional level, from the Public Employment Office (AMS) and estimate labour market flows on disaggregate level using information from the Austrian LFS. Using these data, we disaggregate the labour market into several regional skill labour markets. Following the approach of Veracierto (2011), who uses a simplified version of the Mortensen and Pissarides (1994) model, we 12 Mismatch unemployment in Austria estimate Beveridge curves for Austria and all corresponding disaggregated labour markets. Additionally, we calculate the mismatch unemployment corresponding to each of the disaggreagted levels. Our approach does not allow us to identify all potential causes of mismatch separately. However, following Şahin et al. (2014) we argue that analyzing different levels of disaggregation is informative, especially from a policy perspective. First, we find a substantially increase in mismatch unemployment in Austria after 2014 from about 0.5 percent up to more than 2 percent. Second, we find an increase in most of the Austrian regions after 2014; the increase is especially strong in the region of Vienna, where mismatch unemployment rose from about 1 percent to more than 3 percent. Third, when we consider mismatch unemployment of different skill segments, we find an especially strong increase in mismatch unemployment for manual routine tasks. Mismatch unemployment increase from levels between 1 and 2 percent before 2014 to almost 8 percent after 2014. While the reasons for the shift of the Beveridge curve have been debated substantially in the literature, our analysis confirms that a decrease in matching efficiency after 2014 led to a shift in the Beveridge curve. While so far the reasons for this shift only have been analyzed partially by Christl (2020), our analysis identifies detailed mismatch unemployment on regional and skill level. This is especially important from a policy point of view, since policies to tackle the mismatch problems on the labour market can be targeted especially on the identified labour markets. 13 Mismatch unemployment in Austria References Acemoglu, D., Autor, D., Hazell, J. and Restrepo, P., ‘AI and jobs: Evidence from online vacancies’, NBER working paper 28257, 2021. AMS Österreich, ‘Arbeitsmarktdaten online’. 2020. Data at http://iambweb.ams.or.at/ambweb/. Autor, D. H. and Dorn, D., ‘The growth of low-skill service jobs and the polarization of the US labor market’, American Economic Review, Vol. 103, No 5, 2013, pp. 1553–97. Autor, D. H., Katz, L. F. and Kearney, M. S., ‘The polarization of the US labor market’, American Economic Review, Vol. 96, No 2, 2006, pp. 189–194. Autor, D. H., Levy, F. and Murnane, R. J., ‘The skill content of recent technological change: An empirical exploration’, Quarterly Journal of Economics, Vol. 118, No 4, 2003, pp. 1279–1333. Barbieri, L., Mussida, C., Piva, M. and Vivarelli, M., ‘Testing the employment impact of automation, robots and ai: A survey and some methodological issues’, 2019. Barlevy, G., ‘Evaluating the role of labor market mismatch in rising unemployment’, Economic Perspectives, Vol. 35, No 3, 2011, pp. 82–96. Barnichon, R. and Figura, A., ‘What drives movements in the unemployment rate? A decomposition of the Beveridge curve’, FEDS Finance and Economics Discussion Series 2010-48, 2010. Bock-Schappelwein, J., Famira-Mühlberger, U. and Leoni, T., ‘Arbeitsmarktchancen durch Digitalisierung’, WIFO Studies 60909, 2017. Christl, M., ‘A Beveridge curve decomposition for Austria: Did the liberalisation of the Austrian labour market shift the Beveridge curve?’, Journal for Labour Market Research, Vol. 54, No 1, 2020, pp. 1–15. Christl, M., Köppl-Turyna, M. and Kucsera, D., ‘Structural unemployment after the crisis in Austria’, IZA Journal of European Labor Studies, Vol. 5, No 1, 2016, p. 12. Crawley, A., Welch, S. and Yung, J., ‘Improving estimates of job matching efficiency with different measures of unemployment’, Journal of Macroeconomics, Vol. 67, 2021, p. 103282. Dauth, W., Findeisen, S., Südekum, J. and Woessner, N., ‘German robots-the impact of industrial robots on workers’, 2017. Farber, H. S., ‘Unemployment in the Great Recession: Did the housing market crisis prevent the unemployed from moving to take jobs?’, American Economic Review, Vol. 102, No 3, 2012, pp. 520–25. Goos, M. and Manning, A., ‘Lousy and lovely jobs: The rising polarization of work in Britain’, Review of Economics and Statistics, Vol. 89, No 1, 2007, pp. 118–133. Goos, M., Manning, A. and Salomons, A., ‘Explaining job polarization in Europe: The roles of technology, globalization and institutions’, CEP Discussion Papers 1026, 2010. Klenert, D., Fernandez-Macias, E. and Antón Pérez, J. I., ‘Do robots really destroy jobs? evidence from europe’, Tech. rep., JRC Working Papers Series on Labour, Education and Technology, 2020. Michaels, G., Natraj, A. and Van Reenen, J., ‘Has ICT polarized skill demand? Evidence from eleven countries over twenty-five years’, Review of Economics and Statistics, Vol. 96, No 1, 2014, pp. 60– 77. 14 Mismatch unemployment in Austria Mortensen, D. T. and Pissarides, C. A., ‘Job creation and job destruction in the theory of unemployment’, Review of Economic Studies, Vol. 61, No 3, 1994, pp. 397–415. Moser, C., ‘Daten zur Erwerbstätigkeit in der Mikrozensus-Arbeitskräfteerhebung’, Austrian Journal of Statistics, Vol. 39, No 1&2, 2010, pp. 117–125. Oesch, D. and Rodríguez Menés, J., ‘Upgrading or polarization? occupational change in britain, germany, spain and switzerland, 1990–2008’, Socio-Economic Review, Vol. 9, No 3, 2011, pp. 503–531. Petrongolo, B. and Pissarides, C. A., ‘Looking into the black box: A survey of the matching function’, Journal of Economic Literature, Vol. 39, No 2, 2001, pp. 390–431. Schiman, S., ‘Labor supply shocks and the Beveridge curve’, WIFO Working Papers 568, 2018. Shimer, R., ‘Reassessing the ins and outs of unemployment’, Review of Economic Dynamics, Vol. 15, No 2, 2012, pp. 127–148. Spitz-Oener, A., ‘Technical change, job tasks, and rising educational demands: Looking outside the wage structure’, Journal of Labor Economics, Vol. 24, No 2, 2006, pp. 235–270. Statistik Austria, ‘Mikrozensus-Arbeitskräfteerhebung’. 2020. Data at https://www.statistik. at/web_de/services/mikrodaten_fuer_forschung_und_lehre/datenangebot/ standardisierte_datensaetze_sds/index.html#index3. Veracierto, M., ‘Worker flows and matching efficiency’, Economic Perspectives, Vol. 35, No 4, 2011, pp. 147–169. Şahin, A., Song, J., Topa, G. and Violante, G. L., ‘Mismatch unemployment’, American Economic Review, Vol. 104, No 11, November 2014, pp. 3529–64. 15 Mismatch unemployment in Austria List of Figures Figure 1. Unemployment rates and vacancy rates, by region. . . . . . . . . . . . . . . . . . . . . 5 Figure 2. Unemployment rates and vacancy rates, by skill category. . . . . . . . . . . . . . . . . 5 Figure 3. Transition rates, aggregated data for Austria, 2004–2016. . . . . . . . . . . . . . . . . 6 Figure 4. Job finding rates, by region and skill category. . . . . . . . . . . . . . . . . . . . . . . . 7 Figure 5. Mismatch Indicator and Beveridge Curves, aggregated data for Austria, 2004–2016. 8 Figure 6. Mismatch unemployment, aggregated data for Austria, 2004–2016. . . . . . . . . . . 9 Figure 7. Mismatch unemployment, by region. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 Figure 8. Mismatch unemployment, by skill level. . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 Figure 9. Mismatch unemployment, by region and skill level. . . . . . . . . . . . . . . . . . . . . 12 Figure 10. Job findings rates, by estimation method. . . . . . . . . . . . . . . . . . . . . . . . . . . 21 Figure 11. Model prediction of the unemployment rate, Austria . . . . . . . . . . . . . . . . . . . . 21 Figure 12. Mismatch indicators, by region, 2004–2016. . . . . . . . . . . . . . . . . . . . . . . . . 22 Figure 13. Beveridge curve, by region, 2004–2016. . . . . . . . . . . . . . . . . . . . . . . . . . . . 22 Figure 14. Model prediction of the unemployment rate, by regions . . . . . . . . . . . . . . . . . . 23 Figure 15. Mismatch indicator, by skill level, 2004–2016. . . . . . . . . . . . . . . . . . . . . . . . 23 Figure 16. Beveridge curves, by skill level, 2004–2016. . . . . . . . . . . . . . . . . . . . . . . . . 24 Figure 17. Model prediction of the unemployment rate, by skill level . . . . . . . . . . . . . . . . . 24 Figure 18. Beveridge curves - analytical non-routine tasks . . . . . . . . . . . . . . . . . . . . . . 25 Figure 19. Beveridge curves - interactive non-routine tasks . . . . . . . . . . . . . . . . . . . . . . 25 Figure 20. Beveridge curves - cognitive routine tasks . . . . . . . . . . . . . . . . . . . . . . . . . 26 Figure 21. Beveridge curves - manual routine tasks . . . . . . . . . . . . . . . . . . . . . . . . . . 26 Figure 22. Beveridge curves - manual non-routine tasks . . . . . . . . . . . . . . . . . . . . . . . . 27 16 Mismatch unemployment in Austria List of Tables Table 1. Employment and unemployment shares, by federal state. . . . . . . . . . . . . . . . . . 4 Table 2. Employment and unemployment, by skill category. . . . . . . . . . . . . . . . . . . . . . 4 Table 3. Summarystatisticsbyregion. ................................. 18 Table 4. Summary statistics by region (cont.) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19 Table 5. Summary statistics by skill level. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 Table 6. Classificationofoccupations .................................. 28 17 Mismatch unemployment in Austria Annex Table 3: Summary statistics by region. Variable Mean Std. Dev. Min. Max. N Burgenland unemployment rate 0.06 0.016 0.039 0.091 52 vacancy rate 0.004 0.001 0.003 0.008 52 job finding rate 0.203 0.062 0.084 0.37 51 separation rate 0.012 0.025 -0.035 0.055 51 tightness 14.758 6.491 5.279 30.332 52 Lower Austria unemployment rate 0.055 0.011 0.036 0.079 52 vacancy rate 0.005 0.002 0.003 0.009 52 job finding rate 0.2 0.056 0.094 0.363 51 separation rate 0.011 0.015 -0.021 0.031 51 tightness 11.188 4.465 3.991 21.775 52 Vienna unemployment rate 0.098 0.017 0.071 0.138 52 vacancy rate 0.006 0.002 0.003 0.01 52 job finding rate 0.183 0.043 0.113 0.301 51 separation rate 0.02 0.011 -0.001 0.041 51 tightness 17.891 6.81 7.138 34.947 52 Carinthia unemployment rate 0.074 0.019 0.042 0.11 52 vacancy rate 0.007 0.002 0.004 0.014 52 job finding rate 0.187 0.064 0.096 0.388 51 separation rate 0.015 0.027 -0.035 0.064 51 tightness 11.297 5.309 3.267 23.017 52 Styria unemployment rate 0.059 0.012 0.041 0.084 52 vacancy rate 0.006 0.001 0.004 0.008 52 job finding rate 0.194 0.054 0.107 0.405 51 separation rate 0.012 0.018 -0.023 0.038 51 tightness 10.344 3.58 5.333 19.761 52 Upper Austria unemployment rate 0.041 0.01 0.024 0.061 52 vacancy rate 0.01 0.003 0.007 0.016 52 job finding rate 0.28 0.086 0.1 0.497 51 separation rate 0.011 0.012 -0.015 0.027 51 tightness 4.257 1.657 1.593 8.306 52 Salzburg unemployment rate 0.044 0.008 0.027 0.058 52 vacancy rate 0.01 0.002 0.007 0.015 52 job finding rate 0.249 0.069 0.124 0.436 51 separation rate 0.011 0.012 -0.003 0.039 51 tightness 4.647 1.355 2.408 7.872 52 Source: Own calculations, data on registered unemployed and vacancies obtained from AMS Österreich (2020); employment, job-finiding rate and seperation rate obtained from Statistik Austria (2020). 18 Mismatch unemployment in Austria Table 4: Summary statistics by region (cont.) Variable Mean Std. Dev. Min. Max. N Tyrol unemployment rate 0.051 0.01 0.03 0.069 52 vacancy rate 0.007 0.002 0.005 0.012 52 job finding rate 0.21 0.071 0.089 0.424 51 separation rate 0.012 0.02 -0.015 0.054 51 tightness 7.308 1.894 4.081 11.629 52 Vorarlberg unemployment rate 0.049 0.005 0.04 0.063 52 vacancy rate 0.008 0.002 0.004 0.011 52 job finding rate 0.234 0.053 0.101 0.353 51 separation rate 0.012 0.007 0.001 0.033 51 tightness 6.77 2.357 3.917 13.211 52 Source: Own calculations, data on registered unemployed and vacancies obtained from AMS Österreich (2020); employment, job-finiding rate and seperation rate obtained from Statistik Austria (2020). 19 Mismatch unemployment in Austria Figure 20: Beveridge curves - cognitive routine tasks 0 .01 .02 .03 vacancy rate 0 .025 .05 .075 .1 unemployment rate Burgenland (Cognitive routine tasks) 0 .01 .02 .03 vacancy rate 0 .025 .05 .075 .1 unemployment rate Lower Austria (Cognitive routine tasks) 0 .01 .02 .03 vacancy rate 0 .025 .05 .075 .1 unemployment rate Vienna (Cognitive routine tasks) 0 .01 .02 .03 vacancy rate 0 .025 .05 .075 .1 unemployment rate Carinthia (Cognitive routine tasks) 0 .01 .02 .03 vacancy rate 0 .025 .05 .075 .1 unemployment rate Styria (Cognitive routine tasks) 0 .01 .02 .03 vacancy rate 0 .025 .05 .075 .1 unemployment rate Upper Austria (Cognitive routine tasks) 0 .01 .02 .03 vacancy rate 0 .025 .05 .075 .1 unemployment rate Salzburg (Cognitive routine tasks) 0 .01 .02 .03 vacancy rate 0 .025 .05 .075 .1 unemployment rate Tyrol (Cognitive routine tasks) 0 .01 .02 .03 vacancy rate 0 .025 .05 .075 .1 unemployment rate Vorarlberg (Cognitive routine tasks) Beveridge Curve (till 2014) Beveridge Curve (after 2014) hypothetical BC (till 2014) hypothetical BC (after 2014) Source: Own calculations, based on data from AMS Österreich (2020) and Statistik Austria (2020). Notes: The hypothetical Beveridge curves are estimated with the average matching efficiency before 2014 and after 2014. Figure 21: Beveridge curves - manual routine tasks 0 .01 .02 .03 vacancy rate 0 .05 .1 .15 .2 .25 .3 unemployment rate Burgenland (Manual routine tasks) 0 .01 .02 .03 vacancy rate 0 .05 .1 .15 .2 .25 .3 unemployment rate Lower Austria (Manual routine tasks) 0 .01 .02 .03 vacancy rate 0 .05 .1 .15 .2 .25 .3 unemployment rate Vienna (Manual routine tasks) 0 .01 .02 .03 vacancy rate 0 .05 .1 .15 .2 .25 .3 unemployment rate Carinthia (Manual routine tasks) 0 .01 .02 .03 vacancy rate 0 .05 .1 .15 .2 .25 .3 unemployment rate Styria (Manual routine tasks) 0 .01 .02 .03 vacancy rate 0 .05 .1 .15 .2 .25 .3 unemployment rate Upper Austria (Manual routine tasks) 0 .01 .02 .03 vacancy rate 0 .05 .1 .15 .2 .25 .3 unemployment rate Salzburg (Manual routine tasks) 0 .01 .02 .03 vacancy rate 0 .05 .1 .15 .2 .25 .3 unemployment rate Tyrol (Manual routine tasks) 0 .01 .02 .03 vacancy rate 0 .05 .1 .15 .2 .25 .3 unemployment rate Vorarlberg (Manual routine tasks) Beveridge Curve (till 2014) Beveridge Curve (after 2014) hypothetical BC (till 2014) hypothetical BC (after 2014) Source: Own calculations, based on data from AMS Österreich (2020) and Statistik Austria (2020). Notes: The hypothetical Beveridge curves are estimated with the average matching efficiency before 2014 and after 2014. 26 Mismatch unemployment in Austria Figure 22: Beveridge curves - manual non-routine tasks 0 .01 .02 .03 vacancy rate 0 .05 .1 .15 .2 unemployment rate Burgenland (Manual non-routine tasks) 0 .01 .02 .03 vacancy rate 0 .05 .1 .15 .2 unemployment rate Lower Austria (Manual non-routine tasks) 0 .01 .02 .03 vacancy rate 0 .05 .1 .15 .2 unemployment rate Vienna (Manual non-routine tasks) 0 .01 .02 .03 vacancy rate 0 .05 .1 .15 .2 unemployment rate Carinthia (Manual non-routine tasks) 0 .01 .02 .03 vacancy rate 0 .05 .1 .15 .2 unemployment rate Styria (Manual non-routine tasks) 0 .01 .02 .03 vacancy rate 0 .05 .1 .15 .2 unemployment rate Upper Austria (Manual non-routine tasks) 0 .01 .02 .03 vacancy rate 0 .05 .1 .15 .2 unemployment rate Salzburg (Manual non-routine tasks) 0 .01 .02 .03 vacancy rate 0 .05 .1 .15 .2 unemployment rate Tyrol (Manual non-routine tasks) 0 .01 .02 .03 vacancy rate 0 .05 .1 .15 .2 unemployment rate Vorarlberg (Manual non-routine tasks) Beveridge Curve (till 2014) Beveridge Curve (after 2014) hypothetical BC (till 2014) hypothetical BC (after 2014) Source: Own calculations, based on data from AMS Österreich (2020) and Statistik Austria (2020). Notes: The hypothetical Beveridge curves are estimated with the average matching efficiency before 2014 and after 2014. ’ 27 Mismatch unemployment in Austria Table 6: Classification of occupations ISCO-08 class task category description 111 1 manual routine tasks Legislators and senior officials 112 1 manual routine tasks Managing directors and chief executives 121 1 manual routine tasks Business services and administration managers 122 1 manual routine tasks Sales, marketing and development managers 131 1 manual routine tasks Production managers in agriculture, forestry and fisheries 132 1 manual routine tasks Manufacturing, mining, construction, and distribution managers 133 1 manual routine tasks Information and communications technology service managers 134 1 manual routine tasks Professional services managers 141 1 manual routine tasks Hotel and restaurant managers 143 1 manual routine tasks Other services managers 211 1 manual routine tasks Physical and earth science professionals 212 1 manual routine tasks Mathematicians, actuaries and statisticians 213 1 manual routine tasks Life science professionals 214 1 manual routine tasks Engineering professionals (excluding electrotechnology) 215 1 manual routine tasks Electrotechnology engineers 216 1 manual routine tasks Architects, planners, surveyors and designers 221 1 manual routine tasks Medical doctors 222 1 manual routine tasks Nursing and midwifery professionals 225 1 manual routine tasks Veterinarians 226 1 manual routine tasks Other health professionals 231 1 manual routine tasks University and higher education teachers 232 2 interactive non-routine tasks Vocational education teachers 233 2 interactive non-routine tasks Secondary education teachers 234 2 interactive non-routine tasks Primary school and early childhood teachers 235 2 interactive non-routine tasks Other teaching professionals 241 1 manual routine tasks Finance professionals 242 1 manual routine tasks Administration professionals 243 1 manual routine tasks Sales, marketing and public relations professionals 251 1 manual routine tasks Software and applications developers and analysts 252 1 manual routine tasks Database and network professionals 261 1 manual routine tasks Legal professionals 262 1 manual routine tasks Librarians, archivists and curators 263 1 manual routine tasks Social and religious professionals 264 1 manual routine tasks Authors, journalists and linguists 265 1 manual routine tasks Creative and performing artists 311 3 cognitive routine tasks Physical and engineering science technicians 312 1 manual routine tasks Mining, manufacturing and construction supervisors 313 3 cognitive routine tasks Process control technicians 314 3 cognitive routine tasks Life science technicians and related associate professionals 315 5 manual non-routine tasks Ship and aircraft controllers and technicians 321 3 cognitive routine tasks Medical and pharmaceutical technicians 322 3 cognitive routine tasks Nursing and midwifery associate professionals 325 3 cognitive routine tasks Other health associate professionals 331 3 cognitive routine tasks Financial and mathematical associate professionals 332 2 interactive non-routine tasks Sales and purchasing agents and brokers 333 3 cognitive routine tasks Business services agents ... continued on next page. 28 Mismatch unemployment in Austria Table 6 -- continued from previous page. ISCO-08 class task category description 334 3 cognitive routine tasks Administrative and specialized secretaries 335 3 cognitive routine tasks Regulatory government associate professionals 341 2 interactive non-routine tasks Legal, social and religious associate professionals 342 2 interactive non-routine tasks Sports and fitness workers 343 2 interactive non-routine tasks Artistic, cultural and culinary associate professionals 351 3 cognitive routine tasks Information and communications technology operations and user support technicians 352 3 cognitive routine tasks Telecommunications and broadcasting technicians 411 3 cognitive routine tasks General office clerks 412 3 cognitive routine tasks Secretaries (general) 413 3 cognitive routine tasks Keyboard operators 421 2 interactive non-routine tasks Tellers, money collectors and related clerks 422 2 interactive non-routine tasks Client information workers 431 3 cognitive routine tasks Numerical clerks 432 3 cognitive routine tasks Material-recording and transport clerks 441 3 cognitive routine tasks Other clerical support workers 511 5 manual non-routine tasks Travel attendants, conductors and guides 512 5 manual non-routine tasks Cooks 513 5 manual non-routine tasks Waiters and bartenders 514 5 manual non-routine tasks Hairdressers, beauticians and related workers 515 5 manual non-routine tasks Building and housekeeping supervisors 516 5 manual non-routine tasks Other personal services workers 521 2 interactive non-routine tasks Street and market salespersons 522 2 interactive non-routine tasks Shop salespersons 523 2 interactive non-routine tasks Cashiers and ticket clerks 524 2 interactive non-routine tasks Other sales workers 531 2 interactive non-routine tasks Child care workers and teachers' aides 532 5 manual non-routine tasks Personal care workers in health services 541 5 manual non-routine tasks Protective services workers 611 5 manual non-routine tasks Market gardeners and crop growers 612 5 manual non-routine tasks Animal producers 613 5 manual non-routine tasks Mixed crop and animal producers 621 5 manual non-routine tasks Forestry and related workers 622 5 manual non-routine tasks Fishery workers, hunters and trappers 711 5 manual non-routine tasks Building frame and related trades workers 712 5 manual non-routine tasks Building finishers and related trades workers 713 5 manual non-routine tasks Painters, building structure cleaners and related trades workers 721 5 manual non-routine tasks Sheet and structural metal workers, molders and welders, and related workers 722 5 manual non-routine tasks Blacksmiths, toolmakers and related trades workers 723 5 manual non-routine tasks Machinery mechanics and repairers 731 5 manual non-routine tasks Handicraft workers 732 5 manual non-routine tasks Printing trades workers 741 5 manual non-routine tasks Electrical equipment installers and repairers 742 5 manual non-routine tasks Electronics and telecommunications installers and repairers 751 5 manual non-routine tasks Food processing and related trades workers 752 5 manual non-routine tasks Wood treaters, cabinet-makers and related trades workers ... continued on next page. 29 Mismatch unemployment in Austria Table 6 -- continued from previous page. ISCO-08 class task category description 753 5 manual non-routine tasks Garment and related trades workers 754 4 analytical non-routine tasks Other craft and related workers 811 4 analytical non-routine tasks Mining and mineral processing plant operators 812 4 analytical non-routine tasks Metal processing and finishing plant operators 813 4 analytical non-routine tasks Chemical and photographic products plant and machine operators 814 4 analytical non-routine tasks Rubber, plastic and paper products machine operators 815 4 analytical non-routine tasks Textile, fur and leather products machine operators 816 4 analytical non-routine tasks Food and related products machine operators 817 4 analytical non-routine tasks Wood processing and papermaking plant operators 818 4 analytical non-routine tasks Other stationary plant and machine operators 821 4 analytical non-routine tasks Assemblers 831 5 manual non-routine tasks Locomotive engine drivers and related workers 832 5 manual non-routine tasks Car, van and motorcycle drivers 833 5 manual non-routine tasks Heavy truck and bus drivers 834 4 analytical non-routine tasks Mobile plant operators 835 4 analytical non-routine tasks Ships' deck crews and related workers 911 4 analytical non-routine tasks Domestic, hotel and office cleaners and helpers 912 4 analytical non-routine tasks Vehicle, window, laundry and other hand cleaning workers 921 4 analytical non-routine tasks Agricultural, forestry and fishery labourers 931 4 analytical non-routine tasks Mining and construction labourers 932 4 analytical non-routine tasks Manufacturing labourers 933 4 analytical non-routine tasks Transport and storage labourers 941 4 analytical non-routine tasks Food preparation assistants 951 4 analytical non-routine tasks Street and related service workers 961 4 analytical non-routine tasks Street vendors (excluding food) 962 4 analytical non-routine tasks Other elementary workers 30 GETTING IN TOUCH WITH THE EU In person All over the European Union there are hundreds of Europe Direct information centres. You can find the address of the centre nearest you at: https://europa.eu/european-union/contact_en On the phone or by email Europe Direct is a service that answers your questions about the European Union. You can contact this service: - by freephone: 00 800 6 7 8 9 10 11 (certain operators may charge for these calls), - at the following standard number: +32 22999696, or - by electronic mail via: https://europa.eu/european-union/contact_en FINDING INFORMATION ABOUT THE EU Online Information about the European Union in all the official languages of the EU is available on the Europa website at: https://europa.eu/european-union/index_en EU publications You can download or order free and priced EU publications from EU Bookshop at: https://publications.europa.eu/en/publications. Multiple copies of free publications may be obtained by contacting Europe Direct or your local information centre (see https://europa.eu/european-union/contact_en).