1. Introduction [1]
One of the important determinants for future labour market trends is technological change. Since the beginning of industrialization two questions have always been raised. The first one is whether and to which extent technological change may actually destroy or create jobs. The second question refers to impacts of technological change on the composition of employment, e.g. regarding certain industries, occupations and forms of employment. Economic literature pointed out that technological change is important in many respects, for e.g. economic growth, structural change and productivity (Solow 1956; Romer 1990; Grossman and Helpman 1991).
In the foreseeable future, digitalization is one of the main drivers of technological change. Digital technologies affect the computerisation of production, service delivery and even the private sphere. We can observe an increased speed of technological innovation in the area of digital technology, which is no longer confined to routine manufacturing tasks but may spread to numerous non-routine tasks (Brynjolfsson and Mc Afee 2011). Further advances in areas such as Machine Learning, Artificial Intelligence, Mobile Robotics and the increasing usability of Big Data are still on the way (Frey and Osborne 2013).
Although developments in digital technologies have already gained momentum, the main impacts of this new era of technological change remain to a large degree uncertain and still ahead of us. In order to deal with this issue, the German case is used as an example. For several reasons Germany is an interesting case with respect to potential impacts of digitalization. The country still disposes of a rather large share of employment in manufacturing. Due to the expectation that digitalization may cause comprehensive impacts in production in the first place, potential implications in this respect for the economy as well as for the labour market are of certain interest.
The main purpose of the study is not to generate additional evidence on jobs, which may particularly be at risk through digitalization. Such investigations have already been undertaken for Germany (Bonin, et al., 2015; Brzeski and Burk 2015; Dengler and Matthes 2015; Lorenz et al., 2015). The study will neither carry out new long-term scenarios concerning the implications of the digital revolution on employment. E.g., Wolter et al. (2015), Vogler-Ludwig et al. (2016) as well as Zika et al. (2019) did scenarios focussing on impacts of digitalization within manufacturing (Industry 4.0). This paper will go beyond these studies and focus in addition on one of the important structural labour market issues of digitalization. The study will ask how far previous shifts towards non-standard work and low-wage employment in Germany will probably be reversed or accelerated through the spread of digital technologies.
The paper consists of four parts. The following and second section will present relevant evidence and consider implications of proceeding digitalization for the German labour market. The third and main section assesses the relevance of digitalization for the composition of employment forms and wage dispersion. The question will be how far digitalization can be seen as a new driver for changes in these two areas. In the final section of the study, the main findings will be summarized and used as a basis to deduce rather general and tentative implications for current labour market policies.
2. Digitalization and employment
The digital revolution started with the invention of the microprocessor and its steadily increasing performance. Enforced by the Internet, connectivity will obviously reach completely new dimensions. We are in a process in which the “real” world will continuously be linked to the “virtual” world. Electronic devices and microprocessors connect people with each other, machines with workers and machines with machines. Due to connectivity, even distance seems not to be a relevant problem anymore. Digitalization facilitates a more knowledge-based and decentralized production. It also offers the opportunity to combine production with more intelligent services (“smart services”). Up to now the number of robots is rather small but “smart” automation is spreading fast. However, for economic, legal and social reasons not everything that can potentially be automated will actually be automated.
Digitalization will – like any other type of technological change – drive the progress of the technical equipment of an economy. It will increase productivity in general and labour productivity in particular. This implies that either a given output can be produced with less input or with a given level of input more output can be achieved. Digital technologies also trigger product as well as process innovations. They push new products, which will then enter the market. At the same time, increases in productivity will facilitate lower prices. Both strengthen the competitiveness of the innovator and weaken the market position of competitors.
Such changes typically generate a process of creative destruction (Schumpeter 1912). This causes a permanent incentive for economic agents to be innovative. Therefore, according to standard growth theory technological change is an important source of productivity increase as well as economic growth (Solow 1956). Endogenous growth theory even postulates that it is technological change, which creates long-term economic growth (Romer 1990; Grossman and Helpman 1991). Simulation models indicate that digitalization is already an important driving force for economic growth. It has been estimated that between 1998 and 2012 on average 0.6 percentage points of German average growth per year were due to new digital technologies (Bornemann 2015).
As already mentioned, digitalization is only one major long-term trend which is relevant for the labour market. One has to bear in mind that trends may interact, i.e. one can assume certain interferences. Long-term trends can have rather contrary or even cumulative impacts on certain outcome variables. The combination of digitalization and globalisation may further increase competition. Fewer restrictions on trade and the availability of information at any point and at any time force national economies to specialise even more (Petersen 2015). There might also be an interaction between digitalization and demography. It can be argued that, due to demographic change, potential shortages of labour may be compensated by technological progress (e.g. labour-saving machines).
A starting point for considering potential impacts of digitalization on employment in Germany is the famous study by Frey and Osborne (2013). They expect two waves of automation. The first wave will affect high-risk occupations (probability of automation: more than 70%). According to their estimates, about one half (47%) of total US employment belongs to this category.[2] Their model predicts that most workers in transportation and logistics occupations, office and administrative support workers and labour in production occupations are at risk. In addition, the results further indicate that the lower wages and skills for a given occupation are, the higher is the automation probability and vice versa. This implies that the authors do not expect that the previous trend towards polarization of employment in the U.S. will necessarily continue in the future. The first wave of transformation will last at least for one or two decades. During the second wave the speed of change will slowdown. One of the main reasons mentioned by Frey and Osborne (2013) are engineering bottlenecks to computerisation. In addition, they argue that human labour may still have an advantage in tasks requiring more complex perceptions.
The approach by Frey and Osborne (2013) has already been transferred to Germany (Bonin et al., 2015). The analysis shows that differences between the two risk distributions are rather small. The findings suggest that in Germany fewer workers are employed in occupations with a rather high probability of automation (more than 90%). Instead more workers perform occupations with a comparatively low risk. Using the categorization of risks by Frey and Osborne (2013), the level of high-risk occupations (risk of more 70%) in Germany lies at 42%, which is a little bit lower than in the U.S. (47%). However, other studies not related to Frey and Osborne (2013) found a larger part of jobs that can potentially be substituted by intelligent machines and computers. Brzeski and Burk (2015) estimate that even 59% of German employment seems to be at risk.
Studies like the one by Frey and Osborne (2013) have certain limitations. According to Autor (2014) experts often tend to overestimate the usability and the relevance of new technologies. In particular, comparative advantages of human beings regarding flexibility, discernment and common sense are often neglected or underestimated. The extent to which jobs will disappear due to digitalization does not only depend on the occupation as such but at least as much on the task composition within certain jobs. One has to take into account that tasks within certain jobs may shift from those that can be computerised to those that can be less automated by digital technologies (Autor 2013). This would suggest that one of the main impacts of computerisation – like in previous waves of technological change – is that jobs will be reshaped rather than disappear.
A first question is how far digital technologies may influence job profiles and requirements. There are several research findings which throw light on this issue. For example, Dengler and Matthes (2018) calculated potentials for substitution of digital technologies for different categories: occupations by level of skill requirement, segments of occupations and total employment covered by social security. The results show that all levels of skill requirements are affected by potential substitution. However, the relative risk of low and medium skill requirements are highest to be replaced by digital technologies. This means that not only the level of skills is causing risks of substitution but also how far occupations are dominated by the performance of routine and repetitive tasks as previous studies have shown (Autor et al., 2003). If certain segments of occupations are analysed, the substitution potential in production is by far the highest. In other segments, such as different kinds of service occupations, the potential of substitution is lower but still rather significant. All in all, in 2016 about one quarter of employment relationships covered by social security were performing a job in which 70% of the involved tasks may have the potential to be substituted by computers in future.
Studies focussing only on probabilities of computerisation and automation regarding certain occupations do have another serious shortcoming. They only concentrate on possible job losses and neglect other impacts of technological change such as a reshaping of existing jobs or even job creation. Macroeconomic scenarios take into account such compensation impacts. All together, the studies for Germany do not identify significant job losses (Wolter et al., 2015; Vogler-Ludwig et al., 2016; Zika et al., 2019). However, the results indicate increased turnover and considerable additional shifts between different segments of the labour market such as industries, skills and occupations. Occupations in areas such as information technology or teaching relying on creative skills are more likely to become more important, whereas jobs in manufacturing (e.g. machine- and facility-controlling and maintenance of machinery) or service administration implying a high level of non-manual routine tasks are more likely to shrink. In general, the already ongoing structural change towards a service and knowledge-based economy will be accelerated by digital technologies.
The emergence of digitalization may also have an influence on work-related issues. Employers and employees may be confronted with new requirements for flexibility from their counterpart. This is because it is getting ever less important when and where people work. For workers digital technologies may offer additional options to arrange family work and employment. Also for employers, the issue of working hours gets more attention. The main reason is that the Internet and its various applications remove spatial and systematic barriers of work. A more rapid structural change may in addition have impacts on wage dispersion.
3. Employment forms and wage dispersion
Work arrangements are changing in many ways. Changes refer to their overall composition as well as to the forming of certain types of employment. Regarding the structure of employment forms, we observe in many countries a tendency towards greater diversity. This development concerns the growth of so-called “nonstandard work arrangements” with partly low levels of protection on the one hand and increasing wage disparities on the other hand. Changes with regard to the forming of specific types of employment generate a greater variation of job characteristics. Examples are new types of mobile work as well as crowd employment (Eurofound 2015).
The literature distinguishes, in general, two types of work arrangements: standard and non-standard work (Ostermann 2000; Houseman and Osawa 2003; Mückenberger 1985; Mückenberger 2010). Standard work is usually considered as working full-time in a permanent job. It offers full access to the social security system and implies a clear assignment to a certain employer. Consequently, deviations from standard employment – such as fixed-term contracts or part-time employment – are classified as non-standard. Empirical findings show that compared to standard workers there is on average a greater risk for non-standard workers to lose employment, to be part of the low-wage sector, of being excluded from (also fringe) benefits as well as from firm-based training (Autor 2001; Kalleberg 2011; Jahn and Pozzoli 2011).
However, the classification has certain limits. Some examples may illustrate this. In certain periods of the life course such as education or family work, part-time employment might be of particular interest for individuals because this work arrangement may offer an opportunity to reconcile different activities in an appropriate manner. Furthermore, fixed-term contracts or agency work can facilitate entry to the labour market and can serve as a steppingstone towards more stable employment. Therefore, non-standard employment should not necessarily be considered as precarious work. Of particular importance in this respect is how far individual workers are in an ongoing precarious biography and less whether a certain work arrangement can be considered as potentially precarious.
The following Figure 1 shows the prevalence of standard and non-standard (atypical) employment in the European Union relative to the total population aged 15 to 64 years (Rhein and Walwei 2018). For the EU as a whole the share of standard work has not changed very much in the period between 2002 and 2016 and is still close to 40%. In addition, the data indicate an overall increase in non-standard (dependent) employment by more than three percentage points in the EU since 2002. This increase has apparently not been at the expense of standard work, but rather at the expense of persons not in employment (unemployed plus inactive). A slightly different development can be identified for Germany, where shares of standard work and non-standard work have increased between 2002 and 2016.
However, Figure 1 illustrates that in Germany the growth of non-standard work was stronger than the growth of standard work in that period. By contrast, in Denmark, the Netherlands and Italy the share of non-standard work grew considerably whereas at the same time the share of standard work went down.
Specific forms of employment are used in German companies to a different degree and are not equally distributed amongst workers. Whereas the use of temporary agency work is above average in manufacturing, fixed-term contracts and all variants of part-time employment are disproportionally visible in service industries. Part-time employment is a domain of women as well as older workers and of people who reconcile education and employment. By contrast, younger workers can be found more often in fixed-term contracts and temporary agency work. Temporary workers are more frequently jobless, job entrants or job returners. Workers with no kind of formal qualification are prevalently represented in all types of non-standard employment. Academics frequently start their career with a fixed-term contract. However, sooner or later they will switch to an open-ended contract (Sperber and Walwei 2017).
Developments in the recent past also indicate that in Germany the proportion of workers with low hourly pay apparently grew only marginally. Based on the Socio-Economic Panel and relying on a threshold value of 2/3 of the median hourly wage, Kalina and Weinkopf (2018) show a proportion of low-wage earners amounting to 22.7% for Germany as a whole for 2016, which is only marginally higher than the corresponding percentage of 21.4% that was calculated for 2005. The largest part of the growth thus occurred before 2005, i.e. before the turnaround of the German labour market took place. The rate of low-wage earners grew by almost five percentage points just between 1998 and 2005 (Figure 2). Several studies have found a strong correlation with nonstandard jobs, as low hourly pay is found more frequently in part-time jobs covered by social security and in particular in mini-jobs. Furthermore, analyses of fixed-term jobs and temporary agency work also reveal sometimes considerable gross wage differentials compared to regular employment (Jahn and Pozzoli 2011; Keller and Seifert 2013).
Developments in the past raise the question of what lies behind the changes. Potential drivers for changes in employment forms and wage inequality are:
· structural shifts in employment,
· changes in labour market institutions,
· changing behaviour (and preferences) of employers and workers,
· market power of the parties involved.
The main findings for Germany in this respect can be found in the following box “Employment forms and wage inequality in Germany: Drivers of change”.
Box: Employment forms and wage inequality in Germany: Drivers of change
One driver for growth in nonstandard work arrangements is structural change in total employment. This is of relevance if employment increases in particular segments of the labour market and nonstandard work can be found in these segments more or less often. According to shift-share analyses, long-term changes in the structure of employment, such as the growth of female employment, the increasing importance of service jobs, the ongoing trend of a higher qualification or the constant ageing of the workforce have made at most a small contribution to the change in the composition of work arrangements (see Walwei 2014b). Only with respect to regular part-time employment (i.e., excluding “mini-jobs”) up to 40% of the increase can be associated to changes in the composition of employment by industries and gender. This is due to the fact that those employed in services and women usually have a higher part-time rate than their counterparts.
Concerning the composition of work arrangements, changes in labour market institutions can also make a difference. In general, institutions define the relative attractiveness of work arrangements and either open up or limit options for those concerned. The “Hartz reforms” in Germany between 2003 and 2005 generated push- and pull-effects in this respect. In particular the fourth stage of the Hartz reforms, which emphasised activation and relaxed the criteria of what constitutes a suitable job, is of importance. Since the reforms, recipients of basic social benefits have to accept any employment, which not only affects the labour market entry of people who are in need of assistance but also affects the job-search behaviour of people who want to avoid having to claim basic social benefits, in the sense of a deterrent effect (see Erlinghagen 2010). In this respect, the labour market reforms could have resulted in push-effects towards employment that is less stable and not always sufficient to secure the worker’s livelihood. These effects were accompanied by the pull-effects, in other words the increased scope for action for firms as a result of deregulating non-standard work arrangements as part of the Hartz reforms (see Dietz et al., 2013).
Analyses further show whether there was an increase in atypical employment after the implementation of the Hartz reforms on the basis of their annual growth rates from 2004 to 2012 (see Himsel et al., 2013). In the case of the deregulated forms of employment, such as temporary agency work and marginal part-time employment, early effects of the reforms in the sense of a strong increase can be identified in 2004. However, this appears to be a one-off effect, as the initially strong increase has not continued to the same extent after 2004. The number of people whose only job is marginally part-time showed only a rather small annual growth after 2004. Temporary agency work went up for quite some time but is influenced in the longer-term by the economic situation like no other employment form (see Antoni and Jahn 2009). Possible reasons why temporary agency work did not accelerate further after the labour market reforms may also have to do with re-regulation in the temporary work sector in recent years such as the introduction of a minimum wage in this industry.
A third source of change are the players on the labour market, employers and workers. While using work arrangements employers are confronted with a trade-off because they have to take into account costs and benefits of certain forms of employment. There might be several reasons for companies to use nonstandard jobs (see Walwei 2014a). They can be used to save total labour costs. Nonstandard jobs may also be an option to reduce extra payments to regular workers for overtime. Particularly temporary employment can, in addition, be utilized as a recruiting device and may therefore increase the efficiency of matching labour supply and demand. Non-standard work arrangements offer a high degree of flexibility to adapt available personal resources to variations in product demand. And finally, nonstandard work can be seen as a kind of buffer to protect core workers.
Changes in preferences of workers could also be partly responsible for the change in the composition of work arrangements. In principle, most workers are expected to prefer a regular employment relationship. Nonetheless, some individuals may desire employment forms like part-time work because they can facilitate the compatibility of employment with other activities such as childcare or education and further training during certain periods of their lives (see Stops and Walwei 2014). In addition, risk preferences may change over time. Sometimes a fixed-term contract with a well-reputed employer might be recognized as a kind of stepping stone for a successful career.
Of further importance is that atypical employment following a period of unemployment can facilitate labour market entry (see Hohendanner and Walwei 2013). However, little evidence has so far been found of flexible employment functioning substantially as a bridge to regular employment (see Gensicke 2010; Lehmer 2012; Brülle 2013). Yet it must be taken into consideration that temporary forms of employment need not be less stable in the long run than a permanent job (see Boockmann and Hagen 2005), as fixed-term employment contracts can be converted into permanent ones, and permanent employees can regularly be dismissed. Likewise, even full-time and permanent employment does not necessarily guarantee an income that is sufficient to secure one’s livelihood if the job is in the low-wage sector (see Bruckmeier et al., 2013; Bruckmeier et al., 2015).
A peculiarity of the labour market is that preferences of employers and workers may not always correspond to one another. An important issue in this respect is the relative market power of the parties involved (Houseman and Osawa 2003). Periods of economic slack and high unemployment can therefore push nonstandard work at the expense of standard work. During such periods of excess supply at the labour market, employers can more easily enforce nonstandard work arrangements. As a consequence of the recent labour market recovery in Germany we can assume declining push-effects into these types of employment.
As regards the causes of wage inequality, the Hartz reforms cannot be seen as the sole cause of the increase in wage inequality since the mid-1990s, since they only became effective from 2003 onwards. The relevant literature cites a whole range of factors that may have fostered this development (see Card et al., 2013). The heavy job losses in eastern Germany after reunification put the collective bargaining system to the test and contributed to a decline in union density and collective agreement coverage. Recent studies indicate that the reduction in the coverage of collective agreements between the mid-1990s and the middle of the last decade can explain a considerable part of the growing wage inequality (Antonczyk et al., 2010a; Antonczyk et al., 2010b). Other possible explanatory factors for the stronger wage disparity are growing international trade, outsourcing trends in some sectors of the economy (Autor and Dorn 2013), increasing immigration of workers with low skill levels, specific effects of technological progress on various skill groups and an increased heterogeneity of firms.
***
Possible implications of digitalization on employment forms and wage inequality are manifold. As already mentioned, digitalization will most likely accelerate structural change in employment. There are indications that the increase in service employment and its relative importance may be pushed further. One can also expect an impact of digitalization on employment by gender. Whereas men are more often performing occupations (e.g. in the field of manufacturing) which may be endangered by automation, women are to a greater extent in occupations requiring social tasks being less susceptible to automation. In addition, women can probably get an easier access to jobs in manufacturing or construction because due to “smart” automation such jobs will be physically less demanding than in the past. As shift-share-analyses have shown, both more employment in services and more female employment tend to increase regular part-time employment (excl. the so-called “mini-jobs”).
As already mentioned neither large job losses nor technological unemployment are very much likely because of digitalization. Therefore, one can assume that digital technologies will not greatly push workers into less protected types of nonstandard work. However, this does not mean that within certain segments of the labour market such impacts may not occur at all. Because considerations in the previous section suggest that jobs with low skill requirements are endangered by digitalization, this may have an impact on the quality of attainable jobs for unskilled workers, e.g. regarding their employment security. Already in the recent past, we observed that low-skill workers bear a higher risk of being employed in a less stable job than more qualified workers.
Due to digitalization, new forms of employment can also arise. In this respect, particularly ICT-based mobile work is of interest. The availability of this opportunity depends, of course, on the particular job. While in jobs dominated by manual tasks and by a high frequency of face-to-face contacts mobile work will not be feasible, many other jobs will not require a permanent presence at the workplace like it was the case in the past. Recent evidence for Germany shows that new types of “home office” occur more often in the case of white-collar workers than blue-collar workers, more often for women than for men and more often for executive than for other staff (Grunau et al., 2019). Up to now, there is no indication that ICT-based home office has rapidly increased in the recent past. Workers report that working from home supports the reconciliation of employment work, but also that this type led to a “blurring of boundaries”. Empirical findings indicate that advantages and disadvantages reported of either employers or employees correspond to a considerable degree (Figure 3).
For employers the issue of working hours gets more attention for other reasons. Decisive in this context is that the Internet and its various applications remove spatial and systematic barriers of work. Due to the obvious “blurring of boundaries” it will get more difficult to measure working hours and to regulate them. If workers are less present at the workplace this implies less control of input. It causes rather a greater focus on output-orientation questioning the traditional concept of measuring working hours. This can create a conflict of interests between both sides of the market, particularly with respect to suitable distinctions between working time and leisure time. In addition, there are limits of mobile work because the advantages of direct interaction between human beings cannot be fully utilized. Valuable face-to-face contacts can lose in importance and the building of trust within a team will probably be made more difficult. Up to now, there is no indication that ICT-based home office has rapidly increased in the recent past (Bundesministerium für Arbeit und Soziales 2015; Grunau et al., 2019).
For employers digital technologies offer even more opportunities. This is mainly due to the assumption that the emergence of these new technologies tends to lower transaction costs of market coordination and enable firms to find proper counterparts more easily. In addition, the almost perfect connectivity increases the speed of interaction between potential contract partners. This has at least two implications. First, the opportunity costs of market coordination will substantially be reduced. This will increase the incentive for companies to outsource activities that can be performed more efficient by external suppliers (subcontractors). On the one hand, a tendency towards more outsourcing may generate additional solo entrepreneurs. On the other hand, this may enable firms to concentrate even more on their core activities. Both developments may push self-employment at the expense of dependent employment. Second, the emergence of digitalization may also reduce the opportunity costs of external flexibility and influences its attractiveness in relation to possible alternatives such as internal flexibility. Due to a substantial increase of relevant information, e.g. concerning the magnitude and quality of future orders as well as the degree of utilization, firms know much more about their current and future demand for labour. If they are able to adapt personal resources to variations in product demand as far as possible, they generate cost advantages. In the digital world freelancers might become at least partly a functional equivalent for other types of external flexibility such as fixed-term contracts. Online platforms offering various services by workers to customers may also partly substitute traditional temporary work agencies as long as workers do not have to be necessarily present at the workplace.
Most of the change regarding forms of employment probably refers to self-employment. This does not only – as already mentioned – concern its relative weight compared to dependent employment but also its formation. Results-only work environments may become more important and may fundamentally change the value chain. Particularly new forms of virtual work (crowd employment) can be seen as a forerunner in this respect.
Crowd employment can be considered as a type of gainful employment that utilizes “smart” platforms in order to enable firms or single persons to provide specific services or particular products (Green and Barnes 2013; Saxton et al., 2013). Crowd workers are usually self-employed. Their activities are based on individual tasks or projects rather than on a regular work arrangement. Activities of this kind are mostly carried out separately, implying a kind of global division of tasks. Crowd employment often occurs in the context of specific activities and in some cases it may be used for more complex projects. “Examples of tasks often commissioned through crowd employment are web content and software development, database building and cleaning, classifying web pages, transcribing scanned documents and audio clips, classifying and tagging images, reviewing documents, checking websites for specific content, validating search results, and designing logos and drafting of slogans for the advertising industry” (Eurofound 2015). The examples suggest that crowd employment is heterogeneous and not suitable for all types of tasks, but rather for certain parts of any job (Kittur et al. 2013).
The use of virtual work can have many advantages for firms (Linnhoff-Popien et al., 2015). The speed of finding potential contractors via intelligent platforms is comparatively high, it enables a selection of contractors in terms of quality, it contributes to low fixed costs of employment and it purifies organizational processes within companies. Crowd employment requires organizational changes within firms, though, because they rely on new types of networking. Nevertheless, it is still quite a new phenomenon. Up to now, the spread of crowd employment seems to be marginal, and there are hardly any reliable data available. A survey carried out in Germany indicates that most crowd employment is not done by individuals as their sole source of income but mainly as a side job. In addition, the survey suggests that almost one half of the interviewed crowd workers would – if available – prefer standard employment, whereas the other half prefer the autonomy which is associated with this type of employment (Leimeister et al., 2015). This autonomy of crowd workers can be characterized as work at fingertips: They are free to decide at any time when, where, and to which extent they want to work (at fingertips). Crowd employment can even be seen as a new type of homework and as a new way of arranging work and life (see Figure 4). However, autonomy also implies the risk of fluctuations in demand. The future potential of this type of employment will very much depend on how this area will be regulated in the near future, e.g. with respect to product market regulations or minimum standards concerning remuneration and social security, and how far it paves the way to a successful career as well as ensuring continuous employability.
The aspect of wage inequality may also be influenced by digitalization. One reason is the already mentioned possible impact of these new technologies on low-skilled workers and other difficult-to-place persons for whom it may in the future even be more difficult to find a reasonable job. Workers need to continuously acquire new skills for their job or to switch occupations. In this regard it is also essential that firms probably make more use of outsourcing activities which do not belong to their core business. Such activities offer an opportunity to partly circumvent collective bargaining and may contribute to rising inequality (Goldschmidt and Schmieder 2015). Nonetheless, one has to bear in mind that concerning potential variations of wage disparities digitalization is only one possible impulse. Regulations such as minimum wages will also be relevant in this context because one has to bear in mind that there is a trade-off between achieving equality on the one hand and restricting market entry, as well as pushing circumvention, on the other hand.
4. Conclusions and policy recommendations
Previous research indicates that new technologies are always associated with complex labour market impacts, which are difficult to assess ex ante as well as isolate from other relevant impulses ex post. In the foreseeable future, the emergence of digitalization will most probably be the main driver of technological change. In general, impacts of technological change on the level of employment must not necessarily be negative. Also in the case of emerging digitalization, massive job losses for the economy as a whole seem to be rather unrealistic. This is mainly because new technologies will open new markets, have the potential to create additional demand and can act as a catalyst for a dynamic structural change. Therefore, productivity increases due to technological change must not substantially hurt the labour market. Moreover, a successful application of digital technologies is one of the preconditions to stabilize the level of employment in the long run. However, there are also serious risks and challenges. Severe problems would occur if societies and their governments tried to prevent these new technologies. There are also threats for those industries and firms that fail to make full use of the new technological opportunities and would then lose their competitiveness. Therefore, it is decisive how far the society as a whole will be able to cope with these new technologies.
The emergence of digital technologies will most likely cause various and considerable shifts in the structure of employment. One can expect more digital products, a stronger digitalized production and a larger degree of digital knowledge. Shifts may refer to the composition of employment by industries, occupations, skill levels, tasks and forms of employment. However, at this moment in time it is easier to identify areas which are susceptible to these new technologies than those which may originally be pushed through productivity gains.
Regarding the long-term tendency of more heterogeneity and inequality in employment, digitalization has a certain potential to become a new driver. First of all, this is due to the impact of digital technologies on structural change. Parts of the labour market, which are favoured by the emergence of digitalization, particularly employment in services or employment of women, are those in which one can already today observe a greater spreading of regular part-time work. There are also indications that digital technologies may influence the relationship between dependent employment and self-employment. Due to the lowering of transaction costs through innovative platforms, market coordination is getting more attractive. New types of self-employment such as crowd employment have a potential in certain market segments and may, therefore, gain in importance in future. Since mobile work and working at home will become much easier through digital technologies, traditional concepts of measuring working hours may lose in importance. They might increasingly be replaced by new forms of output-orientation. There are also no indications that the long-term trend towards wage inequality might fundamentally be reversed through new technologies. This is mainly because low-skilled workers and other problem groups of the labour market would probably face even more difficulties to (re-)enter the labour market and well-paid manufacturing jobs are endangered.
Because of a considerable lack of solid evidence in many respects, policy recommendations at this stage need to be rather cautious and should – if they are also intended to be relevant for other Western countries – address only general issues. Of great importance in this context is, first of all, how far the self-employed are insured against unemployment. Such insurance might be organized on an obligatory or a voluntary basis. Up to now, the German legislation already offers an option for the self-employed to insure themselves. However, due to increasing contributions in 2011 the number of insured self-employed recently decreased (Jahn and Springer 2013). This increases the risk of self-employment who must get out of business to be immediately dependent on basic social income. Regarding probable changes of employment forms favouring self-employment voluntary insurances against unemployment for this group need to be promoted more, which also implies additional incentives to enter the insurance system.
Regarding new forms of online work, standards for remunerations are another important topic. Instead of far-reaching and detailed regulations, platforms, which enable to continuously assess clients would be one option to establish “fair” crowd employment. Alternative or complementary options would be regulations similar to minimum wages or “collective” agreements between crowd workers and their clients, which may define reasonable standards of payment (Cherry 2015; Bundesministerium für Arbeit und Soziales 2016). An increased level of social protection for online workers who can be seen as independent contractors can also be an urgent field of action (Weber 2018).
There is also an obvious need for new compromises concerning flexibility. They particularly refer to what is meant by “standard” in the context of work arrangements. One requirement deals with proper arrangements regarding working hours, which should as far as possible match the needs of both sides of the market. In addition, by redefining standards one has to look for an up-to-date solution of a well-known trade-off. Regulatory changes aiming at more worker protection need to always take into account that they may create barriers for market entry, particularly with respect to less productive workers. Therefore, facilitating the upward mobility of workers would be a modern way to neutralize the trade-off.
References
Antonczyk, D., Fitzenberger, B., Sommerfeld, K. (2010a): Rising Wage Inequality, the Decline of Collective Bargaining, and the Gender Wage Gap. Labour Economics, 17(5): 835-847.
Antonczyk, D., DeLeire, T., Fitzenberger, B. (2010b): Polarization and Rising Wage Inequality: Comparing the U.S. and Germany. In: IZA Discussion Paper 4842.
Antoni, M., Jahn, E. (2009): Do changes in regulation affect employment duration in temporary help agencies? Industrial and Labour Relations Review, 62(2): 226-251.
Autor, D. (2014): Polanyi’s paradox and the shape of employment growth. Paper prepared for Federal Reserve Bank of Kansas. Jackson Hole Conference, August 22.
Autor, D. (2013): The “Task Approach” to labour markets: an overview. Journal for Labour Market Research, 46(1): 185-199.
Autor, D. (2001): Why do temporary help firms provide free general skills training? The Quarterly Journal of Economics, 116(4): 1409-1448.
Autor, D., Dorn, D. (2013): The growth of low-skill service jobs and the polarization of the U.S. labour market. American Economic Review, 103 (5): 1553-1597.
Autor, D., Levy, F., Murnane, R. (2003): The skill content of recent technological change. The Quarterly Journal of Economics, 118(4): 1279-1333.
Bonin, H., Gregory, T., Zierahn, U. (2015): Übertragung der Studie von Frey, Osborne (2013) auf Deutschland. Zentrum für Europäische Wirtschaftsforschung (ZEW).
Boockmann, B., Hagen, T. (2005): Die Bedeutung befristeter Arbeitsverhältnisse für die Zugänge und den Verbleib in Beschäftigung. In: Kronauer, M., Linne, G., (eds.): Flexicurity. Die Suche nach Sicherheit in der Flexibilität. Berlin: 149-168.
Bornemann, H. (2015): Wie digitalisiert ist Deutschland? Prognos trendletter. November 2015.
Bruckmeier, K., Eggs, J., Sperber, C., Trappmann, M., Walwei, U. (2015): Arbeitsmarktsituation von Aufstockern: Vor allem Minijobber suchen nach einer anderen Arbeit. IAB-Kurzbericht, No. 19. Nürnberg.
Bruckmeier, K., Eggs, J., Himsel, C., Trappmann, M., Walwei, U. (2013): Aufstocker im SGB II: Steinig und lang - der Weg aus dem Leistungsbezug. IAB-Kurzbericht, No. 14. Nürnberg.
Brülle, J. (2013): Unterschiede in den Arbeitsmarktchancen von atypisch Beschäftigten: Effekte von Beschäftigungsformen oder Erwerbspräferenzen? Zeitschrift für Soziologie, 42 (2): 157-179.
Brynjolfsson, E., McAfee, A. (2011): Race against the machine: How the digital revolution is accelerating innovation, driving productivity, and irreversibly transforming employment and the economy. Digital Frontier Press. Lexington, MA.
Brzeski, C., Burk, I. (2015): Die Roboter kommen. Folgen der Automatisierung für den deutschen Arbeitsmarkt, ing-diba.de
Bundesministerium für Arbeit und Soziales (2015): Mobiles und entgrenztes Arbeiten. Monitor. Bundesministerium für Arbeit und Soziales. Berlin.
Bundesministerium für Arbeit und Soziales (2016): Auswirkungen der Digitalisierung auf Beschäftigung und Erwerbsformen. Arbeiten 4.0.
Card, D.E., Heinig, J., Kline, P. (2013): Workplace Heterogeneity and the Rise of German Wage Inequality. In: Quarterly Journal of Economics, 128: 967-1015.
Cherry, M.A. (2015): Mindestlohn für Crowdarbeit? Regelungen zum gesetzlichen Mindestlohn im digitalen Zeitalter. In: Benner, C. (ed.). Crowdwork – Zurück in die Zukunft? Perspektiven digitaler Arbeit. Bund-Verlag, Frankfurt/Main.
Chui, M., Manyika, J., Miremadi, M. (2015): Four fundamentals of workplace automation. McKinsey Quarterly, November 2015.
Dengler, K., Matthes, B. (2018): Substituierbarkeitspotenziale von Berufen. Wenige Berufsbilder halten mit der Digitalisierung Schritt. IAB-Kurzbericht. No. 4, Nürnberg.
Dietz, M., Himsel, C., Walwei, U. (2013): Wandel der Erwerbsformen * welche Rolle spielen strukturelle Änderungen am Arbeitsmarkt? Arbeit. Zeitschrift für Arbeitsforschung, Arbeitsgestaltung und Arbeitspolitik, 22 (2): 85-104.
Erlinghagen, M. (2010): Mehr Angst vor Arbeitsplatzverlust seit Hartz?: Langfristige Entwicklung der Beschäftigungsunsicherheit in Deutschland. SOEP papers on Multidisciplinary Panel Data Research 279. Berlin.
Eurofound (2015): New forms of employment, Publications Office of the European Union, Luxembourg.
Frey, C.B., Osborne, M.A. (2013): The future of employment: how susceptible are jobs to computerisation? Oxford University.
Gensicke, M., Herzog-Stein, A., Seifert, H., Tschersich, N. (2010): Einmal atypisch, immer atypisch beschäftigt? Mobilitätsprozesse atypischer und normaler Arbeitsverhältnisse im Vergleich. WSI-Mitteilungen, 63 (4): 179-187.
Goldschmidt, D., Schmieder, J.F. (2015): The Rise of Domestic Outsourcing and the Evolution of the German Wage Structure. IZA Discussion Paper Series, No. 9194. 52 pp.
Green, A., Barnes, S.-A. (2013): CrowdEmploy Part I: Crowdsourcing for paid work. An empirical investigation into the impact of crowdsourcing for paid work on employability. Warwick Institute for Employment Research/Institute for Prospective Technological Studies (IPTS), Warwick/Seville.
Grossman, G.M., Helpman, E. (1991): Quality ladders in the theory of growth. The Review of Economic Studies, 58(1): 43-61.
Grunau, P., Ruf, K., Steffes, S., Wolter, S. (2019): Mobile Arbeitsformen aus Sicht von Betrieben und Beschäftigten: Homeoffice bietet Vorteile, hat aber auch Tücken. IAB-Kurzbericht, No. 11, Nürnberg.
Himsel, C., Walwei, U., Dietz, M. (2013): Erwerbslandschaft im Umbruch: Was die Reformen zum Wandel der Beschäftigungsformen beigetragen haben. IAB-Forum, No. 2. Nürnberg: 28-35.
Hohendanner, C., Walwei, U. (2013): Arbeitsmarkteffekte atypischer Beschäftigung. In: WSI-Mitteilungen, 66: 239-246.
Houseman, S.N., Osawa, M. (2003): Nonstandard work in developed countries – Causes and consequences. W.E. Upjohn Institute for Employment Research, Kalamazoo.
Jahn, E., Pozzoli, D.(2011): Does the sector experience affect the pay gap for temporary agency workers? IZA discussionpaper, 5837. Bonn: 1-39.
Jahn, E., Springer, A. (2013): Arbeitslosenversicherung. Auch Selbständige nehmen Unterstützung in Anspruch. IAB-Kurzbericht, No. 26, Nürnberg.
Kalleberg, A.L. (2011): Good Jobs, Bad Jobs. The Rise of Polarized and Precarious Employment Systems in the United States, 1970s to 2000s. New York: Russell Sage Foundation. 292 pp.
Kalina, T., Weinkopf, C. (2018): Niedriglohnbeschäftigung 2016: Beachtliche Lohnzuwächse im unteren Lohnsegment, aber weiterhin hoher Anteil von Beschäftigten mit Niedriglöhnen. IAQReport, Nr. 2018-06. Duisburg.
Keller, B.K., Seifert, H. (2013): Atypische Beschäftigung zwischen Prekarität und Normalität * Entwicklung, Strukturen und Bestimmungsgründe im Überblick. Berlin.
Kittur, A., Nickerson, J., Bernstein, M.S., Gerber, E.M., Shaw, A., Zimmerman, J. (2013): The future of crowd work. Stanford University.
Lehmer, F. (2012): Dient die Arbeitnehmerüberlassung für Langzeitarbeitslose als Brücke in nachhaltige Beschäftigung? In: Sozialer Fortschritt, 61 (8): 190-197.
Leimeister, J.M., Zogay, S., Durward, D., Bretschneider, U. (2015): Neue Geschäftsfelder durch Crowdsourcing: Crowd-basierte Start-ups als Arbeitsmodell der Zukunft. In: Hoffmann, R., Bogedan, C. (eds.): Arbeit der Zukunft, 141-158. Campus. Frankfurt/Main.
Linnhoff-Popien, C., Zaddach, M., Grahl, A. (2015): Marktplätze im Umbruch. Springer-Verlag. Berlin/Heidelberg.
Lorenz, M., Rüßmann, M., Strack, R., Lueth, K.L., Bolle, M. (2015): Man and Machine in Industry 4.0. How Will Technology Transform the Industrial Workforce Through 2025? The Boston Consulting Group.
Mückenberger, U. (2010): Krise des Normalarbeitsverhältnisses. Ein Umbauprogramm. Zeitschrift für Sozialreform. 56(4): 403-420.
Mückenberger; U. (1985): Die Krise des Normalarbeitsverhältnisses, Teil 1 und Teil 2. Zeitschrift für Sozialreform. 31(4): 415-435; 457-475.
Osterman, P. (2000): Work Reorganization in an Era of Restructuring. Trends in Diffusion and Effects on Employee Welfare. Industrial and Labor Relations Review. 53(2): 179-196.
Petersen, T. (2015): Globalisierung, Digitalisierung und Einkommensungleichheit. Policy Brief, No. 1. Bertelsmann-Stiftung. Gütersloh.
Rhein, T., Walwei, U. (2018): Forms of employment in European comparison. In: IAB-Forum, 09.05.2018.
Romer, P. (1990): Endogenous technological change. Journal of Political Economy, 98(5): 71-102.
Saxton, G.D., Oh, O., Kishore, R. (2013): Rules of crowdsourcing: Models, issues and systems of control. Information Systems Management, 30(1): 2-20.
Schumpeter, J. (1912): Theorie der wirtschaftlichen Entwicklung. Dunckler&Humblot, Leipzig.
Solow, R.M. (1956): A contribution to the theory of economic growth. Quarterly Journal of Economics, 70 (1): 65-94.
Sperber, C., Walwei, U. (2017): Treiber des Erwerbsformenwandels: Wer hat welchen Job? In: WSI-Mitteilungen, 70(1): 16-26.
Stops, M., Walwei, U. (2014): Was ist eigentlich noch normal? * ein Beitrag zu den Möglichkeiten und Grenzen einer Bewertung von Erwerbsformen. In: Esser, C.-H. (eds.), Arbeitsmärkte im Wandel der Wirtschafts- und Sozialpolitik. Neue Herausforderungen und Chancen. Festschrift für Franz Egle. Frankfurt am Main: 97-117.
Vogler-Ludwig, K., Düll, N., Kriechel, B. (2016): Arbeitsmarkt 2030. Wirtschaft und Arbeitsmarkt im digitalen Zeitalter. Prognose 2016. Analyse der zukünftigen Arbeitskräftenachfrage und des -angebots in Deutschland auf Basis eines Rechenmodells. Studie im Auftrag des Bundesministeriums für Arbeit und Soziales, München.
Walwei, U. (2014a): Curing the sick man: The German labour market on the way to good health? Institut für Arbeitsmarkt- und Berufsforschung. IAB Current reports. Nürnberg.
Walwei, U. (2014b): Times of change: what drives the growth of work arrangements in Germany? Journal for Labour Market Research, 47 (3): 183-204.
Weber, E. (2018): Setting out for digital social security. IAB-Forum, 13.09.2018, Nürnberg.
Wolter, M.I., Mönnig, A., Hummel, M., Schneemann, C., Weber, E., Zika, G., Helmrich, R., Maier, T., Neuber-Pohl, C. (2015): Industry 4.0 and the consequences for labour market and economy. Scenario calculations in line with the BIBB-IAB qualifications and occupational field projections. IAB-Forschungsbericht, No. 8, Nürnberg.
Zika, G., Schneemann, C., Kalinowsik, M., Maier, T., Winnige, S., Grossmann, A., Mönnig, A., Parton, F., Wolter, M.I. (2019): BMAS-Prognose “Digitalisierte Arbeitswelt”. Berlin: Bundesministerium für Arbeit und Soziales. Forschungsbericht, 526/1K, Berlin. 30 pp.
END NOTES
[1] The contribution is a shortened and partly revised version of the following study of the author: “Digitalization and structural labour market problems. The case of Germany” published as ILO-Research Paper, Geneva. 38 pp.
[2] A more recent study by Chui et al. (2015) obtained an almost similar result. The authors investigated a bottom line of 45% of work activities in the United States which could be automated using already available technologies. If artificial intelligence were to reach the median level of human performance, an additional 13% of work activities could be automated.