L AUTOMATISATION ET L EMPLOI

Transcription

1 L AUTOMATISATION ET L EMPLOI Marco Vivarelli Professor, Università Cattolica del Sacro Cuore, Milano Professorial Fellow UNU-MERIT, Maastricht Research Fellow IZA, Bonn AUDITION CONSEIL D ORIENTATION POUR L EMPLOI, PARIS;

2 TODAY ALARM The pervasive diffusion of ICT and automation has led to a dramatic adjustment of employment, both in quantitative (employment levels) and qualitative terms (skills). Recently, the arrival of 3D printing, self-driving autonomous cars (Tesla, Apple, Google) and agricultural robots has raised again a widespread fear of a new wave of technological unemployment. Moreover, not only agricultural and manufacturing employment appears at risk, but employees in services - including cognitive skills - are no longer safe (see, for instance, how Uber - just a software tool - is fully crowding out taxi companies). Finally, these recent trends have interlinked with the financial and economic crises and with the slow recovery afterwards, often showing a jobless nature (OECD, 2016).

3 however, this is such an old issue: Ned Ludd and Captain Swing against textile and threshing machineries Captain Swing by Eric Hobsbawm and George Rudé, New york, Pantheon Books, 1968

4 RICARDO S SURPRISE.the opinion, entertained by the labouring class, that the employment of machinery is frequently detrimental to their interests, is not founded on prejudice and error, but is conformable to the correct principles of political economy (Ricardo, 1951, vol 1, p. 387; third edition, 1821) However, technological unemployment is considered an exception, occurring only when production does not grow, otherwise a compensation always occurs:

5 THE FAIRY TALE OF TRICKLE-DOWN ECONOMICS NEW MACHINES C P D Y L PC PROCESS INNOVATION DIRECT LABOUR-SAVING EFFECT NO PC Π I D Y L L u W L

6 CLASSICAL AND CONTROVERSIAL ISSUE (1) Machines cannot be constructed without considerable labour, which gives occupation to the hands they throw out of employ. (Say, 1967, p. 87; first ed. 1803); HOWEVER: the machine can only be employed profitably, if it is the (annual) product of fewer men than it replaces. (Marx, 1969, p. 552; first ed ); The introduction of machines is found to reduce prices in a surprising manner. And if they have the effect of taking bread from hundreds, formerly employed in performing their simple operations, they have that also of giving bread to thousands. (Steuart, 1966, vol. II, p. 256; first ed. 1767); HOWEVER:..the increased demand for commodities by some consumers, will be balanced by a cessation of demand on the part of others, namely, the labourers who were superseded by the improvement. (Mill, 1976, p.97; first ed. 1848)

7 CLASSICAL AND CONTROVERSIAL ISSUE (2) I have before observed, too, that the increase of net incomes, estimated in commodities, which is always the consequence of improved machinery, will lead to new saving and accumulation (Ricardo, 1951, vol 1, p. 396; third edition, 1821) ; HOWEVER: The accumulation of capital, though originally appearing as its quantitative extension only, is effected, as we have seen, under a progressive qualitative change in its composition, under a constant increase of its constant, at the expense of its variable constituent. (Marx, 1961, vol. 1; p. 628; first ed. 1867). Entirely new branches of production, creating new fields of labour, are also formed, as the direct result either of machinery or of the general industrial changes brought about by it. (Marx, 1961, vol. 1; p. 445 first ed. 1867); HOWEVER: But the places occupied by these branches in the general production is, even in the most developed countries, far from important (ibidem).

8 CURRENT ECONOMIC THEORY TELLS BASICALLY THE SAME COMPENSATION STORY, THEREFORE: This neo-classical general equilibrium framework can be said to correspond most closely to present-day traditional economic views on technical change and employment. Technological change may indeed result in some temporary unemployment, but with efficiently operating labour and capital markets there is no basic economic problem arising from the introduction of new technology (Freeman, C. and Soete, L., Work for All or Mass Unemployment, London: Pinter, 1994, p.25)

9 THE CRITIQUE NEW MACHINES BUT: Π ; PROC; SCRAP PC L D C P D Y L BUT: ; η<1; SAY S LAW PROCESS INNOVATION DIRECT LABOUR-SAVING EFFECT NO PC Π I D Y L BUT: I PROC; SAY S LAW W L u W L BUT: D ; LOW σ (K,L); PATH DEP.

10 RECENT THEORETICAL MODELS ARE BASED ON THE SAME COMPENSATION FRAMEWORK Examples: Neary, 1981; Stoneman, 1983; Kautsolacos, 1984; Hall and Heffernan, 1985; Waterson and Stoneman, 1985; Dobbs et al., 1987; Layard et al., Indeed, compensation cannot be assumed ex ante (as implicitly done by theoretical studies), since the final employment outcome depends on crucial parameters such as the % of product innovation, expectations, the demand elasticity, the elasticity of substitution between K and L, and so on. In fact, since the 90s, no further relevant theoretical contributions are put forward, with the focus moving to the empirical studies (for a critical discussion of the recent theoretical models and for aggregate and sectoral empirical studies, see Vivarelli, 1995; Vivarelli and Pianta, 2000). Empirical literature is developed at three levels depending on the disaggregation of data (macroeconomic, sectoral and firm level analysis) and using different proxies for technology. The recent literature focuses on the micro level, with pros and cons.

11 E NARROW-MINDED OPTIMISTIC ECONOMISTS VS DOOMSTERS Keynes, J.M., 1930, Economic Possibilities for our Granchildren, Leontief and Duchin, 1986, The future Impact of Automation on Workers, Rifkin, J. 1995, The End of Work, Brynjolfsson and McAfee, A., 2011, Race Against the Machine: How the Digital Revolution is Accelerating Innovation, Driving Productivity, and Irreversibly Transforming Employment and the Economy : the root of the current employment problems is not the Great Recession, but rather a Great Restructuring characterized by an exponential growth in computers processing speed having an ever-bigger impact on jobs, skills, and the whole economy. Frey and Osborne, 2013, The future of employment: how susceptible are jobs to computerisation?, predict that 47% of the occupational categories are at high risk of being automated, including a wide range of service/white-collar/cognitive tasks such as accountancy, logistics, legal works, translation and technical writing, etc.

12 ECONOMIC THEORY IS INCONCLUSIVE. EMPIRICALLY: I N N O V A T I O N - I N P U T R & D E T C PROD PROC PROD & PROC I N N O V A T I O N - O U T P U T JOB CREATION JOB DESTRUCTION

13 MY KEY FINDINGS (1) Pros: R&D fosters labor-friendly product innovation that leads to job creation. Product innovation may imply the emergence of new firms and new sectors and thus new jobs. Price and income compensation mechanisms can counterbalance the initial displacement of workers that occurs following process innovation. Industrial and innovation policies that support R&D and product innovation, especially in high-tech sectors, can foster job creation.

14 MY KEY FINDINGS (2) Cons: Process innovation may displace labor and create technological unemployment. New products may displace older products and so impede the job-creation effects implied by the diffusion of the new activities. Market and institutional rigidities can impede the price and income compensation mechanisms that work to lessen job destruction. The job-creation impact of innovation is often limited to product innovation and to the high-tech sectors. Therefore: Safety nets are necessary for the possible job losses due to process innovation in non-high-tech sectors.

15 SOME REFERENCES Vivarelli, M., The Economics of Technology and Employment: Theory and Empirical Evidence. Aldershot: Elgar. Vivarelli, M., Pianta, M. (eds), The Employment Impact of Innovation: Evidence and Policy. London: Routledge. Piva, M., Vivarelli, M., The skill bias: Comparative evidence and an econometric test. International Review of Applied Economics, 16, Piva, M., Vivarelli, M., Innovation and employment: Evidence from Italian microdata. Journal of Economics, 86, Piva, M., Santarelli, E., Vivarelli, M., The skill bias effect of technological and organisational change: Evidence and policy implications. Research Policy, 34, Bogliacino, F., Vivarelli, M., The job creation effect of R&D expenditures. Australian Economic Papers, 51, Bogliacino, F., Piva, M., Vivarelli, M., R&D and employment: An application of the LSDVC estimator using European data. Economics Letters, 116, Vivarelli, M., Technology, employment and skills: an interpretative framework. Eurasian Business Review, 3, Vivarelli, M., Innovation, employment and skills in advanced and developing countries: A survey of economic literature. Journal of Economic Issues, 48, Van Roy, V., Vertesy, D., Vivarelli, M., Innovation and employment in patenting firms: Empirical evidence from Europe. IZA Discussion Papers 9147, Bonn.

16 THANK YOU

17 SBTC vs RBTC R&D and innovation exhibit a relevant impact on both the quantity (levels) and the quality (skills) of employment. The quality of workers comes in as a critical variable due to the fact that new technologies might ask for specific skills, : the so-called: Skill-biased technological change (SBTC). Initially proposed by Griliches (1969) and Welch (1970), the SBTC hypothesis is based on the idea that there is complementarity between new technologies and skilled workers, given that only the latter are able to implement effectively and efficiently those technologies, meanwhile a substitution effect between new technologies and unskilled workers is likely to happen (see Berman et al., 1994; Machin and Van Reenen,1998). The SBTC hypothesis has been widely supported by the empirical evidence. More recently, the SBTC hypothesis has been encompassed by the polarization approach or Routine-biased technological change (RBTC), where both high-skill, high-wage jobs and low-skill, low-wage jobs seem to be rising. According to this approach, new technologies destroy routine jobs (possibly including cognitive and middle-skill tasks) while creating opportunities in professional categories and skills which turn out to be novel and different from previous ones (including both the high-skill and nonroutinized unskilled tasks such as those in personal and health services; see Autor et al., 2006; Goos et al., 2014).

18 A SHORTAGE IN SKILLED LABOUR IMPLIES UNEMPL0YMENT AMONG UNSKILLED LABOUR TWO POSSIBLE SOLUTIONS: AMERICAN (FLEXIBILITY IN RELATIVE WAGES) VS NORDIC (TRAINING AND RETRAINING)

19 PREVIOUS MICROECONOMETRIC STUDIES Van Reenen, 1997: positive impact of innovation, UK. Doms et al., 1997: positive effect of advanced manufacturing technologies, US. Smolny, 1998: positive impact of product innovation, West Germany. Greenan and Guellec, 2000: positive effect of innovation at the firm-level, but negative at the sectoral level (still positive for product innovation), France. Greenhalgh et al., 2001: positive impact of R&D, UK, but only in the High-Tech. Piva and Vivarelli (2005): positive impact of innovation, Italy. Harrison et al. (2008): positive effect of product innovation and (slightly) negative of process innovation (strong compensation in services), Germany- France-UK-Spain. Hall et al (2008): positive impact of product innovation, Italy. Lachenmaier and Rottmann (2011): positive impact of innovation (including process innovation), no sectoral differences, Germany. Coad and Rao (2011), positive impact of innovation, stronger for fast-growing firms, US (data only from high-tech manufacturing). Bogliacino et al (2012), positive impact of R&D, but only in services and hightech manufacturing, not in the more traditional manufacturing sectors.

20 NOVELTIES OF THE FOLLOWING STUDY (Van Roy, Vertesy, Vivarelli, 2015) IN COMPARISON WITH PREVIOUS LITERATURE European coverage (Lisbon-Barcelona policy targets) vs national datasets Large international longitudinal dataset vs either cross section or short panel Continuous variable (patents) vs proxies of innovation (often dummies) The impact of innovation is also measured from a quality perspective (forward-citation weighted patents that take into account the technological importance of innovation) Sectoral splitting vs aggregate studies: manufacturing, services, high-tech manufacturing (OECD classification) HOWEVER: The patent indicator is characterised by some limitations

21 DATA We matched accounting company data originating from ORBIS with patent and patent quality information from the OECD PATSTAT dataset using firm-patent concordance tables developed by EPO and the Office for Harmonization in the Internal Market (OHIM), while sectoral deflators have been taken from EUROSTAT sources. The matched dataset covers 63,561 EU-based, patenting firms from 27 EU Member States for the years and belonging to manufacturing and service sectors. We then: (1) excluded firms for which either sectoral belonging, employment, value added, fixed assets or cost of labor were missing or not positive (-60%); (2) we dropped outliers in both levels and growth (-4%); ending with a workable unbalanced panel of 19,978 companies (104,074 observations). Weighted patents i, t n = p= 1 it, 1+ Forward citations Max. Forward citations p, t, t, f f

22 SPECIFICATION Dynamic demand for labour augmented with innovation: Where: l = employees y = value added w = gross wage per employee invest = gross investment innov = patent counts or citation-weighted patent counts ε is the idiosyncratic individual and time-invariant firm's fixed effect and ν the usual error term. Lower case letters indicate natural logarithms

23 EMPIRICAL RESULTS (I) FULL SAMPLE Employment Employment Employment t *** 0.670*** (0.016) (0.016) Value added 0.301*** 0.302*** (0.015) (0.015) Patents (0.040) Weighted patents 0.050** (0.021) Gross investments 0.135*** 0.131*** (0.037) (0.037) Labor cost per employee *** *** (0.095) (0.096) Constant 0.408*** 0.425*** (0.059) (0.060) Time, industry and country dummies included included Observations Number of firms Wald test *** *** AR(1) *** *** AR(2) 2.9*** 3.01*** AR(3) Hansen test *** ***

24 EMPIRICAL RESULTS (II) MANUFACTURING VS SERVICES Employment Manufacturing Services Employment t *** 0.686*** 0.589*** 0.585*** (0.015) (0.015) (0.030) (0.030) Value added 0.285*** 0.284*** 0.397*** 0.399*** (0.014) (0.014) (0.030) (0.030) Patents (0.045) (0.091) Weighted patents 0.048** (0.024) (0.040) Gross investments *** 0.160*** (0.036) (0.036) (0.052) (0.051) Labor cost per employee ** ** *** *** (0.102) (0.103) (0.156) (0.152) Constant 0.379*** 0.394*** 0.595*** 0.619*** (0.062) (0.063) (0.090) (0.089) Time, industry and country dummies included included included included Observations Number of firms Wald test *** *** *** *** AR(1) *** *** *** *** AR(2) 2.18** 2.18** 1.81* 1.78* AR(3) Hansen test *** *** *** ***

25 EMPIRICAL RESULTS (III) HIGH-TECH VS LOW-TECH MANUFACTURING FIRMS Employment High-tech R&D Low-tech R&D Employment t *** 0.671*** 0.692*** 0.694*** (0.017) (0.017) (0.020) (0.019) Value added 0.291*** 0.293*** 0.289*** 0.283*** (0.016) (0.016) (0.018) (0.018) Patents 0.115*** (0.043) (0.079) Weighted patents 0.080*** (0.025) (0.038) Gross investments 0.069** 0.063** (0.030) (0.030) (0.036) (0.036) Labor cost per employee *** *** ** * (0.113) (0.113) (0.130) (0.130) Constant 0.477*** 0.499*** 0.345*** 0.366*** (0.068) (0.070) (0.087) (0.082) Time, industry and country dummies included included included included Observations Number of firms Wald test *** *** *** *** AR(1) *** *** *** *** AR(2) Hansen test *** *** *** ***

26 CONCLUSIONS AND POLICY IMPLICATIONS Our findings confirm the possible labor-friendly nature of innovation at the firm level, in line with prior empirical research. However, our sectoral estimates show that this positive employment impact is statistically significant only in high- and medium-tech manufacturing sectors, while irrelevant in low-tech manufacturing and services. Therefore, patented innovations fully display their labor-friendly nature in the new and emerging sectors, characterized by higher technological opportunities, by higher demand elasticity and by a likely dominance of product innovation. These outcomes prove that the aim of the EU2020 strategy - that is to develop an European economy based on knowledge and innovation - points in the right direction also in terms of job creation. Moreover - since our impact variable takes into account the quality of the introduced innovation - for policy makers it is also reassuring to know that the demand for labor may further increase as the quality of innovation increases. However, the positive and significant employment impact of innovation is not equally detectable across the different sectors. This is something that should be taken into account by a European innovation policy which considers employment as one of its specific targets.

27 CAVEATS It is important to keep in mind that this study has only tested the labor-friendly nature of patented innovation, while neglecting the possible labor-saving impact of non-patented process innovation. This means that embodied technological change and process innovation with their possible adverse impact on employment are probably underestimated in this work. Our citation-weighted patent indicator may be a more sophisticated measure of innovation than sheer patent counts, but it should be noted that patents are imperfect indicators of innovation, particularly for firms in the service sectors. This study has been conducted on a sample of medium-large IPR-intensive firms; therefore, generalizing our results to more aggregate levels is not straight-forward and must take into consideration possible biases in our data coverage.