Service offshoring and wages: worker-level evidence from Italy

Transcription

1 Service offshoring and wages: worker-level evidence from Italy Elisa Borghi Università Carlo Cattaneo - LIUC Rosario Crinò Centro de Estudios Monetarios y Financieros (CEMFI) Final Workshop MIUR-PRIN Project Production, R&D and Knowledge Offshoring: Economic Analyses and Policy Implications September 25,2013

2 Introduction Research Question Effects of service offshoring on the wages of Italian workers

3 Introduction Research Question Effects of service offshoring on the wages of Italian workers Methodology Worker-level wage regressions

4 Introduction Research Question Effects of service offshoring on the wages of Italian workers Methodology Worker-level wage regressions Motivation ICT development and fear of service offshoring

5 Introduction Research Question Effects of service offshoring on the wages of Italian workers Methodology Worker-level wage regressions Motivation ICT development and fear of service offshoring Main results

6 Introduction Research Question Effects of service offshoring on the wages of Italian workers Methodology Worker-level wage regressions Motivation ICT development and fear of service offshoring Main results No significant reductions in workers wages

7 Introduction Research Question Effects of service offshoring on the wages of Italian workers Methodology Worker-level wage regressions Motivation ICT development and fear of service offshoring Main results No significant reductions in workers wages Contribution to widening the wage gap between more and less skilled employees

8 Introduction Research Question Effects of service offshoring on the wages of Italian workers Methodology Worker-level wage regressions Motivation ICT development and fear of service offshoring Main results No significant reductions in workers wages Contribution to widening the wage gap between more and less skilled employees Heterogeneity in the effects of service offshoring

9 Literature Fraction of jobs potentially at risk of offshoring is quite large (roughly 25% in developed countries ) [Blinder (2006, 2009); Kroll (2007); Moncarz et al. (2008); Blinder and Krueger (2009); Kletzer (2009)]

10 Literature Fraction of jobs potentially at risk of offshoring is quite large (roughly 25% in developed countries ) [Blinder (2006, 2009); Kroll (2007); Moncarz et al. (2008); Blinder and Krueger (2009); Kletzer (2009)] Actual impact of service offshoring on the labor market has been moderate in many developed countries (2 4% of job separation in U.S. in the first half of past decade) [Bhagwati et al. (2004); Mankiw and Swagel (2006); Kirkegaard (2007)]

11 Literature Fraction of jobs potentially at risk of offshoring is quite large (roughly 25% in developed countries ) [Blinder (2006, 2009); Kroll (2007); Moncarz et al. (2008); Blinder and Krueger (2009); Kletzer (2009)] Actual impact of service offshoring on the labor market has been moderate in many developed countries (2 4% of job separation in U.S. in the first half of past decade) [Bhagwati et al. (2004); Mankiw and Swagel (2006); Kirkegaard (2007)] Industry- or firm-level studies: small effects on domestic employment. (even positive in some cases, by boosting productivity and expanding scale of firms operations) [Recent contributions: Liu and Trefler (2011); Hijzen et al. (2011); Crinò (2012)]

12 Literature Evidence on Italy: Service offshoring has not caused significant reductions in domestic employment over the last two decades [Falzoni and Tajoli (2011) ; Crinò (2010)]

13 Literature Evidence on Italy: Service offshoring has not caused significant reductions in domestic employment over the last two decades [Falzoni and Tajoli (2011) ; Crinò (2010)] Service offshoring affects the skill composition of labor demand, favouring high-skilled workers [Crinò (2010a, 2010b, 2012)]

14 Literature Evidence on Italy: Service offshoring has not caused significant reductions in domestic employment over the last two decades [Falzoni and Tajoli (2011) ; Crinò (2010)] Service offshoring affects the skill composition of labor demand, favouring high-skilled workers [Crinò (2010a, 2010b, 2012)] Geishecker and Gorg (2012): individual level wage data for U.K.: raise in the skill premium

15 Literature Evidence on Italy: Service offshoring has not caused significant reductions in domestic employment over the last two decades [Falzoni and Tajoli (2011) ; Crinò (2010)] Service offshoring affects the skill composition of labor demand, favouring high-skilled workers [Crinò (2010a, 2010b, 2012)] Geishecker and Gorg (2012): individual level wage data for U.K.: raise in the skill premium Wage level regression: material offshoring raises the skill premium [Hummels et al. (AER - forthcoming)]

16 Focus on wage effects Contribution

17 Contribution Focus on wage effects Rich micro-level data (matched employer-employee data)

18 Contribution Focus on wage effects Rich micro-level data (matched employer-employee data) Worker-, firm- and industry-level characteristics considered

19 Contribution Focus on wage effects Rich micro-level data (matched employer-employee data) Worker-, firm- and industry-level characteristics considered Focus on job stayers: control for worker-firm pairs

20 Matched Employer-Employee Data Administrative records from the Italian National Social Security Institute (INPS)

21 Matched Employer-Employee Data Administrative records from the Italian National Social Security Institute (INPS) All individuals born in four different dates with at least one contract registered at INPS in a given year

22 Matched Employer-Employee Data Administrative records from the Italian National Social Security Institute (INPS) All individuals born in four different dates with at least one contract registered at INPS in a given year Full time workers 16-65

23 Matched Employer-Employee Data Administrative records from the Italian National Social Security Institute (INPS) All individuals born in four different dates with at least one contract registered at INPS in a given year Full time workers ,000 individuals; 60,000 per year

24 Matched Employer-Employee Data Administrative records from the Italian National Social Security Institute (INPS) All individuals born in four different dates with at least one contract registered at INPS in a given year Full time workers ,000 individuals; 60,000 per year Time span:

25 Employer-employee data Table 2 - Descriptive Statistics on the Matched Employer-Employee Data All sample Manufacturing Services White-Collar Blue-Collar Mean Std. Dev. Mean Std. Dev. Mean Std. Dev. Mean Std. Dev. Mean Std. Dev. a) Workers Age Daily wage Gender: Female Occupation: Executives Occupation: Managers & white-collar Occupation: Blue-collar Occupation: Apprentices Industry: Services b) Firms Size (employees) Organization: Single firm Organization: Parent firm Organization: Subsidiary Industry: Services Source: Administrative records from INPS.

26 17 industries, ATECO81 Offshoring Indicators

27 Offshoring Indicators 17 industries, ATECO81 Share of imported private services in total input purchases

28 Offshoring Indicators 17 industries, ATECO81 Share of imported private services in total input purchases Imports of services: ISTAT s Import Matrices

29 Offshoring Indicators 17 industries, ATECO81 Share of imported private services in total input purchases Imports of services: ISTAT s Import Matrices Total input purchases: ISTAT s Input-Output accounts

30 Offshoring Indicators 17 industries, ATECO81 Share of imported private services in total input purchases Imports of services: ISTAT s Import Matrices Total input purchases: ISTAT s Input-Output accounts 4 different services:

31 Offshoring Indicators 17 industries, ATECO81 Share of imported private services in total input purchases Imports of services: ISTAT s Import Matrices Total input purchases: ISTAT s Input-Output accounts 4 different services: Finance and insurance

32 Offshoring Indicators 17 industries, ATECO81 Share of imported private services in total input purchases Imports of services: ISTAT s Import Matrices Total input purchases: ISTAT s Input-Output accounts 4 different services: Finance and insurance Computing and information

33 Offshoring Indicators 17 industries, ATECO81 Share of imported private services in total input purchases Imports of services: ISTAT s Import Matrices Total input purchases: ISTAT s Input-Output accounts 4 different services: Finance and insurance Computing and information R&D

34 Offshoring Indicators 17 industries, ATECO81 Share of imported private services in total input purchases Imports of services: ISTAT s Import Matrices Total input purchases: ISTAT s Input-Output accounts 4 different services: Finance and insurance Computing and information R&D business services

35 Offshoring indicators M sht = imports of services (s= service; h= industry; t= time)

36 Offshoring indicators M sht = imports of services (s= service; h= industry; t= time) Total service imports at the industry level: MS ht = 4 s=1 M sht

37 Offshoring indicators M sht = imports of services (s= service; h= industry; t= time) Total service imports at the industry level: MS ht = 4 s=1 M sht Proxy for service offshoring (normalizing by the total purchases of intermediate inputs in each industry, I ) SOS ht = MS ht I ht

38 Offshoring indicators M sht = imports of services (s= service; h= industry; t= time) Total service imports at the industry level: MS ht = 4 s=1 M sht Proxy for service offshoring (normalizing by the total purchases of intermediate inputs in each industry, I ) SOS ht = MS ht I ht Four disaggregate indicators for all the services covered by the Import Matrices.

39 Offshoring indicators Table 3 - Descriptive Statistics on the Offshoring Indicators a) Manufacturing Mean St. dev change % SOS SOS_FI SOS_CI SOS_RD SOS_BS MOS b) Services Mean St. dev change % SOS SOS_FI SOS_CI SOS_RD SOS_BS MOS SOS: service offshoring, aggregate; SOS_FI: Finance and insurance; SOS_CI: computer and information; SOS_RD: R&D; SOS_BS: business services; MOS: Material offshoring. Source: ISTAT.

40 Empirical model where: ln w ijht = α ij + α t + β 1 Γ ht + θ Θ it + ω Ω jt + λ Λ ht + ε ijht Γ ht = service offshoring

41 Empirical model ln w ijht = α ij + α t + β 1 Γ ht + θ Θ it + ω Ω jt + λ Λ ht + ε ijht where: Γ ht = service offshoring Θ it = individual characteristics

42 Empirical model ln w ijht = α ij + α t + β 1 Γ ht + θ Θ it + ω Ω jt + λ Λ ht + ε ijht where: Γ ht = service offshoring Θ it = individual characteristics Ω jt =firm characteristics

43 Empirical model ln w ijht = α ij + α t + β 1 Γ ht + θ Θ it + ω Ω jt + λ Λ ht + ε ijht where: Γ ht = service offshoring Θ it = individual characteristics Ω jt =firm characteristics Λ ht = industry characteristics

44 Empirical model ln w ijht = α ij + α t + β 1 Γ ht + θ Θ it + ω Ω jt + λ Λ ht + ε ijht where: Γ ht = service offshoring Θ it = individual characteristics Ω jt =firm characteristics Λ ht = industry characteristics α ij = worker-firm fixed effect

45 Empirical model ln w ijht = α ij + α t + β 1 Γ ht + θ Θ it + ω Ω jt + λ Λ ht + ε ijht where: Γ ht = service offshoring Θ it = individual characteristics Ω jt =firm characteristics Λ ht = industry characteristics α ij = worker-firm fixed effect α t = time effect

46 Results: Service Offshoring and workers wages Benchmark Individual Charact. Firm Charact. Material Offsh. (1) (2) (3) (4) SOS [0.031] [0.029] [0.029] [0.029] age 0.085*** 0.084*** 0.085*** [0.004] [0.004] [0.004] age *** *** *** [0.000] [0.000] [0.000] Occup: Managers ** ** ** [0.053] [0.053] [0.053] Occup: Blue coll ** ** ** [0.060] [0.060] [0.060] Occup: Apprentices *** *** *** [0.052] [0.053] [0.053] Occup: Others [0.085] [0.083] [0.083] Employees [0.004] [0.005] MOS ** [0.001] Observations R-squared Groups The dependent variable is log daily wages. All specifications control for worker-firm and year fixed effects. Standard errors reported in brackets are corrected for clustering within industry. ***, ** and * denote significance at the 1, 5 and 10 percent level, respectively.

47 Results: Service Offshoring and workers wages Industry Charact. Openness Import Penetration Exports & Imports (5) (6) (7) (8) SOS [0.040] [0.038] [0.042] [0.041] Production [0.167] [0.137] [0.150] [0.136] Full-time equivalent [0.135] [0.146] [0.134] [0.146] Capital intensity *** *** *** *** [0.006] [0.007] [0.006] [0.007] Material intensity [0.115] [0.113] [0.120] [0.121] Openness ** [0.011] Import penetration *** [0.010] Exports * [0.011] Imports *** [0.008] Observations R-squared Groups The dependent variable is log daily wages. All specifications control for worker-firm and year fixed effects. Standard errors reported in brackets are corrected for clustering within industry. ***, ** and * denote significance at the 1, 5 and 10 percent level, respectively.

48 Results: Service Offshoring and skill premium Benchmark Individual Charact. Firm Charact. Material Offsh. Industry Charact. Openness Import Exports & Penetration Imports (1) (2) (3) (4) (5) (6) (7) (8) SOS [0.028] [0.030] [0.030] [0.030] [0.041] [0.039] [0.043] [0.041] SOS wc *** * 0.014* 0.013* 0.013* [0.007] [0.008] [0.008] [0.008] [0.007] [0.007] [0.006] [0.006] Observations 497, , , , , , , ,016 R-squared Number of groups 208, , , , , , , ,335 The dependent variable is log daily wages. All specifications control for worker-firm and year fixed effects. Standard errors reported in round brackets are corrected for clustering within industry. ***, ** and * denote significance at the 1, 5 and 10 percent level, respectively.

49 Results: Offshoring of individual services and wages Benchmark Individual Charact. Firm Charact. Material Offsh. Industry Charact. Openness Import Exports & Penetration Imports (1) (2) (3) (4) (5) (6) (7) (8) SOS_FI [0.066] [0.053] [0.052] [0.047] [0.074] [0.074] [0.075] [0.074] SOS_RD [0.113] [0.100] [0.100] [0.087] [0.086] [0.098] [0.093] [0.101] SOS_CI * 0.036* 0.094*** [0.020] [0.020] [0.019] [0.027] [0.028] [0.028] [0.026] [0.029] SOS_BS * * ** ** ** ** [0.078] [0.075] [0.074] [0.069] [0.085] [0.080] [0.087] [0.080] Observations 497, , , , , , , ,016 R-squared Number of groups 208, , , , , , , ,335 The dependent variable is log daily wages. All specifications control for worker-firm and year fixed effects. Standard errors reported in round brackets are corrected for clustering within industry. ***, ** and * denote significance at the 1, 5 and 10 percent level, respectively.

50 Conclusion Service offshoring does not cause significant reductions in workers wages

51 Conclusion Service offshoring does not cause significant reductions in workers wages Service offshoring may contribute to widening the wage gap btw. more and less skilled employees (but small effect)

52 Conclusion Service offshoring does not cause significant reductions in workers wages Service offshoring may contribute to widening the wage gap btw. more and less skilled employees (but small effect) Offshoring of business services exerts negative effects on wages, while offshoring of other services has virtually no impact.

53 Conclusion Service offshoring does not cause significant reductions in workers wages Service offshoring may contribute to widening the wage gap btw. more and less skilled employees (but small effect) Offshoring of business services exerts negative effects on wages, while offshoring of other services has virtually no impact. Implications:

54 Conclusion Service offshoring does not cause significant reductions in workers wages Service offshoring may contribute to widening the wage gap btw. more and less skilled employees (but small effect) Offshoring of business services exerts negative effects on wages, while offshoring of other services has virtually no impact. Implications: Fear of service offshoring (Amiti and Wei, 2005) largely exaggerated

55 Conclusion Service offshoring does not cause significant reductions in workers wages Service offshoring may contribute to widening the wage gap btw. more and less skilled employees (but small effect) Offshoring of business services exerts negative effects on wages, while offshoring of other services has virtually no impact. Implications: Fear of service offshoring (Amiti and Wei, 2005) largely exaggerated Specific government policies may help to mitigate adjustment costs