How Does the Minimum Wage Affect Wage Inequality and Firm Investments in Fixed and Human Capital? Evidence from China

Transcription

1 How Does the Minimum Wage Affect Wage Inequality and Firm Investments in Fixed and Human Capital? Evidence from China Tobias Haepp and Carl Lin National Taiwan University & Chung-Hua Institution for Economic Research Beijing Normal University & Institute for the Study of Labor 4th Conference on Regulating for Decent Work International Labour Office Geneva, Switzerland July 9, 2015

2 (I)

3 (I) China has introduced a complex system of minimum wages ILO Minimum Wage Fixing Convention ratified in Enterprises Minimum Wage Regulations passed in New Minimum Wage Regulations passed in 2004.

4 (I) China has introduced a complex system of minimum wages ILO Minimum Wage Fixing Convention ratified in Enterprises Minimum Wage Regulations passed in New Minimum Wage Regulations passed in Previous research on Chinese MW has focused on labor market indicators: Employment of female workers (Wang and Gunderson, 2012; Jia, 2014) Employment of young adults and low-skilled workers (Fang and Lin, 2013) Labor productivity (Mayneris et al., 2014)

5 (II)

6 (II) of MW in China has been accompanied by intense debate Proponents argue that minimum wages help to provide worker protection induce incentives for production upgrading

7 (II) of MW in China has been accompanied by intense debate Proponents argue that minimum wages help to provide worker protection induce incentives for production upgrading Opponents argue that minimum wages raise production costs harm firm competitiveness

8 (II) of MW in China has been accompanied by intense debate Proponents argue that minimum wages help to provide worker protection induce incentives for production upgrading Opponents argue that minimum wages raise production costs harm firm competitiveness What is the effect on firm investments? Key channel for skill-upgrading and long-term firm competitiveness

9 Minimum wages and firm investments:

10 Minimum wages and firm investments: Competitive labor markets: fixed capital investment Negative scale effect due to increase in labor costs. Substitution or complementarity effect between the two factors of production.

11 Minimum wages and firm investments: Competitive labor markets: fixed capital investment Negative scale effect due to increase in labor costs. Substitution or complementarity effect between the two factors of production. Competitive labor markets: human capital investment Workers accept lower wages to finance their training. Introducing a lower bound on wages reduces ability to finance training.

12 Minimum wages and firm investments: Competitive labor markets: fixed capital investment Negative scale effect due to increase in labor costs. Substitution or complementarity effect between the two factors of production. Competitive labor markets: human capital investment Workers accept lower wages to finance their training. Introducing a lower bound on wages reduces ability to finance training. Non-competitive labor markets: fixed & human capital investment Wages are below marginal productivity. When labor costs rise through minimum wages, employers raise investments to maintain profit margins.

13 Minimum wages and firm investments:

14 Minimum wages and firm investments: Fixed capital investments: mixed evidence Negative effect in Indonesia (Rama, 1999) Positive effect in EU, but not in US (Pischke, 2005) No effect in the UK (Riley and Bondibene, 2013)

15 Minimum wages and firm investments: Fixed capital investments: mixed evidence Negative effect in Indonesia (Rama, 1999) Positive effect in EU, but not in US (Pischke, 2005) No effect in the UK (Riley and Bondibene, 2013) Human capital investments: mostly no effects Negative effect (Neumark and Wascher, 2001) No effect in US (Acemoglu and Pischke, 2003; Fairris and Pedace, 2004) No effect in UK (Arulampalam et al., 2004)

16 Minimum wages and firm investments: Fixed capital investments: mixed evidence Negative effect in Indonesia (Rama, 1999) Positive effect in EU, but not in US (Pischke, 2005) No effect in the UK (Riley and Bondibene, 2013) Human capital investments: mostly no effects Negative effect (Neumark and Wascher, 2001) No effect in US (Acemoglu and Pischke, 2003; Fairris and Pedace, 2004) No effect in UK (Arulampalam et al., 2004) Key characteristics of previous studies Mostly from advanced economies. Most studies analyze cross-section data and/or one-time increase in MW.

17 sources: minimum wage and firm data from China

18 sources: minimum wage and firm data from China Chinese minimum wage data Minimum wages differ even at the county level. Our data set contains minimum wage levels in 2605 counties over time.

19 sources: minimum wage and firm data from China Chinese minimum wage data Minimum wages differ even at the county level. Our data set contains minimum wage levels in 2605 counties over time. China Annual Census of Industrial Firms (CASIF) Implemented by National Bureau of Statistics Contains all public enterprises Contains all other enterprises with revenue above RMB 5 million CASIF firms contribute about 90% of Chinese industrial output Years included: 2000 until 2007 (i.e. years surrounding the introduction of new Chinese minimum regulations)

20 National minimum wage and incidence of increases over time

21 Provincial minimum wages over time (in RMB per month) Province East Beijing Fujian Guangdong Hainan Hebei Jiangsu Shandong Shanghai Tianjin Zhejiang Northeast Heilongjiang Jilin Liaoning Central Anhui Henan Hubei Hunan Jiangxi Shanxi West Chongqing Gansu Guangxi Guizhou Inner Mongolia Ningxia Qinghai Shaanxi Sichuan Xinjiang Yunnan Notes: Minimum wages have been calculated as time-weighted and population-weighted average values based on county level minimum wage data. Values have been deflated to the price level in year 2000.

22 Mean Std. Dev. Min. Max. Obs Fixed capital investment Human capital investment Min. wage / Avg. wage Fixed capital investment Human capital investment Min. wage / Avg. wage Fixed capital investment Human capital investment Min. wage / Avg. wage Fixed capital investment Human capital investment Min. wage / Avg. wage Fixed capital investment Human capital investment Min. wage / Avg. wage Fixed capital investment Human capital investment Min. wage / Avg. wage Summary statistics for key variables over time Variable Fixed capital investment Human capital investment Min. wage / Avg. wage Note: have been deflated to the price level in year The unit of measurement for training expenditures per worker is thousand RMB.

23 (I)

24 (I) Use firm-level average wages to identify affected firms (e.g. Draca et al. (2011))

25 (I) Use firm-level average wages to identify affected firms (e.g. Draca et al. (2011)) Split firms into treatment and control group 0 treatdumit = 1 if awit mwjt if awit < mwjt (1)

26 (I) Use firm-level average wages to identify affected firms (e.g. Draca et al. (2011)) Split firms into treatment and control group 0 treatdumit = 1 if awit mwjt if awit < mwjt (1) Split firms & measure intensity of treatment 0 treatintit = mwjt awit if awit mwjt if awit < mwjt (2)

27 Change in log wage for treated vs. non-treated companies Quantile Control group Treatment group Notes: The average wage of companies in the 1st and 20th wage quantile is RMB and RMB, respectively. The treatment groups shrink to less than thirty companies above the 20th quantile and the remaining quantiles have therefore been omitted.

28 Change in log wage for treated vs. non-treated companies by ownership type Quantile Control LOEs Treatment Control SOEs Treatment NMCOEs Control Treatment Control FOEs Treatment Notes: The average wage of companies in the 1st and 20th wage quantile is RMB and RMB, respectively. LOEs refers to local private or collective firms, SOEs refers to state-owned enterprises, the non-mainland Chinese owned enterprises (NMCOEs) are those owned by investors from either Hong Kong, Macao or Taiwan and FOEs are foreign-owned enterprises. The treatment groups shrink to less than thirty companies above the 10th quantile in the state-owned sector and the remaining quantiles have therefore been omitted for all ownership types.

29 Impact on inter-firm wage inequality

30 Impact on inter-firm wage inequality Focus on pre- vs. post-reform year (2003 vs. 2005)

31 Impact on inter-firm wage inequality Focus on pre- vs. post-reform year (2003 vs. 2005) Calculate the share of affected firms by county

32 Impact on inter-firm wage inequality Focus on pre- vs. post-reform year (2003 vs. 2005) Calculate the share of affected firms by county Calculate changes in inequality measures by county Gini coefficient 10pc/50pc ratio

33 Impact on inter-firm wage inequality

34 (II)

35 (II) Employ our instruments to estimate the effect of the MW policy on investments

36 (II) Employ our instruments to estimate the effect of the MW policy on investments Fixed capital investments: Error-correction model (GMM methodology) Iit = α treati,t 1 + β0 ki,t 1 + β1 yi,t + β2 yi,t 1 + β3 (ki,t 2 yi,t 2 ) εi,t (3)

37 (II) Employ our instruments to estimate the effect of the MW policy on investments Fixed capital investments: Error-correction model (GMM methodology) Iit = α treati,t 1 + β0 ki,t 1 + β1 yi,t + β2 yi,t 1 + β3 (ki,t 2 yi,t 2 ) εi,t (3) Human capital investments: Censored regression models (Logit and Tobit) Hit = α treati,t 1 + β x, + εi,t i t (5)

38 Fixed capital investment regression results GMM models (1) (2) Treatment dummy (lag) (0.005) (0.013) Investment rate (lag) (0.220) (0.220) Change in log output (0.083) (0.083) Change in log output (lag) (0.048) (0.048) Error correction term (0.047) (0.048) Profit per capital (0.007) (0.007) Profit per capital (lag) (0.006) (0.006) Debt per capital (0.000) (0.000) Debt per capital (lag) (0.004) (0.004) Average wage (0.000) (0.000) Treatment intensity (lag) Observations Firms Number of instruments Hansen Test (p-value) AR(1) (p-value) AR(2) (p-value) Notes: Standard errors in parentheses. The respective significance symbols denote: p < 0.10, p < 0.05, p < 0.01.

39 Human capital investment regression results Logit models (3) (4) Treatment dummy (lag) Treatment intensity (lag) Average wage Log workforce size Labor productivity State owned Foreign owned Exporter dummy Tobit models (5) (6) (0.002) (0.018) (0.000) (0.023) (0.112) (0.037) (0.031) (0.015) (0.055) (0.000) (0.026) (0.125) (0.034) (0.025) (0.014) Union dummy Technical staff (%) University degree (%) Female staff (%) Constant Firm fixed effects Industry fixed effects Observations Log likelihood Chi2 Prob Chi2 > 0 σu σe ρ (0.000) (0.001) (0.006) (0.002) (0.002) (0.001) (0.001) (0.018) (0.009) (0.006) (0.005) (0.005) (0.000) (0.001) (0.006) (0.002) (0.002) (0.001) (0.001) (0.018) (0.009) (0.006) (0.005) Notes: Bootstrapped standard errors are shown in parentheses. The significance symbols denote: p < 0.10, p < 0.05, p < 0.01.

40 Fixed capital investment regression results by firm ownership type LOEs GMM models SOEs NMCOEs (0.008) (0.007) Investment rate (lag) (0.271) (0.239) Change in log output (0.097) (0.126) Change in log output (lag) (0.054) (0.072) Error correction term (0.053) (0.080) Average wage (0.000) (0.000) Treatment dummy (lag) Observations Groups Number of instruments Hansen Test (p-value) AR(1) (p-value) AR(2) (p-value) (0.008) (0.153) (0.096) (0.049) (0.051) (0.000) FOEs (0.011) (0.260) (0.111) (0.058) (0.058) (0.000) Notes: Standard errors are shown in parentheses. The respective significance symbols denote: p < 0.10, p < 0.05, p < 0.01.

41 Human capital investment regression results by firm ownership type Logit models SOEs NMCOEs LOEs Treatment dummy (lag) (0.023) Average wage (0.000) Log workforce size (0.031) Labor productivity (0.158) Exporter dummy (0.021) Union dummy (0.087) (0.000) (0.110) (0.422) (0.089) (0.066) (0.000) (0.055) (0.485) (0.041) FOEs Technical staff (%) University degree (%) Female staff (%) Firm fixed effects Industry fixed effects Observations Log likelihood Chi2 Prob Chi2 > 0 σu σe ρ FOEs (0.002) (0.000) (0.001) (0.008) (0.001) (0.001) (0.021) (0.011) (0.007) (0.004) (0.000) (0.001) (0.011) (0.003) (0.004) (0.038) (0.019) (0.018) (0.006) (0.000) (0.002) (0.023) (0.003) (0.003) (0.058) (0.026) (0.018) (0.007) (0.000) (0.002) (0.021) (0.002) (0.004) (0.055) (0.021) (0.018) Notes: Bootstrapped standard errors are shown in parentheses. The significance symbols denote: (significant) constant term has been omitted to save space. Tobit models SOEs NMCOEs LOEs (0.093) (0.000) (0.061) (0.451) (0.033) p < 0.10, p < 0.05, p < The

42

43 Minimum wage treatment triggers wage growth in affected firms

44 Minimum wage treatment triggers wage growth in affected firms This results in a reduction in inter-firm wage inequality

45 Minimum wage treatment triggers wage growth in affected firms This results in a reduction in inter-firm wage inequality All firms decrease their human capital investments in response to the minimum wage

46 Minimum wage treatment triggers wage growth in affected firms This results in a reduction in inter-firm wage inequality All firms decrease their human capital investments in response to the minimum wage Chinese private firms increase their fixed capital investments in response to the minimum wage

47 Minimum wage treatment triggers wage growth in affected firms This results in a reduction in inter-firm wage inequality All firms decrease their human capital investments in response to the minimum wage Chinese private firms increase their fixed capital investments in response to the minimum wage Our investment results can be explained by neoclassical theories of the labor market

48 Thank you for your attention! Comments and suggestions are most welcome: