· Web viewThe firm-level measure of the labor share does not face with the problem of having to separate labor income from self-employed's mixed income. The labor income share is

Wage share and corporate policies in personnel

management:

A firm-level study

Byung-Hee Lee1 and Deok Soon Hwang2

1. Introduction

The share of labor in national income has been declining in most

developed countries since the 1980s. Many explanations has been

suggested to account for the decline in the labor share. They include

technological progress, globalization, institutional and policy changes

and so on. This paper will address the role of corporate policies in

personnel management.

Korea has been experiencing big changes in labor market towards

flexibilization since the 1997 Asian financial crisis. At the same time the

labor share, adjusted for self-employment income, declined by about 10

percentages points in Korea. Increase in non-regular workers and

outsourcing are some of the leading examples of HR policies that

businesses adopted as a way to pursue higher short-term profitability.

This study analyzes the impact of such changes in corporate HR policies

on labor income share.

1 Senior Research Fellow, Korea Labor Institute, e-mail address: [email protected]

2 Senior Research Fellow, Korea Labor Institute, e-mail address: [email protected]

1

In this context, this paper investigates a functional income distribution

at the firm level. It is within individual firms that divide between labor

and capital of economic gains is ultimately decided. A firm-level analysis

is also relatively free from the problem of composition bias that arises by

the shift in employment from labor intensive sectors to capital intensive

sectors. Moreover it also does not require separation of labor income and

capital income from the mixed income of the self-employed (Siegenthaler

and Stucki, 2014). The data used in this study is panel data. The panel

nature of data enables us to resolve the issue of endogeneity.

I explores the determinants on the wage share on the firm level. The

role of technical progress (measured by capital-labor ratio and R&D),

imperfect market competition, union density will be examined. Especial

concern in the paper is the corporate policies in personnel management.

I focus the effect of both the ratio of nonregular work and

subcontracting.

2. Literature Review

The causes of changes in the labor income are often identified as

changes in the industrial structure, technological change, globalization,

the degree of monopoly in the product market, labor and management's

bargaining power, and the increase in the financialization. Empirical

study results vary as to the impact of each factor, which lead to divergent

policy implications.

The Neoclassical economics tend to look to capital-augmenting

technical change and globalization as the main causes of the decline in

2

labor income share in recent decades. Bentolila and Saint-Paul (2003)

concludes that the decline in labor share is attributed to capital

accumulation and capital-augmenting technical progress. Arpaia et al.

(2009) points to technological factors, such as capital-augmenting

technological development and complementarity between capital and

skilled labor. OECD (2012) also concludes that 80% of the fall in the

within-industries labor share of OECD members between 1990-2007 is

attributable to technological development and increase in capital

intensity. The European Commission (2007) demonstrated that skill-

biased technological development is the main cause of labor income

share decline, followed by globalization. If the fall in labor income share

is mainly due to such structural changes, it leaves little room for policy

intervention. What policymakers can do is to run macroeconomic policies

in a way that promotes capital accumulation and technological

advancement and emphasizes training and education for low-skilled

workers who are most negatively affected by the drop in labor income

share (caused by structural changes) (Sang-Heon Lee, 2014).

In contrast, Michal Kalacki and post-Keynesian economists place more

weight on policies and institutions. Stockhammer (2013) and ILO (2012)

highlight the role of institutional changes such as financialization of the

economy and weakening of the labor market institutions and welfare

state. In developed countries, the impact is larger in the order of

financial globalization, institutional changes, globalization and

technological changes. In developing countries, it is in the order of

financialization, globalization and institutional changes while

technological changes actually offset the decline in labor income share

3

because of the catch-up effect. Dunhaupt (2013) demonstrates that

financialization of the economy is the main cause of the fall in labor

income share in OECD members. The spread of shareholder value-

maximizing management ends up putting pressure on wage, by focusing

on short-term profits and lower labor cost and increasing financial

income such as dividends and interest.

Almost all studies used cross-country or industrial data, rarely firm-

level data. Siegenthaler and Stucki (2014) uses panel data on Swiss

companies' innovation activities collected across four waves between

2001-2010 and find that the main factor deceasing labor share is the

increase in the share of workers using ICT in the firm. Growiec (2012)

uses the quarterly panel data of the private sector in Poland between

1995-2008 to identify the determinants of labor income share. He takes

into account such factors as ownership structure, labor market

conditions, market structure and age of the company.

This study firm-level panel data to analyze the impact of each of the

following factors on labor market share: technological factors (capital

intensity, innovation), level of competition in the product market, union

density, use of non-regular workers, and outsourcing.

3. Data for analysis

The data used in this study combines data from Workplace Panel Survey

(WPS) conducted by the Korea Labor Institute with accounting data.

Every two years from 2005 to 2011, the WPS surveyed about 2,000

workplaces with 30 or more full-time workers in all industries with the

4

exception of the agriculture, forestry and fishing sector and the mining

sector. All data from the first to the fourth waves are currently available

from this survey for which the fourth wave panel retention rate was

62.5% and where lost samples were replaced with workplaces of a

similar size that is in a similar industry sector. The analysis in this study

only utilizes data on private sector workplaces. This study analyzes

corporations in the private sector.

For the figures of capital, this study uses accounting data on the year-

end balance of tangible assets, which represent the sum of the value of

land, buildings, machinery and vehicles. For the figures of intermediate

input, accounting data on the cost of sales in used. All variables

representing production input and final output are deflated with the

Production Price Index. Whereas the WPS conducts survey at the

establishment level, financial information used in this study comes from

the level of firms. This discrepancy is reconciles through the

standardization of data. For firms with multiple establishments, the

proportion of workers in a certain establishment out of the total workers

in the firm is multiplied with the variables to convert the financial

information to establishment-level. Meanwhile, the per capital variables

(including per capita value-added) are all results of dividing each

variable with the total number of employees.

Employment types in the WPS are classified as follows. Total employees

is defined as the sum of regular and non-regular workers. Non-regular

workers are either directly or indirectly employed. Directly-employed

non-regular workers include fixed-term contract workers and part-timers.

Indirectly-employed workers include temporary agency workers who are

5

dispatched to workplaces under the Act on the Protection of Temporary

Agency Workers, in-house subcontract and contract company workers

who are hired by subcontractors but who provide labor in the workplace

of the principal contractor who are not entitled to the Act on the

Protection of Temporary Agency Workers, and independent contractors

who provide commissioned labor as self-employed individuals.

The firm-level measure of the labor share does not face with the

problem of having to separate labor income from self-employed's mixed

income. The labor income share is calculated as the percentage of labor

cost out of the value-added at factor cost (i.e. sum of operating surplus,

financial costs and labor cost). Samples with operating loss are excluded

as they show overblown labor income share. In the end, an unbalanced

panel of 4 years is created consisting of 1,579 establishments of only

corporations in the private sector.

The trend in labor income share in this study is compared with that of

Financial Statement Analysis (FSA) conducted by Bank of Korea (see

[Figure 1]). FSA does not allow for understanding of the time-series

trends of the labor income share because of frequent changes in

corporate accounting standards and survey scope. It expanded the

population for sample design in 2007 to include all corporations subject

to corporate income tax, including small businesses. Figure 1 shows an

identical level of labor income share between the two surveys in 2007,

then lower level for the WPS in 2009 and 2011. However the trends are

similar. It dropped in 2009 as corporate financial situation deteriorated

following the global financial crisis then rebounded in 2011, but remains

lower than the pre-crisis level.

6

Figure 1. Trends in the labor share on the firm level

(Unit: %)

2005 2006 2007 2008 2009 2010 2011 201258

60

62

64

66

68

70

72

WPSFSA (2007-10)FSA (2009-12)

Note: There is a time-series interruption in 2009 in the statistics of FSA due to change in the survey method. Source: Workplace Panel Survey and Financial Statement Analysis.

4. Empirical Methodology

The determinants of the labor share will be analyzed using the

estimation equation in Bentolila and Saint-Paul (2003). They assumes a

production function where the elasticity of substitution remains constant.

Qi=¿

In this equation, Qi is the output of firm i, and K i and Li are input of

capital and labor respectively. Ai is capital-augmenting technical change

and Bi is labor-augmenting technical change. ε is the parameter whose

relationship with the elasticity of substitution σ is represented as

σ ≡1/(1−ε). α is the distribution parameter.

7

In a perfectly competitive market, labor demand will be determined at

the point where the price of labor is equivalent to the value of the

marginal product of labor. In this context, the labor income share can be

calculated as follows.

LSi=(1−α)¿¿

In this equation, the closer ε is to 0, the production function converges

to the Cobb-Douglas production function, while the labor income share

converges to 1−α.

They use capital/output ratio in the equation, but this study uses

capital/labor ratio in accordance with the EU Commission (2007). When

capital intensity is k i=K i /Li and technical parameter ratio is T i=Ai/Bi, the

labor share is equal to:

LSi=(1−α )α ¿¿

).

The equation above shows that the labor share is determined by capital

intensity, technical parameter ratio and elasticity of substitution between

labor and capital. How much capital intensity and technical parameter

ratio respectively affect labor share depends on the elasticity of

substitution. When labor and capital are substitutive (ε<0 or σ>1),

increase in capital intensity reduces labor income share. But if they are

complementary (ε>0 or σ<1), it increases the labor income share. If it is

the Cobb-Douglas production function (ε=0 or σ=1), labor income share

remains unchanged despite increase in capital intensity. Meanwhile, the

technical parameter ratio T i changes depending on the nature of

technical change. If labor and capital are substitutive, capital-

augmenting technical change increase T i and reduces labor income

8

share. But labor-augmenting technical change has the opposite effect: it

reduces T i and increases labor income share.

The above assumes that the labor income share is determined by

capital intensity and technological factors such as technical parameter

ratio. But if the product market and labor market are not perfectly

competitive, the institutions that affect competition in the product

market and relative bargaining power in the labor market also affect the

labor income share.

When firms are in imperfect competition in the product market, price is

determined at a higher level than the marginal production cost. If the

ratio between product price (pi) and marginal cost (MC i) is the mark-up

rate (μi=pi /MC i), when firms seek maximum profit in an imperfectly

competitive market, the labor income share is determined as follows:

LSi=g (k i , T i)/ μi. That is, the mark-up rate and labor income share have a

negative correlation.

And the impact of bargaining power of labor to management on labor

income share depends on the bargaining model. In a right-to-manage

model where labor and management negotiate only the wage, and where

the management has the full authority to make employment decisions,

the impact would depend on the elasticity of substitution between

production factors. That is, when workers' bargaining power increases

and succeeds in improving wage, hiring would decrease. But the labor

income share could either rise or fall depending on the elasticity of

substitution between capital and labor. But in an efficient bargaining

model where labor and management negotiate for both wage and

employment, workers' bargaining power is bigger and thus hiring does

9

not decrease with higher wage. If workers' relative bargaining power is θi

, the labor income share is determined as: LSi=θi+(1−θi)g(k i ,T i). This

means that increase in workers' bargaining power brings up the labor

income share regardless of elasticity of substitution.

The estimation model in this paper is as follows:

ln LS¿=βo+β1 ln k¿+β2T¿+ β3 μ¿+β4θ¿+β5θ¿+∑jβ j X¿

j+ηi+λt+ϵ ¿

As for the factors that affect labor income share, capital intensity,

technological factors, and the extent of competition in the product and

labor markets, have been included in accordance with the previous

studies. As for the capital intensity, it was represented using the year-

end value of real per capita tangible assets in the accounting data. WPS

lacks quantitative information on R&D. As for technological factors, this

study uses qualitative information on innovation types. Each innovation

type is given the following dummies: "process improvement with only

production engineering without R&D," "R&D implemented only when

necessary, to the extent of introducing new technologies developed by

other companies," and "leading innovation with R&D." As for the extent

of competition in the product market, dummy variables are generated by

using 1 for very weak or weak competition for the main product and 0 for

others. The extent of competition in the labor market is measured using

the union density rate.

To measure the impact of corporate policies in personnel management

on the labor income share, the main topic of interest of this study, the

corresponding variable (S¿) was added to explanatory variables in the

equation. The proportion of non-regular workers is calculated by the 10

number of non-regular workers by the number of all workers. As for

outsourcing, the value of 1 is given to the principal company in

subcontracting transactions for the firm's main product (that only

contracts out) and 0 for others.

In the estimation model above, ηirepresents the unobservable

heterogeneity of the firm. If such unobserved factor exists, it could lead

to the problem of endogeneity where correlation exists between the

variable of this study's interest and the error term, resulting in

overestimation of the effect of the variable of interest. Unobserved time-

invariant heterogeneity is normally estimated using panel regression

analysis. But fixed-effect model is not only limited to surviving firms but

uses information on within-firm variations only (Siegenthaler and Stucki,

2014). Thus it is not able to account for the inter-firm differences in such

structural and strategic factors as production techniques, market

circumstances and use of labor.

This study attempted to control for the inter-firm heterogeneity by

including the labor income share of the previous term in the pooled OLS

estimation. Including previous term's labor income share not only

controls for part of the endogeneity due to missing variables but also the

endogeneity caused by reverse causation. If a firm's behavior is affected

by past labor income share, previous term's dependent variables can

control for the changes in corporate strategy according to the labor

income share.

But this does not fully address the problem of endogeneity. Labor

income share, investment and choice in technical factors such as R&D

can still be associated with firms' unobserved heterogeneity. Unobserved

11

demand shock or productivity change can affect labor income share and

explanatory variables, creating false correlation. Levinsohn and Petrin

(2003) focuses on firms' practice of first adjusting the intermediate input

when faced with an unobserved exogenous shock. Intermediate goods

are not included in calculation of the value-added, and thus have no

direct impact on the share of labor or capital. The ratio of intermediate

input controls for the variable heterogeneity and helps reduce the

simultaneity bias. In this study, the share of manufacturing cost out of

value-added is included in the estimation model.

In accordance to the analysis by Growiec (2012) concluding that the

characteristics of the business establishment also affect the labor income

share, the log value of total employees, age of the establishment and

industry (manufacturing or not) are included. Year dummy has also been

added to control for the effect of economic cycles.

Basic statistics as of 2011 for these variables are presented in <Table

1>. As of 2011, business establishments that use non-regular workers

account for 68.5% of the total, and their labor income share, at 64.7%, is

lower than that of the establishments that do not have non-regular

employees (69.5%). Those that use non-regular labor are found to be

relatively high in capital intensity, R&D intensity, number of employees,

union organization rate, share of principal companies and company age,

with less competition in the market.

Meanwhile, 14.7% of total business establishments were found to be

principal companies that only contract out in subcontracting

transactions. By type of subcontracting, principal companies have much

lower labor income share than subcontractors or those who do not

12

engage in subcontracting transactions.

Table 1. Sample Characteristics of 2011

Non-regular workers Outsorucing

In use Not in use Principal Others

Mean (SD) Mean (SD) Mean (SD) Mean (SD)

Labor income share 0.647 (0.23

2)0.695 (0.220) 0.597

(0.247)

0.673 (0.22

4)

Log (per capita capital) 4.074 (2.03

4)3.761 (2.247) 4.715

(1.874)

3.847 (2.12

0)

No R&D 0.077 (2.03

4)0.116 (0.321) 0.029

(0.169)

0.100 (0.30

0)Process improvement with no R&D

0.200 (0.26

7)0.164 (0.371) 0.117

(0.322)

0.201 (0.40

1)

R&D only when needed 0.230 (0.40

0)0.215 (0.412) 0.248

(0.434)

0.221 (0.41

5)Leading innovation with R&D

0.494 (0.42

1)0.505 (0.501) 0.606

(0.490)

0.479 (0.50

0)Weak competition in market

0.046(0.20

9)0.041 (0.199) 0.051

(0.221)

0.043 (0.20

3)

Union density rate 0.222 (0.31

6)0.151 (0.292) 0.278

(0.331)

0.186 (0.30

4)

Non-regular workers (%) 0.226 (0.23

7)0 0.196

(0.227)

0.148 (0.22

1)

Principal (contracting out) 0.170 (0.37

6)0.099 (0.299) 1 0

Log (# of employees) 5.575 (1.21

4)4.757 (1.195) 5.809

(1.276)

5.232 (1.24

5)

Company age 25.49 (16.0

0)22.32 (13.19) 28.78

(16.15)

23.75 (14.9

6)

Manufacturing 0.616 (0.48

7)0.683 (0.466) 0.708

(0.456)

0.625 (0.48

4)

13

Sample size 636 293 137 792

% of applicable companies 0.685 0.315 0.147 0.853

5. Estimation Result

<Table 2> shows estimation results of determinants on labor income

share. Column (1) includes intermediate input to control for firms' time-

variant heterogeneity, while Column (2) additionally includes previous

term's labor income share.

Column (1) shows, firstly, capital intensity has a significant negative

effect on labor share. According to the discussion above, this result

implies that labor has substitutive relationship with capital. Secondly,

when labor and capital are mutually substitutive, the impact on labor

income share depends on the nature of technical change. Although it was

not presented separately, the "innovation-leading" companies stating that

innovation is the key to their competition strategy have a significant

positive correlation with capital intensity. Innovation-leading type is

likely to cause capital-augmenting technical change. The hypothesis that

capital-augmenting technical change will pull down the labor income

share if labor and capital are substitutive is supported in the result.

Innovation-leading type has a significantly lower labor income share than

other types. Thirdly, firms with less market competition have significantly

lower labor income share than others. In a product market with

imperfect competition, firms are likely to set the price at higher than the

production cost. Fourthly, it was also found that the higher the union

density rate, the higher the labor income share. This is so despite the

substitutive relationship between labor and capital because employment 14

does not go down as wage rises, or goes down only by a minimal extent.

Meanwhile, in terms of the impact from corporate HR policies, the topic

of interest of this study, it was found that the higher the proportion of

non-regular workers, the lower the labor share. The definition of "non-

regular workers" in this study also includes indirectly-employed workers,

but their cash and value-in-kind compensation is not included in the

firms' payroll. Although non-regular workers' productivity and wage

cannot be compared, it can be interpreted that the use of non-regular

labor reduces the labor income share not because it ends up distributing

more of the value-added to regular workers but because it helps increase

the share of capital income. In terms of principal companies' labor

income share, it is significantly lower than subcontractors or those that

do not engage in subcontracting. If subcontracting has unfair conditions,

more of the benefits will belong to principal companies, but the fact that

labor income share is lower even when other factors are controlled for

indicates that more of them are accrued to capital income rather than the

rent distribution within the principal company.

As for impact of other control variables, it was found that the larger the

company size, the lower the labor income share. Although it was not

possible to discern any difference between new businesses and

continuing establishments due to limitations of panel data, the age of

company does not appear to have a significant impact. In terms of the

year effect, labor income share was much lower in 2009 compared to

2005, and was also significantly lower in 2011.

Estimation Result of column (2) where the previous term's log labor

income share has been added to control for endogeneity shows that

15

although the significant of the estimation coefficients fell, the sign is

largely similar. The estimation result where the impact of previous term's

labor income share was found to be large implies that the unobserved

firm characteristics are large.

Table 2. Estimation of determinant of labor share in the firm level

(Pooled OLS)

(1) (2)

Log(capital intensity) -0.089 (0.004) *** -0.029 (0.005) ***

Process improvement without R&D

-0.042 (0.035) -0.036 (0.038)

R&D only when needed -0.039 (0.034) -0.026 (0.036)

Leading innovation with R&D -0.084 (0.033) ** -0.062 (0.035) *

Market competition is weak -0.149 (0.034) *** -0.054 (0.036)

Union density rate 0.136 (0.028) *** -0.009 (0.028)

Non-regular workers % -0.173 (0.039) *** -0.095 (0.040) **

Principal (contracting out) -0.041 (0.020) ** -0.025 (0.021)

Share of manufacturing cost 0.001 (0.003) 0.008 (0.003) **

Share of manufacturing cost, squared

0.000 (0.000) 0.000 (0.000)

Log (# of employees) -0.024 (0.007) *** -0.009 (0.007)

Age of workplace 0.000 (0.001) 0.000 (0.001)

Manufacturing 0.018 (0.017) -0.049 (0.018) ***

2007 0.007 (0.022)

2009 -0.079 (0.022) *** -0.107 (0.020) ***

2011 -0.051 (0.022) ** -0.004 (0.020)

Constant 0.083 (0.048) * 0.093 (0.050) *

Lag of dependent variable 0.693 (0.017) ***

Adj R-squared 0.135 0.517

N 3,621 2,110

16

Notes: 1) The reference group has the following characteristics: weak market competition, not a principal company (i.e., subcontractor or no subcontracting), service sector, 2005.

2) Standard errors are in parentheses. 3) * significant at 10% level; ** significant at 5% level; *** significant at

1% level.

To control for the unobserved firm characteristics, a number of

different models have been additionally estimated. The results are

presented in <Table 3> for comparison. Results of the random effect

model that uses both inter-firm and intra-firm information are similar to

those of the pooled OLS, but quite different from those of the fixed-effect

model where only intra-firm information is used. In the fixed-effect

model, impact of the variables of interest is not significant, except for

that of R&D/innovation. But the fixed-effect model is preferred because

the Hausman test cannot reject the null hypothesis that the covariance

between unobserved firm characteristics and explanatory variables is 0.

This also shows the importance of controlling for endogeneity. But if it is

important to account for the structural and strategic differences between

firms, pooled OLS or random effect model results would be better

suitable in identifying the determinants of labor income share.

In terms of the effect of corporate policies in personnel management,

pooled OLS or random effect model shows that 10%p increase in non-

regular workers reduces labor income share by 0.9%. And a principal

company's labor income share is 0.03%p lower than that of

subcontractors or those that do not engage in contracting.

Table 3. Estimation of determinant of labor share in the firm level 17

(Pooled OLS, REM & FEM model)

(2)Pooled OLS

(3)REM

(4)FEM

Log(capital intensity) -0.029 (0.005)*** -0.065 (0.005)*** 0.008 (0.00

8)Process improvement without R&D

-0.036 (0.038) -0.062 (0.029)** -0.061 (0.03

2)*

R&D only when needed -0.026 (0.036) -0.047 (0.028)* -0.045 (0.03

1)Leading innovation with R&D

-0.062 (0.035)* -0.083 (0.027)*** -0.069 (0.03

0)**

Market competition is weak -0.054 (0.036) -0.086 (0.030)*** -0.026 (0.03

2)

Union density rate -0.009 (0.028) 0.040 (0.032) -0.047 (0.06

1)

Non-regular worker % -0.095 (0.040)** -0.089 (0.034)*** 0.046 (0.04

1)

Principal (contracting out) -0.025 (0.021) -0.032 (0.016)** -0.017 (0.01

7)

Share of manufacturing cost 0.008 (0.003)** 0.023 (0.004)*** 0.074 (0.00

6)***

Share of manufacturing cost, squared

0.000 (0.000) 0.000 (0.000)*** -0.001 (0.00

0)***

Log (# of employees) -0.009 (0.007)Age of firms 0.000 (0.001)Manufacturing -0.049 (0.018)***

2007 0.008 (0.015) -0.004 (0.01

5)

2009 -0.107 (0.020)*** -0.096 (0.015)*** -0.129 (0.01

6)***

2011 -0.004 (0.020) -0.070 (0.016)*** -0.108 (0.01

6)***

Constant 0.093 (0.050)* -0.169 (0.034)*** -0.637 (0.04 **

18

9) *

Lag of dependent variable 0.693 (0.017)***

Adj R-squared 0.517 Wald chi2 296.66 ***

F value 21.52 ***N 2,110 3,621 3,621

Notes: 1) The reference group has the following characteristics: weak market competition, not a principal company (i.e., subcontractor or no subcontracting), service sector, 2005.

2) Standard errors are in parentheses. 3) * significant at 10% level; ** significant at 5% level; *** significant at

1% level.

6. Summary and Policy Implications

This paper analyzed the determinants of labor income share at the firm

level. Main results are summarized as follows. Labor income share in a

firm is determined not only by technical factor but also other factors such

as the degree of monopoly in the product market, employees' bargaining

power and corporate labor strategy. Capital intensity and R&D have a

negative effect on labor income share. In comparison, the weaker the

competition in the product market, the lower the labor income share, and

the higher the union density rate, the higher the labor income share. And

the higher the non-regular employee share, the lower the labor income

share. Principal companies that only contract out are found to have a

lower labor income share.

Based on such analysis, this paper focuses on the following

implications.

First, the analysis result that higher share of non-regular employees 19

leads to lower labor income share implies that even if pecuniary gain is

created from utilizing non-regular labor, it is accrued to capital income,

not to under regular workers. It also means that there should be policy

efforts to ensure fair distribution of the fruits of production between

labor and capital, going beyond simply trying to ease inequality in the

labor income.

Second, , the analysis result that labor share for principal company in

subcontracting transactions is higher than others even when various

factors are controlled for shows that the benefits of subcontracting

transactions are attributed to the principal company's capital income.

This implies that efforts to improve unfair subcontracting transactions

must be undertaken to ensure fair distribution.

References

Arpaia, A., E. Pérez and K. Pichelmann (2009), "Understanding Labour Income

Share Dynamics in Europe", Economic Papers 379, Directorate General

Economic and Monetary Affairs, European Commission.

Bentolila, S. and G. Saint-Paul (2003), "Explaining movements in the labor

share", Contributions to Macroeconomics, 3(1), 1–3.

Dunhaupt, P. (2013), “The effect of financialization on labor’s share of income”,

Working Paper No. 17/2013, Berlin: Institute for International Political

Economy, Berlin School of Economics and Law.

European Commission (2007), "The labour income share in the European

Union", Employment in Europe, 237–272. Luxembourg: Office for Official

Publications of the European Communities.

20

Growiec, J. (2012), "Determinants of the Labor Share: Evidence from a Panel of

Firms", Eastern European Economics, 50(5), 23-65.

ILO (2012), Global Wage Report 2012/13: Wages and Equitable Growth,

Geneva: ILO.

Levinsohn, J. and A. Petrin (2003), "Estimating Production Functions Using

Inputs to Control for Unobservables," Review of Economic Studies, 70(2),

317-341.

OECD (2012), "Labour Losing to Capital: What Explains the Declining Labour

Share?", Employment Outlook, Paris: OECD.

Siegenthaler, M. and T. Stucki (2014). "Dividing the Pie: the Determinants of

Labor’s Share of Income on the Firm Level", KOF Working Paper No. 352.

Stockhammer, E. (2013), "Why Have Wage Shares Fallen? An analysis of the

Determinants of Functional Income Distribution", Mark Lavoie and

Engelbert Stockhammer, Wage-led Growth: An Equitable Strategy for

Economic Recovery, Palgrave macmillan and the ILO.

21