Management International Review

, Volume 53, Issue 5, pp 741–762

Management Compensation Systems in MNCs and Domestic Firms

Cross-National Empirical Evidence

Authors

    • Management SchoolUniversity of Sheffield
  • Chris Brewster
    • Henley Business SchoolUniversity of Reading
  • Mehmet Demirbag
    • Management SchoolUniversity of Sheffield
  • Geoffrey Wood
    • Warwick Business SchoolWarwick University
Research Article

DOI: 10.1007/s11575-013-0175-2

Cite this article as:
Le, H., Brewster, C., Demirbag, M. et al. Manag Int Rev (2013) 53: 741. doi:10.1007/s11575-013-0175-2

Abstract

  • This is a study of the relationship between institutional settings and managerial compensation systems, based on extensive cross-national survey evidence.

  • We compare differences in practices between Multinational Corporations (MNCs) and domestic firms across a range of capitalist archetypes.

  • We find that MNCs are more likely to promote compensation systems that incentivise managers in line with organisational performance compared to domestic firms. Our findings also reveal persistent diversity reflecting firm type and institutional setting. We find that the gap between MNCs and domestic firms in terms of the usage of incentive-related compensation is less pronounced in Liberal Market Economies than in other settings. This suggests that it is a combination of being an MNC and the specific home locale that moulds approaches to managerial compensation. This reflects considerable hybridisation of practices within and between settings.

Keywords:

Management compensationMNCsShareholder rightsInstitutional setting

Introduction

This paper examines the relationship between institutional settings and managerial compensation systems based on extensive cross-national survey evidence. We compare differences in practices between Multinational Corporations (MNCs) and their domestic counterparts across a range of capitalist archetypes. Our focus is the link between a firm’s geographical diversification strategies and two dimensions of managerial compensation, namely monetary and non-monetary rewards. Contingent compensation that rewards managers in line with overall organisational financial performance (such as stock options) plays an important role in aligning the interests of managers and shareholders within liberal market economies (LMEs) (Dore 2000; Erturk et al. 2008; Froud et al. 2006). In contrast, within coordinated market economies (CMEs), the incentives under which managers operate are more diffuse, and encompass the ability to develop and maintain relationships with stakeholders and industry associations (Dore 2000). We explore whether MNCs differ from their domestic counterparts in terms of using different compensation mechanisms and whether the difference, if any, varies with the institutional settings of the MNCs’ countries of origin. To do this we use a unique dataset, which contains information about domestic and multinational firms operating in 18 countries. Our sample is considerably larger and more comparative than those used in previous studies of management compensation, which mainly focus on developed economies such as Canada, Germany, Japan, the UK, or the USA (Crossland and Hambrick 2007; Fey and Furu 2008; Oxelheim and Randoy 2005; Southam and Sapp 2010). In contrast our analysis is based on a dataset that covers a range of institutional settings, encompassing developed, emerging and transitional countries in or in close proximity to Europe.

Our contribution to the literature is three-fold. First, we find that firms may compensate for the challenges posed by institutional settings in a manner that has the potential to make for complementarities rather than simply additional inefficiencies or costs (Gordon and Roe 2004, p. 16). Second, we find further evidence on the consequences of spatial dispersion on the compensation strategies by MNCs. Third, we find that firms rarely practice “pure” forms of “calculative” or “co-operative” human resources management, contrary to what is suggested in much of the literature (Carpenter et al. 2001); rather, firms are likely to “mix and match” interventions in response to the pressures posed by locale and firm type. The rest of the paper is structured as follows. The next section provides a review of related literature and outlines our hypotheses. Section 3 describes our methods and data. Section 4 presents our empirical results. Section 5 discusses the results and their implication. Section 6 summaries and concludes.

Related Literature and Hypotheses Development

Institutional Settings, Geographic Diversification and Management Compensation

Diversification, industrial and geographical, increases firm complexity and hence the information-processing task that the management faces. According to Fama and Jensen (1983) it is costly to transfer specific information relevant to decisions between internal agents in complex organisations. This proposition has two important implications. Firstly, managers in complex firms need to gather and process more information for their decisions than their counterparts in less complex firms. Secondly, the demands of information-processing increase with firm complexity. Diversification is an important contributor to firm complexity, increasing the information-processing task that the management faces (Finkelstein and Hambrick 1989; Musteen et al. 2009) especially due to divisional interdependence in related-diversified firms and internal capital markets in unrelated-diversified firms (Jones and Hill 1988). Henderson and Fredrickson (1996) report that compensation is indeed positively related to the information-processing demands resulted from firm’s diversification scope.

Existing literature suggests that geographical diversification may influence the management compensation practice differently compared to industrial diversification. Roth and O’Donnell (1996) point out that geographical diversification offers some unique aspects of firm complexity (see also Gomez-Mejia and Palich 1997). Managers of geographically diversified firms, i.e. MNCs, are expected to have different skills (e.g., language skills and wider global knowledge) compared to managers of domestic firms (Oxelheim and Randoy 2005). Given the complex environment in which MNCs operate, managers also have increased responsibility (Oxelheim and Randoy 2005; Carpenter and Sanders 2004). Further Nohria and Ghoshal (1994) argue that geographical diversification heightens the ambiguity surrounding management’s actions (Fey and Furu 2008; Kim et al. 2005). Hence, a compensation premium and a stronger emphasis on contingent pay may be necessary as internationalisation increases (Wiseman and Gomez-Meija 1998; Oxelheim and Randoy 2005), and will be enhanced as the skills and competences required to manage international firms become scarcer (Oxelheim and Randoy 2005).

A particularly important distinction between the management compensation mechanisms used by MNCs and domestic firms is rooted in the assertion that the former is moulded by the institutional settings of both home and host countries while the latter is under the influence of the home country institutional setting only (Filatotchev and Wright 2011; Oxelheim and Randoy 2005; Schuler and Rogovsky 1998). In comparing institutional settings, an influential school of thinking has been the comparative capitalisms literature (see Jackson and Deeg 2008). Here Dore (2000) draws a key distinction between the shareholder dominated varieties of capitalism, where managers are closely reined in and have to maximise returns, and the more stakeholder orientated varieties of capitalism, where the interests of workers and owners are systematically mediated with a stronger emphasis on retention and investment. In a landmark 2001 collection, Hall and Soskice (2001) define the former as Liberal Market Economies (LMEs), examples being the USA and the UK, and the latter as Coordinated Market Economies (CMEs), examples being Germany, Japan and Scandinavia. Significantly, Hall and Soskice (2001) hold that both models are equally viable, with each context having its own complementary sets of rules and practices. Hall and Soskice (2001) also acknowledge that many developed countries fit into neither category and that other societies may persistently combine features of both such as the emerging market economies of post-communist Central and Eastern Europe and the mixed market economies of Mediterranean Europe (Hancke et al. 2007)1

Within the context of LMEs, management compensation, more specifically rewards linked to overall financial performance via share-based incentives, is considered as an important corporate governance mechanism to align the interests of managers with those of shareholders in order to mitigate agency problems (see e.g., Jensen and Meckling 1976; Murphy 1999; Froud et al. 2006; Dore 2000). From this shareholder value perspective, managerial compensation is seen as a powerful incentive mechanism for the governance of MNCs’ cross-border operations, ensuring that MNCs originated from LMEs do indeed maintain the shareholder value agenda (Filatotchev and Wright 2011; Tihanyi et al. 2009; Sanders and Carpenter 1998; Jensen and Murphy 1990). Gomez-Mejia (1992) argues that management compensation should be designed to ensure that the extent of a firm’s diversification really improves its performance. This is because other mechanisms to reduce agency problems such as monitoring or bonding of managers may be more difficult to implement in MNCs due to ownership in different cultures (Sanders and Carpenter 1998; Tihanyi et al. 2009). Indeed Sanders and Carpenter (1998) find that compensation is linked to more severe agency problems between shareholders and managers arising from a firm’s geographical dispersion of its production.

In contrast, in CMEs managers are less likely to operate under such incentives and, hence, will devote more attention to building relations with a wide range of stakeholders. This reflects a very different historical understanding of the firm as an entity that has dense and rich ties to a wide range of stakeholders, and has obligations to the industry and society at large (Hall and Soskice 2001; Whitley 1999). Hence, managers would be seen more as stewards of an entity that has a wider societal purpose in its own right, rather than simply as a vehicle for the enrichment of shareholders (Hall and Soskice 2001; Whitley 1999). As shareholder rights are weaker, reining in managers to carry out a narrow brief as agents of the former alone would be seen as undesirable; rather, the focus would be on encouraging managerial stewardship and longer term commitment (Hall and Soskice 2001; Whitley 1999). In turn, this would suggest that compensation would be less contingent, and above all, less narrowly orientated towards share price performance.

However, developments in work and employment relations point to both the persistence of these differences and their erosion (Kaufman 2011, p. 34). Sanders and Tuschke (2007), for example, report that German firms that are exposed to multiple institutional environments, more specifically that of the USA, are more likely to adopt stock-based pay practices. In the run up to the 2008 financial crisis, the relatively strong performance of the USA and the UK vis-à-vis Germany led to suggestions that the LME model would ultimately triumph on performance grounds, with work and employment relations standards being driven down (Kaufman 2011, p. 40); arguably, a feature of this has already been a general move to more conditional and contingent terms and conditions of tenure and compensation. Jessop (2012) argues that within global capitalism, differences in capitalist archetypes notwithstanding, neo-liberalism represents the dominant ecosystem, and therefore, LME-type practices are increasingly likely to be encountered elsewhere. Hence, it could be argued that those firms that are most embedded in the global economy are most likely to be prone to such diffusion. Kaufman (2011, p. 34) argues that MNCs may act as agents of standardisation or “norm entrepreneurs”, and hence, it could be argued that compensation systems within MNCs are more likely to approximate ones commonly encountered in LMEs (see also Dore 2008). Gooderham et al. (2006) report that US MNCs tend to use incentive-based compensation, along with other calculative LME-type practices, but they are more easily able to do so in LME-type countries than in CME-type countries.

In light of the existing literature, it could be argued that managerial compensation may be bound up with context and the relative strength of the shareholder value maximisation agenda in different settings (Dore 2000; Erturk et al. 2008). However, not much has been done to examine management compensation practices, especially in MNCs, in various institutional settings (Gooderham et al. 2006).

Although the early literature on comparative capitalism focuses on the LME/CME dichotomy, it soon becomes apparent that many non-liberal market economies do not easily fit into the general CME categorisation. Firstly, there are the economies of Mediterranean Europe (Mixed Market Economies) (Hancke et al. 2007; Amable 2003). These economies are characterised by weakly aligned institutions, whose coverage of the large SMEs and informal economies is poor. The latter areas of economic activity coexist with large organisations, many of which are either statist, or were formally under state ownership and/or owe their status to state patronage either presently or at some time in the past. Up until the present crisis, such large organisations were characterised by high levels of job security, with bureaucratic reward systems, with standardised pay rates per job, and promoting for seniority and relative access to patronage. Secondly, there are the Transition Market Economies of Central and Eastern Europe. Over twenty years after the demise of state socialism, they retain distinct characteristics, a product of long-term historical legacies, combined with the shocks of the adjustments of the 1990s (Hancke et al. 2007). The influx of large numbers of Western European MNCs into such countries would vest the approaches adopted by such companies with great importance, especially given ongoing institutional fluidity. Whilst the non-liberal Rhineland (Continental European) and Nordic (Social Democratic) economies are commonly subsumed in to the CME category, Amable (2003) notes that they retain important differences, inter alia, in terms of the regulation of labour markets and corporate governance. Although unions are stronger in the Nordic countries, security of tenure is weaker, as are inter-firm ties at industry level (Amable 2003). Again, as industry specific training is weaker in these countries (Amable 2003), it could be argued that managers are less likely to have industry specific qualifications. This may result in managers having weaker ties to specific industries and, hence, being more likely to be swayed or attracted by contingent short term approaches to reward, rather than simply the intrinsic rewards of stewardship of the firm. Palley (1997) argues that if tenure is weaker, individuals are less likely to be around to enjoy the returns generated by long term projects, and hence be more orientated to rewards that yield more immediate benefits. In continental European economies such as Germany, Sanders and Tuschke (2007, p. 40) argue that share-based incentives have been historically construed as antithetical to its national corporate governance logic and, hence, such practice is not popular among domestic firms.

Statement of Hypotheses

Owing to the complexity of MNCs, and their role as norm entrepreneurs, it could be argued that the usage of incentive-related compensation which links managerial compensation to broader organisational performance is more likely in these firms (Sako and Kotosaka 2012; Dore 2008; Holmstrom 1979; Jensen and Murphy 1990). Whitley (2001, p. 31) argues that the international business environment is relatively anomic, facilitating an integration of activities and practices across national boundaries, and encouraging ownership-based coordination and, hence, a stronger emphasis on shareholder rights. Here we distinguish between the share-based incentives, such as stock options, and the non-monetary incentives, a lesser discussed element of management compensation. O’Donnell (2000) asserts that non-monetary rewards, such as secondments and inter-organisation tasks, enhance cooperation in the unique intra-firm international interdependent organisational structure of MNCs. In the context of MNCs, we argue that compared to domestic firms, MNCs are more likely to use both share-based and non-monetary incentives. Hence, we hypothesise that:

Hypothesis 1:

Everything else being equal, the use of share-based incentives is higher for the management in MNCs than for the management in domestic firms.

Hypothesis 2:

Everything else being equal, the use of non-monetary incentives is higher for the management in MNCs than for the management in domestic firms.

It is commonly held that the LME model is the capitalist archetype where institutions and practices are most closely aligned to securing shareholder primacy (see for example Hall and Soskice 2001; Dore 2000). Hence, it could be argued that MNCs in LMEs are likely to use incentive-related compensation to promote shareholder value (Filatotchev and Wright 2011; Sanders and Carpenter 1998); as stakeholder rights are weaker in such contexts, firms are freer to concentrate on the bottom line (Dore 2000). For example, Lane (2001, p. 86) notes, that German MNCs operating in the USA have taken advantage of the setting to try out new organisational norms and people management practices. For similar reasons, such practices are also likely to be widely used by domestic firms in LMEs to incentivise managers. However the gap in using incentive-related compensation between domestic firms and MNCs could be much larger in other contexts. For example, it could be argued that MNCs in non-liberal settings are likely to adopt share-based incentives when removed from the immediate constraints of their home environment (Sanders and Tuschke 2007; Whitley 2010). Institutional restraints on firms are likely to weaken as they cross national boundaries, favouring more instrumental approaches to managing people (Whitley 2001). For similar reasons, it could be argued that the usage of other forms of managerial incentives are more likely to be closely aligned to domestic firm practices in LMEs than in other settings (Dore 2000). Hence, we hypothesise that:

Hypothesis 3:

Everything else being equal, the difference in the use of share-based incentives in MNCs compared to domestic firms is smaller in LMEs than in other varieties of capitalism.

Hypothesis 4:

Everything else being equal, the difference in the use of non-monetary incentives in MNCs compared to domestic firms is smaller in LMEs than in other varieties of capitalism.

Methods

Sample and Variables

In this paper we use data from the 2004/2005 Cranet survey. This is the round of the survey which gives us the best range of countries for our analysis. The survey gathers evidence on HRM practices from organisations and we use evidence from eighteen countries in or geographically close to Europe. Details on the Cranet surveys and their methodological foundations may be found in Nordhaug et al. (1996); Croucher et al. (2010); and Brewster et al. (2007). This survey provides comparative firm level evidence, in contrast to much existing work comparing institutional framework to firm outcomes, which tends to focus on macro-trends (Roe 2003; Baums and Scott 2005) or selected case study evidence. Table 1 provides details of the definitions of the variables. Focusing on various mechanisms that provide ‘value-increasing incentives’ (Jensen and Murphy 1990, p. 226), we construct Share-based incentive, a dummy variable which is set to one if the management is offered at least one of the following schemes: Employee share schemes or stock options; and zero otherwise. For non-monetary rewards we construct Non-monetary incentive, which focuses on managerial career development, to measure the non-monetary incentive-related compensation. This variable is the sum of scores on the following methods used for managerial career development: (1) special tasks/projects to stimulate learning, (2) involvement in cross-organisational/disciplinary/functional tasks, (3) participation in project team work; (4) networking; (5) formal career plans; (6) succession plans; (7) planned job rotation; (8) ‘high-flier’ schemes (i.e. particular human resource development schemes targeted at the talented earmarked for potentially rapid promotion up the organisational hierarchy); (9) experience schemes; and (10) secondments to other organizations2 For each method, a score of 0 is given if it is not at all used; 1 if it is used to a small extent; 2 if it is used to a large extent; and 3 if it is used entirely. This measure is similar to the concept of non-monetary career-related awards used by O’Donnell (2000), who argues that incentives other than financial compensation are particularly important in the context of MNCs. Schemes such as ‘identification as high potential employee’ or ‘increased international experience’ would be useful ‘to facilitate the cooperation needed with greater interdependence’ in MNCs (O’Donnell 2000, p. 533).

Table 1

Definitions of the variables

MNC

A dummy variable which is set to one if the firm is a multinational company and zero if the firm is a domestic company.

Share-based incentive

A dummy variable which is set to one if the management is offered at least one of the following: employee share schemes, or stock options; and zero otherwise.

Non-monetary incentive

This variable measures the non-monetary elements of compensation. It is the sum of scores on the following methods for managerial career development: (1) special tasks/projects to stimulate learning, (2) involvement in cross-organisational/disciplinary/functional tasks, (3) participation in project team work; (4) networking; (5) formal career plans; (6) assessment centres; (7) succession plans; (8) planned job rotation; (9) ‘high-flier’ schemes; (10) experience schemes; (11) secondments to other organisations. The score for each method takes on a value of:

0 if it is not used at all;

1 if it is used to a small extent;

2 if it is used to a large extent; and

3 if it is used entirely.

Lnsize

The size of the firm measured by the logarithm of total number of employees.

Profit

An ordinal variable which takes on a value of:

2 if the firm’s the gross revenue over the past 3 years has been well in excess of costs;

1 if the firm’s the gross revenue over the past 3 years has been sufficient to make a small profit;

0 if the firm’s the gross revenue over the past 3 years has been enough to break even;

-1 if the firm’s the gross revenue over the past 3 years has been insufficient to cover costs; and

-2 if the firm’s the gross revenue over the past 3 years has been so low as to produce large losses.

M&A

A dummy variable which is set to one if the firm has been involved in a merger and acquisition, or takeover over the past three years; and zero otherwise.

Our key variable of interest to test Hypothesis 1 is a dummy variable, MNC, which is set to one if the firm is multinational and zero if it is a domestic (non-MNC) firm. We appreciate that this dichotomous categorisation does not reveal the scope of the firm beyond whether it is multinational or not. There are clear differences between MNCs that only operate in two neighbouring states and those that have a truly global scope. However, in all cases, even operating in two neighbouring states involves at least some greater variation in internal practices than operating in a single market. Here we focus on the relationship between institutional and specific firm-level management compensation practices. As such, this does not take account of differences in strategic orientation. Some MNCs may have chosen to expand geographically, whilst others may have attained multinational status without having planned this at a strategic level. A closer evaluation of differences in strategic orientation would represent fertile ground for future research. We also control for other firm-specific characteristics that may influence management compensation. These are: Firm size (Lnsize) measured as the logarithm of total number of employees; firm profitability (Profit); and recent corporate restructuring activities such as merger and acquisitions, takeovers or demergers (M&A).

Our study only looks at the relationship between managerial compensation, and whether a company is an MNC or not, and does not explore the differences in the managerial compensation practice within MNCs, that is, between head offices and subsidiaries. It is recognised that subsidiaries may, in reality, have very different internal dynamics to head offices and an evaluation as to differences in pay systems between them would constitute a further important area for future research. Due to data constraints, our study also does not examine the impact of countries of domicile on the compensation practices in MNCs. We recognise that a closer examination of differences in practice between countries of headquarters and domicile would represent a fruitful area for further analysis3

Descriptive Statistics

Table 2 reports the descriptive statistics on the variables used in this paper (Panel A) and the distributions of firms across countries (Panel B). Panel A shows that 45 % of the sample firms are multinational. Twentynine percent of the sample firms report the use of either employee share schemes, or stock options or both in management compensation. There is a significant variation across firms in terms of the score for non-monetary incentive-related compensation, ranging from 0 to 24 with an average of only 7.264. Table 3 shows the correlation matrix of the variables used in the paper.

Table 2

Descriptive statistics

Panel A Descriptive statistics

Variable

Mean

Median

Std. dev.

Min

Max

Obs.

MNC

0.448

0

0.497

0

1

1,728

Share-based incentive

0.294

0

0.456

0

1

1,728

Non-monetary incentive

7.264

7

4.364

0

24

1,728

Lnsize

6.061

5.858

1.251

1.792

12.260

1,728

Profit

1.046

1

1.117

− 2

6

1,728

M&A

0.503

1

0.500

0

1

1,728

Panel B Distribution across countries

Country

Number

Percentage (%)

VOC

Austria

88

5.093

Continental

Belgium

66

3.819

Continental

Bulgaria

65

3.762

Transition

Czech

25

1.447

Transition

Denmark

176

10.185

Social democratic

Estonia

28

1.620

Transition

Finland

70

4.051

Social democratic

Germany

123

7.118

Continental

Greece

61

3.530

Mixed

Iceland

36

2.083

N/A

Israel

47

2.720

Mixed

Norway

81

4.688

Social democratic

Slovakia

202

11.690

Transition

Slovenia

67

3.877

Transition

Sweden

90

5.208

Social democratic

The Netherlands

81

4.688

Continental

Tunisia

105

6.076

N/A

UK

317

18.345

Liberal

The variety of capitalism (VOC) classification follows the Amable (2003) and Hancke et al.’s (2007) broad classifications

Table 3

The correlation matrix

 

MNC

Share-based incentive

Non-monetary incentive

SIZE

Profit

M&A

MNC

1

0.162

0.169

0.165

0.056

0.217

Share-based incentive

 

1

0.191

0.173

0.045

0.194

Non-monetary incentive

  

1

0.238

0.095

0.139

Lnsize

   

1

0.046

0.259

Profit

    

1

0.020

M&A

     

1

Estimation Methods

In order to test our hypotheses we estimate the following equation:

Where Yi is the dependent variable, which is either Share-based incentive or Non-monetary incentive according to the variable definitions above.

We use probit regression to estimate model (1) when Share-based incentive, which is discrete, is used as the dependent variable. When Non-monetary incentive, which is a count variable, is used as the dependent variable, we use the Negative Binomial 2 (NB2) regression (see Cameron and Trivedi 2005 for more details). All regressions are estimated using the White-robust estimator of variance.

We estimate model (1) for the whole sample and for sub-samples classified according to the national institutional frameworks using the Amable’s (2003) and Hancke et al.’s (2007) classification. Further, besides controlling for firm size by the explanatory variable Lnsize, for each sample/sub-sample we estimate model (1) separately for small and large firms, i.e. firms with the number of employees below and above the whole sample median, respectively. This is to address the concern that our results may be driven merely by size4 as MNCs in our sample are significantly larger than their domestic counterparts (see Table 4 below). Indeed, existing literature points to firm size as a key determinant of the management compensation packages where managers in larger firms receive higher pay and more incentives (Gomez-Mejia 1994; Conyon et al. 1995; Murphy 1999; Gabaix and Landier 2007). If our results hold for large firms only such results are sensitive to firm size. If, on the other hand, no discernible pattern is detected when we compare the estimation results for small and large firms, our results are not driven by the impact of the presence of large MNCs in the data. To ensure that our results are robust we also perform the following robustness checks: i) conduct analysis for three terciles of firms, small, medium and large; ii) conduct analysis for small and large MNCs, i.e. MNCs with the number of employees below and above the whole sample median; and iii) conduct analysis for small and large firms using the median firm size of the variety of capitalism in which firms operate rather than the whole sample median firm size. All the aforementioned analyses produce qualitatively similar results and thus we only report results where we divide firms into small and large according to the whole sample median firm size5

Table 4

MNCs and domestic firms: the univariate test

Share-based incentive

Non-monetary incentive

Lnsize

M&A

Profit

MNCs

Domestic

MNCs

Domestic

MNCs

Domestic

MNCs

Domestic

MNCs

Domestic

0.376

0.227***

8.081

6.601***

6.289

5.875***

1.115

0.989***

0.624

0.406***

This table reports means and the results of the difference in means tests of the incentive-related variables and other firm-characteristics for MNCs and domestic firms. All variables are defined in Table 1

*, ** and *** denote significance of the difference in means at the 10 %, 5 % and 1 % level, respectively

Analysis and Results

MNCs and Incentive-related Management Compensation: The Univariate Analysis

Table 4 reports the means and the results of the difference in means tests of the incentive-related variables in MNCs and domestic firms. We find that, compared to domestic firms, MNCs are more likely to implement share-based compensation and provide more non-monetary incentives. This univariate analysis, however, does not take into account any other firm-specific characteristics that may determine the management compensation such as firm size, profitability or restructuring activities. This issue is taken up in the next section. We also report the means and the results of the difference in means tests of some firm-specific characteristics of MNCs and domestic firms. MNCs seem to be larger, more likely to engage in M&A compared to single-country firms and are more profitable.

MNCs and Incentive-related Management Compensation: The Multivariate Analysis

Table 5 reports the estimation results using Share-based incentive and Non-monetary incentive as the dependent variables, respectively. We report the estimation results for small and large firms separately. In all specifications the coefficients of MNC are positive and statistically significant, suggesting that MNCs are more likely to use incentive-related compensation than domestic firms. Further, the coefficients of MNC are positive and statistically significant in both small and large firms. This indicates that the impact of being multinational is not merely driven by the fact that MNCs tend to be larger firms. The findings confirm that, as suggested by the univariate analysis, what MNCs do is distinct from the behaviour of domestic firms, and they therefore strongly support Hypotheses 1 and 2. It seems that MNCs do play the role of norm-entrepreneurs aiming to align managers more closely to the shareholder agenda (Sako and Kotosaka 2012). The coefficients on MNC in the regressions for Share-based incentive in small and large firms are of similar magnitude, suggesting that the multinational effect is similar among small and large firms in terms of using share-based incentive. The coefficient for Non-monetary incentive in small firms however is nearly three times larger than that in large firms, suggesting that the gap in using non-monetary rewards between MNCs and domestic firms is less pronounced among large firms. This implies that large domestic firms use more non-monetary rewards than smaller domestic firms.

Table 5

MNCs and incentive-related compensation

 

Share-based incentive

Non-monetary incentive

 

Small

Large

Small

Large

 

(1)

(2)

(3)

(4)

MNC

0.319***

0.388***

0.227***

0.072*

 

(0.003)

(0.000)

(0.000)

(0.072)

Lnsize

0.041

0.226***

0.082*

0.088***

 

(0.696)

(0.000)

(0.087)

(0.000)

Profit

0.048

0.053

0.054***

0.069***

 

(0.276)

(0.226)

(0.005)

(0.000)

M&A

0.336***

0.333***

0.063

0.033

 

(0.001)

(0.001)

(0.150)

(0.382)

Constant

− 1.671***

− 2.791***

1.380***

1.384***

 

(0.009)

(0.000)

(0.000)

(0.000)

Obs.

838

890

838

890

Wald χ2

83.750

105.090

146.774

169.971

Log-Pseudolikelihood

− 423.34

− 506.52

− 2318.03

− 2551.39

This table reports results of the probit regressions for the association between firms’ geographical diversification strategies and share-based incentives and NB2 regression for the association between firms’ geographical diversification strategies and non-monetary incentives. The dependent variable is Share-based incentive in columns 1 and 2 and Non-monetary incentive in columns 3 and 4. The estimation results are for small firms (firm size below the sample median) in columns 1 and 3 and large firms (firm size above the sample median) in columns 2 and 4. All variables are defined in Table 1. Figures in parentheses are the p value. Standard errors (not reported) are obtained from the White-robust estimator of variance. Country dummies are included but not reported for brevity

*, ** and *** denote significance at the 10 %, 5 % and 1 % level, respectively

The results in Table 5 also confirm the finding of existing research that firm size is positively related to management compensation systems (Gomez-Mejia 1994; Conyon et al. 1995; Murphy 1999; Gabaix and Landier 2007), reflecting firm complexity and/or greater resources available to implement more sophisticated reward systems. It should be noted that size does not have a significant impact on the use of shared-based incentives in small firms while the coefficient of MNC remains significant. This provides further support for our finding that being multinational is a factor determining the design of management compensation packages.

It can also be seen that M&A has a positive and statistically significant impact on the use of share-based incentives in both small and large firms. The degree of firm diversification changes through mergers and acquisitions (Amihud and Lev 1981; Baker et al. 1988), which allows new opportunities to renege on implicit contracts and to institute new agendas: Given the global ecosystemic dominance of the shareholder value model, this may suggest that such processes will be associated with a stronger emphasis on contingent managerial pay (Jessop 2012; Streeck 2009). Firms that engage in M&A therefore may find the need to implement devices to align the interests of managers with that of shareholders. Again, managers in target firms may be bound by implicit contracts with staff, which may need to be reformed or adjusted to suit the new overall organisational culture. More importantly, the results show the higher likelihood of MNCs using managerial incentive schemes even after we control for variations in these firm-specific characteristics, including size, profitability and M&A activities that influence how management compensation is determined.

To examine whether our results hold in different national institutional settings and especially whether the MNC effect in the LMEs differs from that in other settings, we rerun the regression for Model (1) for two sub-samples: i) firms operating in the Liberal Market Economies (LMEs) and ii) firms operating in all other varieties of capitalism (Non-LMEs)6 We then compare the magnitude of the coefficient for MNC in the estimation results for each sub-sample. We further estimate model (1) for the sub-sample of firms operating in the CMEs, as defined by Hall and Soskice (2001), which includes both the Continental Market Economies and the Social Democratic Economies as classified in Amable (2003) and Hancke et al.’s (2007).

Table 6 reports the estimation results using Share-based incentive as the dependent variable. The coefficient for MNC is statistically significant only in the regression for large firms in the LMEs (column 2). But it is strongly statistically significant in both the regressions for small and large firms in all the other VOCs (columns 3 and 4) and in the CMEs (columns 5 and 6). Further, the statistically significant coefficient for MNCis the smallest in magnitude in the LMEs. While this confirms the finding in Table 5 that MNCs are more likely to use share-based incentives than domestic firms, it suggests that such multinational effect is much less pronounced in the LMEs, especially compared to the CMEs. This strongly supports Hypothesis 3. Table 7 reports the estimation results using Non-monetary incentive as the dependent variable. The coefficient of MNC in the LMEs is not statistically significant in either the regression for small firms or that for large firms (columns 1 and 2). But it is statistically significant in both the regressions for small and large firms in all the other VOCs (columns 3 and 4) and in the regressions for small firms in the CMEs (column 5). This strongly supports Hypothesis 4 that the multinational effect regarding the use of incentive-related compensation in MNCs is less pronounced in the LMEs. Together the results in Table 6 and Table 7 imply that this incentive-related pay practice is already popular among domestic firms in the LME setting but not among their counterparts in other settings.

Table 6

MNCs, share-based incentive schemes and the varieties of capitalism

 

LMEs

All Non-LMEs

CMEs

 

Small

Large

Small

Large

Small

Large

 

(1)

(2)

(3)

(4)

(5)

(6)

MNC

0.432

0.408*

0.288**

0.435***

0.471***

0.587***

 

(0.182)

(0.051)

(0.013)

(0.000)

(0.002)

(0.000)

Lnsize

− 0.458

0.339***

0.036

0.201***

− 0.067

0.233***

 

(0.150)

(0.000)

(0.743)

(0.001)

(0.622)

(0.000)

Profit

− 0.002

0.048

0.113**

0.021

0.191***

0.068

 

(0.987)

(0.611)

(0.039)

(0.696)

(0.006)

(0.287)

M&A

0.724***

0.343

0.357***

0.317***

0.360**

0.354**

 

(0.008)

(0.115)

(0.001)

(0.009)

(0.013)

(0.022)

Constant

1.161

− 3.055***

− 1.597**

− 2.689***

− 1.196

− 3.089***

 

(0.493)

(0.000)

(0.013)

(0.000)

(0.125)

(0.000)

Obs.

127

190

661

609

380

395

Wald χ2

12.190

27.476

67.142

71.928

51.687

59.614

Log-Pseudolikelihood

− 58.384

− 113.09

− 357.062

− 340.192

− 200.403

− 213.518

This table reports results of the probit regressions for the association between firms’ geographical diversification strategies and incentive schemes. The dependent variable is Share-based incentive. All variables are defined in Table 1. The estimation results are for small firms in the LMEs (firm size below the sample median) in column 1 and large firms in the LMEs (firm size above the sample median) in column 2; for small and large firms in all other varieties of capitalism other than the LMEs in columns 3 and 4 respectively; and for small and large firms in the CMEs, including the Continental Market Economies and Social Democratic Economies, in columns 5 and 6 respectively. Figures in parentheses are the p value. Standard errors (not reported) are obtained from the White-robust estimator of variance. Country dummies are included but not reported for brevity

*, ** and *** denote significance at the 10 %, 5 % and 1 % level, respectively

Table 7

MNCs, non-monetary incentives and the varieties of capitalism

 

LMEs

Non-LMEs

CMEs

 

Small

Large

Small

Large

Small

Large

 

(1)

(2)

(3)

(4)

(5)

(6)

MNC

− 0.042

− 0.037

0.240***

0.107**

0.156***

0.060

 

(0.754)

(0.660)

(0.000)

(0.024)

(0.008)

(0.263)

Lnsize

0.171

0.112***

0.064

0.064***

0.104

0.056***

 

(0.200)

(0.000)

(0.179)

(0.000)

(0.101)

(0.002)

Profit

0.044

0.055*

0.056**

0.091***

0.045*

0.107***

 

(0.328)

(0.095)

(0.021)

(0.000)

(0.094)

(0.000)

M&A

− 0.045

0.069

0.063

0.040

0.036

0.032

 

(0.641)

(0.437)

(0.198)

(0.364)

(0.526)

(0.530)

Constant

0.810

1.139***

1.481***

1.496***

1.305***

1.563***

 

(0.260)

(0.000)

(0.000)

(0.000)

(0.000)

(0.000)

Obs.

127

190

661

609

380

395

Wald χ2

3.028

36.709

71.962

102.842

38.731

81.357

Log-Pseudolikelihood

− 329.94

− 529.14

− 58.384

− 113.098

− 1052.229

− 1124.694

This table reports results of the NB2 regressions for the association between firms’ geographical diversification strategies and incentive schemes. The dependent variable is Non-monetary incentive. All variables are defined in Table 1. The estimation results are for small firms in the LMEs (firm size below the sample median) in column 1 and large firms in the LMEs (firm size above the sample median) in column 2; for small and large firms in all other varieties of capitalism other than the LMEs in columns 3 and 4 respectively; and for small and large firms in the CMEs, including the Continental Market Economies and Social Democratic Economies, in columns 5 and 6 respectively. Figures in parentheses are the p value. Standard errors (not reported) are obtained from the White-robust estimator of variance. Country dummies are included but not reported for brevity

*, ** and *** denote significance at the 10 %, 5 % and 1 % level, respectively

Results in Table 6 and Table 7 also reveal that in the CMEs the multinational effect is stronger for the share-based incentives compared to the non-monetary incentives. It appears that firms may use a combination of practices. This is contrary to what is suggested in much of the literature (von Glinow and Teagarden 2006; Carpenter et al. 2001) that firms tend to use either “calculative” or “co-operative” practices. When we rerun the regression for Model (1) for the sub-samples of firms in the Social Democratic Economies, Continental European Economies; Mixed Market Economies and Transition European Emerging Market Economies,7 it appears that the gap between MNCs and domestic firms in terms of using the share-based incentives is more pronounced among firms in the Social Democratic Economies and Continental European Economies. On the contrary the gap between MNCs and domestic firms in terms of using the share-based incentives is more pronounced among firms in the Mixed Market Economies and Transition Market Economies. These results reveal a more detailed and diverse picture regarding practices of incentivising and monitoring managers in MNCs and domestic firms across setting. The results are presented in Appendices A and B.

Discussion

An immediate question that arises from the findings is whether the multinational effect on the contingent pay systems, and more specifically, the usage of shared-based compensation, is simply an effect of their complexity or of their role as “norm entrepreneurs”, pioneering LME-like practices within other settings. The agency theory tends to assume that, irrespective of context, managers will, if left to their own devices, pursue an agenda of “empire building” in the interests of their own personal prestige (rather than because of, say, bureaucratic economies of scale) (Jensen and Meckling 1976; Murphy 1999). This would suggest that larger firms will experience the same agency problems irrespective of setting, and that the need for incentive based compensation—above all, share-based incentives—will be similar in all larger firms, irrespective of geographical scope. In other words, is the greater prevalence of incentive-related compensation in MNCs that we document simply because they tend to be larger, rather than because they straddle national boundaries and are thus better equipped to challenge embedded local ways of doing things? We find that the latter is indeed the case. Our findings suggest that the agency theory provides a weaker explanation than the relative ability of MNCs to depart from—and, in doing so, undermine—more cooperative models that see the firm as fulfilling a wider range of functions than simply the short term maximisation of shareholder value (O’Donnell 2000; Sanders and Tuschke 2007; Dore 2008; Streeck 2009).

Overall, when comparing the Liberal Market Economies with other varieties of capitalism we find that the gap between MNCs and domestic firms is much less pronounced in liberal markets. This is in line with the proposition of the Variety of Capitalism approach that, compared to other frameworks, in the Liberal Market Economies where shareholders’ rights are stronger and shareholder wealth is typically measured via the share price (Dore 2000; Hall and Soskice 2001), both MNCs and domestic firms adopt mechanisms to align the interest of managers with those of shareholders. In other contexts only MNCs that are embedded in multi-institutional settings are more likely to adopt such mechanism.

Hence, we find that institutional effects are uneven, varying according to firm scope and setting. What is likely to cause such internal diversity? In the first instance, this may be an issue of complementarity; practices yield more favourable outcomes when combined than an analysis of their constituent parts would suggest (Gordon and Roe 2004; Crouch 2005). It may be that the manner in which managerial compensation systems are structured—combining different types of monetary and non-monetary incentives—work together so well that they outweigh any costs incurred through incompatibility with national institutional arrangements. Why then, do domestic firms not mimic such arrangements? The latter could reflect the additional benefits of having standardised managerial compensation systems across an MNC regardless of national setting in terms of economies of scale and in lowering transaction costs when relocating managers; in other words, the gains to domestic firms could be more marginal. MNCs are more likely to be able to rely on cross border linkages and wider knowledge of different ways of doing things to challenge or be different from domestic firms (Lane 2005). Moreover, domestic firms are likely to have denser and more deeply embedded ties to their peers and other stakeholders: This would serve to discourage “norm innovation” (see also Sanders and Tuschke 2007). Global trends towards liberalisation of state policy open up room for innovation; whilst some firms—particularly those with limited resources—may be reluctant to cast aside familiar ways of doing things, others are likely to do so. This means that change is unlikely to be uniform and uncontested. Indeed, firms may experiment with new practices in one area, and retain more familiar ones that have known benefits in others. This may either reflect a “mix and match” approach, combining different approaches in order to gain benefits from both, or a greater desire to ensure managers remain aligned to a shareholder value agenda in such contexts. It could also reflect a desire by firms to capitalise on the growing weaknesses of unions in many national settings.

Conclusion

This paper reveals persistent diversity reflecting firm type and national setting. The findings suggest that MNCs are more likely to use incentive-related compensation than domestic firms. This could reflect their role in pioneering practices associated with the shareholder value model, and liberal market capitalism more generally, in other settings (Sako and Kotosaka 2012; Dore 2008). Such firms may indeed play the role of ‘norm entrepreneurs’, pioneering new sets of practices, which ultimately serve as a basis for fundamental change across a particular settings (Sako and Kotosaka 2012; Dore 2008). However, this difference between MNCs and domestic firms is less pronounced in LMEs compared to other settings, probably a reflection of the wider usage of contingent compensation in the former, and the greater extent to which LME-type practices are more prone to diffuse into other models, rather than vice versa. Our study does highlight the extent to which MNCs headquartered in other countries are prone to influence from the LME model; as Jessop (2012) notes, the latter model has global ecosystemic dominance, and it would appear that firms most inserted in the global economy are indeed more likely to adopt specific firm practices associated with this model. At the same time, there may be limits to such diffusion. MNCs may, at least in part, follow standardised practices owing to bureaucratic economies of scale and standardisation, whose benefits might outweigh any costs incurred as a result of incompatibility with local norms and expectations in different settings. In contrast, even if of similar size, domestic firms can reap similar economies simply through adhering to a dominant local model, whilst not incurring any of the latter costs.

Appendix (A):MNCs, share-based incentive schemes and the varieties of capitalism

 

Social Democratic

Continental European

Transition

Mixed

 

Small

Large

Small

Large

Small

Large

Small

Large

MNC

0.559***

0.575**

0.323

0.552***

− 0.213

0.193

1.032**

0.252

 

(0.003)

(0.015)

(0.219)

(0.010)

(0.302)

(0.493)

(0.043)

(0.498)

Lnsize

0.014

0.242**

− 0.183

0.245***

0.322*

− 0.391**

− 0.711

0.540*

 

(0.938)

(0.017)

(0.323)

(0.007)

(0.058)

(0.026)

(0.218)

(0.011)

Profit

0.173**

− 0.005

0.216

0.189*

0.177

− 0.091

− 0.427*

− 0.026

 

(0.033)

(0.955)

(0.122)

(0.076)

(0.126)

(0.447)

(0.061)

(0.881)

M&A

0.322*

0.457**

0.450*

0.255

0.423**

0.246

0.599

0.281

 

(0.075)

(0.034)

(0.082)

(0.256)

(0.034)

(0.302)

(0.179)

(0.471)

Constant

− 1.363

− 3.362***

− 0.579

− 3.227***

− 2.648***

1.960

3.112

− 4.699***

 

(0.195)

(0.000)

(0.574)

(0.000)

(0.008)

(0.108)

(0.294)

(0.001)

Obs.

225

192

155

203

240

147

41

67

Wald χ2

21.999

34.555

12.694

29.322

21.940

11.557

8.834

12.647

Log-Pseudolikelihood

− 135.91

− 106.20

− 63.839

− 105.99

− 124.46

− 81.172

− 20.113

− 36.331

This table reports results of the probit regressions for the association between firms’ geographical diversification strategies and incentive schemes in the Social Democratic Economies, Continental European Economies, Transition Market Economies and Mixed Market Economies. The dependent variable is Share-based incentive. All variables are defined in Table 1. The regressions are estimated for small firms (firm size below the sample median) and large firms (firm size above the sample median) separately. Figures in parentheses are the p value. Standard errors (not reported) are obtained from the White-robust estimator of variance. Country dummies are included but not reported for brevity

*, ** and *** denote significance at the 10 %, 5 % and 1 % level, respectively

Appendix (B):MNCs, Non-monetary incentives and the varieties of capitalism

 

Social democratic

Continental European

Transition

Mixed

 

Small

Large

Small

Large

Small

Large

Small

Large

MNC

0.179**

0.027

0.133

0.078

0.301***

0.216*

0.428*

0.162

 

(0.014)

(0.745)

(0.173)

(0.273)

(0.003)

(0.056)

(0.081)

(0.312)

Lnsize

0.079

0.061**

0.111

0.053**

0.012

0.036

0.219

0.260***

 

(0.230)

(0.015)

(0.328)

(0.033)

(0.881)

(0.580)

(0.485)

(0.002)

Profit

0.052

0.093***

0.027

0.123***

0.119**

0.011

− 0.120

0.142

 

(0.116)

(0.006)

(0.576)

(0.001)

(0.032)

(0.799)

(0.335)

(0.155)

M&A

− 0.026

0.035

0.128

0.033

0.095

0.179*

0.024

− 0.408**

 

(0.717)

(0.611)

(0.164)

(0.652)

(0.316)

(0.054)

(0.937)

(0.025)

Constant

1.524***

1.689***

1.270**

1.552***

1.795***

1.897***

0.521

0.113

 

(0.000)

(0.000)

(0.041)

(0.000)

(0.000)

(0.000)

(0.752)

(0.855)

Obs.

225

192

155

203

240

147

41

67

Wald χ2

30.233

36.332

10.780

40.715

31.363

21.724

6.208

14.311

Log-Pseudolikelihood

− 612.63

− 533.58

− 438.11

− 590.51

− 705.18

− 428.46

− 119.03

− 199.13

This table reports results of the NB2 regression for the association between firms’ geographical diversification strategies and non-monetary incentives in the Social Democratic Economies, Continental European Economies, Transition Market Economies and Mixed Market Economies. The dependent variable is Non-monetary incentive. All the regressions are estimated for small firms (firm size below the sample median) and large firms (firm size above the sample median) separately. All variables are defined in Table 1. Figures in parentheses are the p value. Standard errors (not reported) are obtained from the White-robust estimator of variance. Country dummies are included but not reported for brevity

*, ** and *** denote significance at the 10 %, 5 % and 1 % level, respectively

Footnotes
1

It is acknowledged that in such contexts, it is more difficult for mutually supportive complementarities to emerge and persist; hence, they will face pressures to converge to one or other more mature model (Hall and Gingrich 2004; Hancke et al. 2007). Amable (2003) offers a similar classification, with two main differences. Firstly, he does not deal explicitly with Eastern and Central Europe other than as countries en route to one or other of the more mature archetype. Secondly, he makes a very useful distinction between continental European capitalism (i.e. continental European CMEs) and Scandinavia (social democratic capitalism) and argues that the latter are distinct archetypes (Amable 2003), without one necessarily being inherently superior to the other.

 
2

“High-flier schemes” are development programmes for selected younger staff who might make senior management; “experience schemes” are where staff have the opportunity to ‘act up’ in a managerial role above their own; and “secondments” are where staff are given experience in other, linked or external, organisations.

 
3

In an initial analysis we find that foreign subsidiaries of MNCs are more likely to implement incentive-related pay practice than domestic firms in their countries of domicile. Further, we find that foreign subsidiaries of MNCs from the USA that operate in the Social Democratic, Continental European and Liberal Market Economies are much more likely to use incentive-related compensation compared to those of non-USA MNCs.

 
4

We thank an anonymous referee for this point.

 
5

These results are available from the authors upon request.

 
6

As we note above, we report the result in which firms are classified as small and large according to the whole sample median. Our results remain qualitatively the same when firms are classified according to the medians of the varieties of capitalism in which they operate.

 
7

As noted above, Amable does not accord much attention to this region, other than in terms of relative convergence to other models. Hence, we have used Hancke et al.’s terminology for a region still undergoing structural change.

 

Acknowledgments

We thank the editor and two anonymous referees for helpful comments/suggestions that significantly improve the paper. We would also like to thank the members of the Cranet network who collected the data and the Cranfield School of Management for co-ordinating the network. All errors remain ours.

Copyright information

© Springer-Verlag Berlin Heidelberg 2013