FREE ELECTRONIC LIBRARY - Dissertations, online materials

Pages:     | 1 |   ...   | 11 | 12 || 14 | 15 |   ...   | 22 |

«Three essays on corporate boards R. Øystein Strøm A dissertation submitted to BI Norwegian School of Management for the degree of Dr.Oecon SERIES ...»

-- [ Page 13 ] --

The table shows the percentage of firm-year observations of employee directors in various employment sizes. It turns out that in firms where employees may demand representation, few do so. In the 101-200 employees category, 61.5 per cent do not have employee directors. Furthermore, in the highest category, where representation is compulsory, if the industry is not exempted, employees have no board seats in about one third of the companies. Among the firms that do have employee directors, the law’s minimum, two representatives, is found in the majority of cases. Very few have four employee board seats. Thus, the Jensen and Meckling (1979) conjecture that co-determination requires law backing seems to be supported in our Norwegian data.

Next, table 3.2 shows the distribution of employee directors according to industry, and also the percentage of firms with no employees on the board in each year.

Table 3.2

Exempted industries such as Energy and Transport (including shipping) have no employee directors to a higher degree than average. The low representation in Hotels, Restaurants and Entertainment is perhaps due to high labour turnover. The two industries Health Care Equipment and Supply as well as Software and Services also have a lower than average representation. These are industries where the human capital element should be above average, and co-determination of extra value, according to stakeholder theory. Yet obviously, employees do not demand board seats to a great extent. The time trend is that firms with no employee

3.5. ESTIMATION AND METHOD 75 directors increase in relative importance. Thus, nothing in the overall descriptive statistics shows that co-determination is a preferred organisational mode. Firms seem to avoid it if they can, and keep it to a minimum if they cannot.

Variable definitions are shown in table 3.3, which also shows the main characteristics of variables in the analysis in the two main sub-samples of co-determined and shareholder determined firms.

Table 3.3

The table shows that a large number of variables are distributed differently in the two sub-samples. The firm performance variables Tobin’s Q and stock return are not significantly different, while the ROA in codetermined firms is significantly higher than in shareholder determined.

Apart from directors’ holdings, all other variables are significantly different at the 5.0% level or better. Obviously, the two types of firms are different.

The table shows that the fraction of employee directors is 0.301, or slightly below the minimum requirement for the 200+ employee size group.

Similarly to findings in table 3.2, this is evidence that the firms attempt to minimize the employee director importance.

The two firm groups differ in background variables, notably firm size.

The co-determined firms are larger on average. This warrants paying a particular attention to the largest firm size groups in regressions, in order to control for firm size biases.

3.5 Estimation and method

I estimate the relationships in (3.2) with simultaneous equations regressions on the full samples as well as sub-samples. The equations spell out behavioural relationships between variables. Since the equilibrium model of governance is not known, reduced form estimation is not possible (Greene, 2003, section 15.2). The equations are behavioural, but not structural in the sense of belonging to an equilibrium model.

The fixed effects methods (Woolridge, 2002) is common to all regressions. Fixed effects estimation amounts to removing the individual heterogeneity of firms contained in the fixed effect ci 12. Remember the error 12 For every individual firm, an overall average is constructed. Then, from each company observation the overall, individual average is subtracted.


term in the system (3.2) is uit, which contains the fixed effect ci and a idiosyncratic effect vit, which varies over time and companies. i refers to firm number i, and t is the time period. When demeaning the variables, the fixed effect element disappears. So does the constant term.

I use the three-stage least squares (3SLS) methodology in estimations.

The 3SLS is an instrumental variables estimation method where the instruments are the predicted values of the dependent variable in a regression on all the explanatory variables in the system (Greene, 2003, p. 398). The predicted values are found from GLS regressions, and iterations are taken until convergence is achieved. Meaningful overall measures, such as R2 in OLS regressions, are not available. Instead, I include a Wald test (Greene, 2003, p. 107) to study whether all coefficients in a given equation are zero.

The danger in simultaneous equation estimation lies in the model specification (Greene, 2003). If, for instance, a misspecification has occurred in the first equation, the mistake may contaminate all other equations as well.

To investigate if this propagation of misspecification is a serious problem, I perform several robustness tests, see below.

I perform estimations in the full sample and for sub-samples. First, the model (3.2) is estimated on the full sample with Tobin’s Q as firm performance, and then on sub-samples of co-determined and shareholder determined firms. The sub-sample tests will reveal whether results from the overall sample really apply to co-determined firms alone, or if the employee director effect is merely due to difference in sampling. I further partition the sample to include only firms with more than 200 employees, when co-determination is compulsory. This will remove firm size effects.

In robustness tests, I perform an estimation with all index variables included individually (the right hand side of (3.1)), as well as an estimation of a wider definition of the board index13, this time including nonsignificant effects in Bøhren and Strøm (2007) as well. Further robustness tests include replacing Tobin’s Q with ROA and stock return as dependent variable, and replacing leverage with the dividend payout rate. Also, I remove the lagged firm performance in order to investigate whether paramInaddition to the variables in (3.1), I include Outside owner concentration, Independence, CEO director, Exported and imported directors, and board age dispersion. Outside owner concentration is the sum of squared equity fractions across all the firm’s outside owners; Independence is the board tenure of the non-employee directors minus the tenure of the CEO; CEO director equals 1 if the CEO is a member of his company’s board and zero otherwise; Exported CEO is the number of outside directorships held by the firm’s CEO; Imported CEO is the proportion of CEOs from other companies on the board; Board age dispersion is the standard deviation of board age.

3.6. ECONOMETRIC EVIDENCE 77 eter estimates remain stable. The last robustness test is a test of the Fauver and Fuerst (2006) information hypothesis, which I interpret to mean that in information intensive industries firm performance is improved with codetermination. This regression should show if their positive employee director result is also the case in Norway.

The explanatory variables are assumed to be simultaneous with firm performance. Since board members are predominantly elected in the late spring, the new board should also have had some time to make a noticeable impact upon firm performance, measured at year-end. This assumption is reasonable given some market efficiency.

3.6 Econometric evidence

Do employee directors improve firm performance, and are governance mechanisms at least partly endogenously determined?

This section reports simultaneous regression results of the model (3.2).

I estimate for the whole sample and then turn to sub-samples of co-determined and shareholder determined companies, and for firms with more than 200 employees. All regressions are done with standardized values.

This means that comparisons of economic importance can be read off from coefficient values.

I start with estimations of the model (3.2) for the entire sample. Table

3.4 shows the estimation results.

Table 3.4

The Wald tests show that no equation supports the null hypothesis that all coefficients are zero. Comparing the two sections of the table, signs and coefficient values are very much the same. Thus, I restrict comments to the case of systematic risk in the upper section.

The co-determination hypothesis says that the employee director variable is negative to firm performance, and positive to the board index and leverage. Table 3.4 confirms this except for the leverage, where only the sign is as predicted. Furthermore, the co-determination hypothesis implies a positive impact on average wage. Here too, only the sign is conrmed. This weaker result may be due to pay being determined by external market conditions. Thus, the direct and indirect effects of co-determination are partly confirmed. Consequently, employee directors carry a negative association with firm performance, and shareholders tend to take


compensatory actions to alleviate the influence of employee directors. The board index is at least partly endogenously determined.

Are the board index and the leverage positively related to firm performance and negatively to average wage? For the board index, this is confirmed for firm performance, but only the sign is as expected for average wage. Thus, a better composed board will improve firm performance.

On the other hand, leverage is against the free cash flow hypothesis expectations in both firm performance and average wage. A higher leverage indicates a lower firm performance and higher average wage. In conclusion, the governance hypothesis is not fully confirmed.

The negative association between leverage and firm performance conrms findings in empirical studies ( Barclay et al., 1995; Rajan and Zingales, 1995; and Brick et al., 2005). I offer two alternative explanations to the free cash flow hypothesis; the fear of higher decision costs in a situation with three decision makers, that is, shareholders, employees, and banks; and the negative signalling effect of a high leverage (Myers, 1977).

Also note the complementarity between the board index and leverage (Agrawal and Knoeber, 1996). The sign is negative and significant. Thus, the two governance mechanisms are substitutes rather than complements.

The Hermalin and Weisbach (1998) reverse causation hypothesis is only partly confirmed, as the board index is positive and leverage is negative.

Lower leverage should bring lower monitoring intensity. The results are significant, indicating that good performance leads to a better board index, and to an easier debt burden. In all, endogeneity is confirmed, as both the board index and the leverage are at least partly determined from the presence of employee directors and from past performance.

Are shareholders able to neutralize the employee director by adjustments in the board index and the leverage, taking the employee director relationship to average wage into consideration as well? Since the variables are standardised to have average zero and standard deviation 1.0 in regressions, coefficients can be compared. They show that the direct effect is stronger than the indirect to the board index. For the negative direct employee director effect is now 0.119, while the indirect effect upon the board index is positive and 0.314. Since the board index is now 0.122 to firm performance, the positive, indirect impact of employee directors through the board index is only 0.038 (= 0.122 × 0.314), or 31.1% of the direct board index effect. The shareholders are able to compensate 31.9% of the negative direct effect of employee directors through adjustments to board characteristics. Furthermore, the employee director also impacts

3.6. ECONOMETRIC EVIDENCE 79 positively upon average wage, which is negatively related to firm performance. Even though the average wage is not significant in the overall sample, it is for co-determined firms, as I shortly report. The same applies to leverage. Likewise, the economic significance of the indirect effects from the lagged firm performance is very low, being 0.01 for both the board index and the leverage.

Thus, the economic magnitude of the indirect effects from employee directors or past firm performance upon firm performance is small compared to the direct effect of the board index and the leverage. Endogeneity matters, but not very much.

The volatility measure in the lower section of table 3.4 gives two interesting relationships in the board index and the leverage equations. It turns out that only leverage has the expected positive and significant sign. The Raheja (2005) theory of board composition implies that the board index is positively related to firm risk. For volatility the opposite sign obtains.

Next, the model is studied in sub-samples. If regulation plays a role, a less than optimal board composition is likely to follow. Therefore, we should observe stronger and more significant coefficients in the co-determined firms than in the shareholder determined. Table 3.5 is a report on the two sub-samples of firms.

Table 3.5

Note that the Wald test shows rejection of the null hypothesis that all variables have zero significance. Furthermore, a Chow dummy variable test rejects the hypothesis that the coefficients of the sub-samples are equal to the those in the overall sample. Thus, there is a difference between codetermined and shareholder determined firms.

In the co-determination sub-sample the employee director effects are even more pronounced than in the overall sample. The negative employee director impact upon firm performance is about 45% higher than in the overall sample and the indirect effect on the board index increases even more. Now, the employee director variable is significant in relation to leverage and to average wage. Thus, the co-determination hypothesis is even more strongly confirmed in the sub-sample of only co-determined firms than in the overall sample.

Pages:     | 1 |   ...   | 11 | 12 || 14 | 15 |   ...   | 22 |

Similar works:

«Wesham During the Great War by Jennifer Ray A thesis submitted in partial fulfilment for the requirements for the degree of MA (by Research) at the University of Central Lancashire March 2014 Chapter 1 Introduction and Literature Review The Great War has, since its conclusion, been the topic of extensive study. Initially, research focused on the military aspect of the conflict. However, during the 1960s, studying the social effect of the conflict developed as a distinct genre of research. More...»

«African Journal of Accounting, Economics, Finance and Banking Research, Vol. 2. No. 2. 2008. Dulacha Barako and Peter Gatere OUTSOURCING PRACTICES OF THE KENYAN BANKING SECTOR Dulacha G. Barako Curtin University of Technology, Australia Dulacha.Barako@cbs.curtin.edu.au Peter K. Gatere Central Bank of Kenya, Kenya ABSTRACT This paper presents contemporary evidence of outsourcing practices of the Kenyan banking sector. A questionnaire was sent to forty commercial banks operating in Kenya....»

«Urban road pricing acceptance Björn Hårsman Inregia 2001-11-11 To be presented at the IMPRINT-EUROPE seminar in Brussels November 21 and 22, 2001 Contents 1. Introduction 2 2. Lessons from economic theory 3 3. Surveys and interviews 8 4. Experience with road pricing 11 5. Conclusions 15 References 17 1 1. Introduction At noon on Thursday February 1, 1990, a cordon toll system opened in the city of Oslo. At the same time a road tunnel, financed by the expected toll revenues, was opened under...»

«July 15, 2015 www.citizen.org Financial Services Conflict of Interest Act Outlining the need for increased revolving-door and reverse revolving-door legislation Acknowledgments This report was written by Craig Holman, Ph.D., government affairs lobbyist, and Emma Stein, researcher, Public Citizen’s Congress Watch division, and edited by Lisa Gilbert, director, Public Citizen’s Congress Watch division. About Public Citizen Public Citizen is a national non-profit organization with more than...»

«THESES OF DOCTORAL (PhD) DISSERTATION ZOLTÁN SZABÓ KAPOSVÁR UNIVERSITY FACULTY OF ECONOMIC SCIENCES 2015 KAPOSVÁR UNIVERSITY FACULTY OF ECONOMIC SCIENCE Department of Marketing and Trade Head of doctoral school: Prof. Dr. SÁNDOR KEREKES DSc Doctor of the Hungarian Academy of Sciences, Professor Supervisor: Dr. habil. Viktória Szente PhD associate professor Co-supervisor: Prof. Dr. habil. József Kocsondi CSc university professor MEDICAL WATER BASED SPA TOWN DEVELOPMENT POSSIBILITES...»

«The Impact of Africa on China Jing GU and John HUMPHREY Institute of Development Studies Brighton, UK. 1 1 Introduction No-one can doubt the increasing importance of China in Africa, or the impact that its growing presence has had on both African politicians and also aid donors. The presence of so many African leaders at the November 2006 Beijing Summit is proof of the former. The impact on donors arises from the rapid rise in China's trade with the continent, the increasing activity of Chinese...»

«Kansas City Southern In The De Ramus Era Not, for best 8 amount will apply two companies bottom-line and better. Your partnership will make the Kansas City Southern in the DeRamus Era part of concept and particularly further. They is such in I are ironically your trademark worth is out organization loans them would use. Of a much day really of the change even is ahead of your areas. Capitalize you as the worry advertising government of corporations if you, and gain of money gets up. Now, they...»

«HOLLIDAYSBURG BOROUGH COUNCIL MEETING HELD IN THE COUNCIL CHAMBERS OF THE MUNICIPAL BUILDING THURSDAY, OCTOBER 11, 2012, AT 7:00 P.M. The meeting was called to order at 7:00 P.M. by President John P. Stultz, Jr., who presided. Present: Council Members Joseph A. Pompa; Timothy A. Beresnyak; Mark L. Shawley; Timothy M. Baranik; and Richard A. Scholton. Also present: Borough Manager Mark G. Schroyer; Mayor Joseph R. Dodson; Attorney Nathan Karn; Finance Director James Gehret; and Borough Secretary...»

«Nieves Pérez-Solórzano Borragán EU Accession and Interest Politics in Central and Eastern Europe ABSTRACT EU-related issues and patterns of behaviour are permeating the new member states’ domestic environments.The literature assessing the domestic impact of the Eastward Enlargement focuses on the institutional capability of the new member states to meet the accession criteria as defined in Copenhagen in 1993. This article expands the remit of the debate on Europeanisation in the new member...»

«Financing Firms in India∗ Franklin Allen Rajesh Chakrabarti Sankar De Finance Department Finance Area Centre for Analytical Finance The Wharton School Indian School of Business Indian School of Business University of Pennsylvania Hyderabad 500 032, India Hyderabad 500 032, India allenf@wharton.upenn.edu rajesh_chakrabarti@isb.edu Sankar_De@isb.edu Jun “QJ” Qian Meijun Qian Finance Department Finance Department Carroll School of Management NUS Business School Boston College National...»

«Preface Two years ago, the Grace, Kennedy Foundation Lecture was a dream. The many and varied problems of the Caribbean region were crying out for constructive analysis leading hopefully to viable solutions. The Directors of this Foundation felt that if some of their resources were devoted to serious consideration of major issues by persons competent to evaluate them, their money would be well spent. It is now no longer a dream. The first GFK Lecture was delivered in 1989 and its author, the...»

«At a glance The near collapse of the world economy triggered Regulatory reform and the role regulatory reforms with the of internal audit power to fundamentally change how financial services organizations do business. As the financial services industry responds to an unprecedented level of evolving regulatory reform, internal audit functions must adapt to meet the challenge. Today's environment presents internal audit with a unique opportunity to position itself as a trusted advisor to the...»

<<  HOME   |    CONTACTS
2016 www.dissertation.xlibx.info - Dissertations, online materials

Materials of this site are available for review, all rights belong to their respective owners.
If you do not agree with the fact that your material is placed on this site, please, email us, we will within 1-2 business days delete him.