Quantitative Paradigm

Germann, Cornel

doi:10.1007/978-3-658-40817-6_7

Cornel Germann²

786 Accesses

Abstract

Chapter 7 refers to the quantitative research instrument, description of survey sample, research results, and limitations.

You have full access to this open access chapter, Download chapter PDF

Chapter 7 elaborates on the quantitative paradigm of the dissertation. Keeping in mind succession planning theory (Chapter 4 ) and qualitative insights (Chapter 6), the research instrument and the description of the sample are first explained. To familiarise the reader with the results, the quantitative insights are then analysed and their limitations are demonstrated.

7.1 Research Instrument

The next chapter presents detailed information on the survey method (web-based survey), its application, reliability, and validity, as well as the theoretical-methodological context relevant to the presentation of the sample data.

7.1.1 Web-Based Survey

In the sphere of the board, the survey method is a common way of gaining access to the boardroom. It allows to analyse current trends and dynamics with a manageable time and effort for the participating organisations and people (Boyle et al., 2016, p. 402). Therefore, the survey method is suitable for tracking readiness for board practices or identifying new trends.

The survey was web-based and contained closed questions (Bell et al., 2019, pp. 232–233). Applying the two-dimensional survey format has several advantages (summarised in the following) (Rea et al., 2014, pp. 12−13, 51−55): On the one hand, with a web-based technique, the questionnaire can be completed at the respondent’s convenience without time constraints. Since there is no personal contact between the respondent and the researcher, anonymity can be ensured and the wording of the questions can be generalised, allowing the questionnaire to be longer and more complex. It also allows the use of visual aids that may encourage the respondents to make more prominent statements. On the other hand, the questions in a purely closed format tend to be clearer and more comprehensible in contrast to a social survey with an open questionnaire. It also makes a reduction in complexity and survey duration possible (~ 10 – 15 minutes). Moreover, the uniform character of closed questions facilitates the analytical comparisons of the results.

7.1.2 Quantitative Application

The survey was performed from October to December 2021. The specific time period is aligned with that of other board research studies (e.g. Huse, 2009b). The underlying timing was chosen as it is the time of the year with least board activity, in the hope that this would increase the response rate. Participants were asked personally by email to take part in the study. The email addresses were known from interview results, could be identified through own research, or were provided with the help of independent gatekeepers (Section 5.3.2).^{Footnote 1} For persons whose email address could not be identified, the investor relations department was contacted with a request to forward the survey. As in the interview process, the participants were informed that the results would be treated anonymously and confidentially.

As explained in Section 5.3.2, Table 7.1 shows that the primary target group for the web-based survey was the head nomination committee and the chairperson. The initial sample consisted of 215 chairpersons and 149 nomination committee chairs, with a total of 365 persons identified.^{Footnote 2} From that, 32 board interlocks (chair and/or chair nomination committee) and 33 dual mandates (chair equals chair nomination committee) were excluded, leading to an adjusted sample population of 299.

At the end of the survey period, 87 chairpersons and heads of nomination committees had successfully completed the survey. After analysing the data set, seven adjustments were made due to missing values (listwise deletion). Therefore, the final sample consisted of 80 participants with a response rate of 27%. The response rate was thus above the average range of 10 – 25% for high-level professionals, with the lower limit more likely for board chair and head nomination committee as data access is more difficult (Huse, 2009b, pp. 367–368).

Table 7.1 Survey – Sample and Response Rate

Full size table

Following previous board survey studies (e.g. Clune et al., 2014; Nicholson et al., 2006), the closed and web-based survey consisted primarily of five-point Likert scale questions or a subset of predetermined response options. In contrast to three- or seven-point scale formats, as exemplified in Devlin et al.’s (1993) study, the five-point scale appears to be less confusing and its scale descriptions easier to read.^{Footnote 3} It is also thought to be more suitable for European survey studies (Bouranta et al., 2009, p. 280). More importantly, however, the choice of a five-point scale made it possible to compare the results with other board studies and tailor the questions specifically to the target group without (overly) limiting the range of options. To meet the respective level of measurement (Bernard, 2017, pp. 45–48), the survey questions also consisted of nominal variables for demographics and ordinal variables for participants’ opinions (see appendix A7 in the electronic supplementary material).

Overall, the questionnaire consisted of eight dimensions: (1) introduction (information on the research project), (2) preliminary remarks (information for the survey and first question); (3) organisational demographics (industry, market capitalisation, ownership); (4) succession planning (institutionalisation, process management, key contingencies); (5) stakeholder management (key stakeholders); (6) competences (roles, profile, experience); (7) voluntary disclosure (transparency, strategy, channel, time/timing); and (8) participant demographics (gender, age, position, experience).

7.1.3 Quantitative Reliability and Validity

In quantitative studies, quantitative reliability refers to the accuracy and consistency of a study (Cohen et al., 2007, p. 146). According to Golafshani (2015, p. 599), study results can be considered reliable if the researcher reaches the same conclusions when conducting several more attempts. Quantitative validity specifies how accurate the results are and/or whether the underlying survey questions truly measure what is intended (Groves et al., 2009, p. 274). Quantitative construct validity thereby refers to the interplay between data and construct, which is obviously key for validation purposes (Golafshani, 2015, p. 599; Wainer & Braun, 1988, p. 25).

To check for quantitative reliability and validity, the survey was based on the theoretical and qualitative insights gained earlier in this thesis. To avoid misunderstandings and ambiguities, questions were pretested by an academic supervisor, a co-worker, a member of the board of directors, and a company secretary, and their feedback was incorporated (Collins, 2003, p. 231). The purpose of pilot testing was to ensure clarity (questions are understandable), comprehensiveness (questions are relevant), and acceptability (questions meet methodological/ethical standards) (Rea et al., 2014, p. 38). By pretesting and including the results from the extensive literature review and the qualitative insights, it can be expected that the questionnaire covered all relevant aspects for the research questions under study and that the quantitative reliability and validity criteria have been met.

7.1.4 Methodological Analysis

Whenever general findings for a population are to be quantified from a specific sample, a researcher employs a descriptive and/or inferential statistic approach (summarised in the following) (Fitzgerald et al., 2001, p. 287): Descriptive statistics, on the one hand, present the fact or distribution of the studied phenomena and show the exact, unfiltered responses of the participants. This allows the generalisability of the results to be studied and initial conclusions to be drawn, thereby serving as a point of reference for additional statistical analyses. Inferential statistics, on the other hand, assess the significance and strength of the relationships between variables. It examines the (non-)existence of relationships within the sample relative to its population. Depending on the underlying assumptions (distribution, homogeneity, interval/ratio/ordinal data), scholars either apply parametric or non-parametric techniques.^{Footnote 4}

Responding to the research questions (Section 5.2) and the survey’s data parameters (abnormal distribution, non-homogeneity, ordinal data), this thesis first applies descriptive statistics and then performs a Kruskal-Wallis H and a Dunn-Bonferroni post hoc test. While the descriptive approach summarises the central tendencies and variabilities, the non-parametric method allows to identify any discrepancies between the group respondents (Corder & Foreman, 2014, p. 118).

The non-parametric group comparison approach draws on the interviews (Chapter 6) and on chair studies (e.g. Dulewicz et al., 2007), which found that stakeholder and shareholder demands vary and thus affect succession planning. Following Farah et al. (2020, p. 2), who urged to study leadership succession in different organisations/types of contexts, the three underlying pairs of groups relate to market capitalisation (size), shareholder structure (block holdings), and board experience (proxy for intergenerational disparities). In contrast to the survey questions posed (up to 6 category items), the number of items within the groups is reduced given they survey’s distribution (Section 7.2) and theoretical-logical relationships. The group sample variables are thus as follows (where N = number of observations across groups):

Market capitalisation:	Group 1 (N = 36) = Less than CHF 2,001 million
	Group 2 (N = 23) = CHF 2,001 – 10,0000 million
	Group 3 (N = 21) = More than CHF 10,000 million
Shareholder structure: ^{Footnote 5}	Group 1 (N = 21) = Individual(s)/family
	Group 2 (N = 25) = Institutional investor/activist
	Group 3 (N = 16) = Government/(pension) fund
Board experience:	Group 1 (N = 34) = Less than 10 years
	Group 2 (N = 29) = 10 – 20 years
	Group 3 (N = 17) = More than 20 years

For the descriptive approach, the absolute values (in %) are reported. Insights were provided by visual charts and specific explanations. To address the research questions and strengthen the results, if applicable, constructs from single items were aggregated (absolute values subsumed and then expressed in %). Construct formation was done according to the previous theoretical insights gained earlier. For the non-parametric approach, Kruskal-Wallis’ H-value, Dunn-Bonferroni’s p-value, and Pearson’s r-value are described in more detail below.

The Kruskal-Wallis H test is the non-parametric alternative of a one-way analysis of variance (ANOVA) (Kruskal & Wallis, 1952, pp. 584–585). It compares the (in)equality of two or more independent groups on an ordinal scale. By examining the medians of the population (θi), Kruskal-Wallis’ null hypothesis (H_O) asserts that there is no statistical difference between the groups under study (whereby the significance level is given by the chi-square (χ²) distribution). Calculating the Kruskal-Wallis statistic (H) and the degrees of freedom (df) requires sorting the data in ascending order, allocating ranks to the data points, and summing the ranks of the groups under study. To determine the H-statistic, the following formula is used, where N = number of observations across groups, R_i = sum of the ranks in the group sample, n_i = group sample size, and k = number of groups (Corder & Foreman, 2014, p. 118):

$$H = \frac{12}{{N\left( {N + 1} \right)}}\mathop \sum \limits_{i = 1}^{k} \frac{{R_{i}^{2} }}{{n_{i} }} - 3\left( {N + 1} \right)$$

$$df = k - 1$$

The H test expects the measured value to be unique. If, however, two or more survey participants have identical ranking score values, a tie correction must be applied. Under those circumstances, the initial value H is divided by the tie correction value C_H, where T = number of tie values and N = number of observations across groups (Corder & Foreman, 2014, p. 118):

$$C_{H} = 1 - \frac{{\sum \left( {T^{3} - T} \right)}}{{N^{3} - N}}$$

Kruskal and Wallis find that there is a difference between two samples (Weaver et al., 2017, p. 354), but do not identify the specific sample or the magnitude of the difference (Beatty, 2018, p. 82). Thus, in order to control for type I error, i.e. to reject the null hypothesis if it is true, and to identify the significant sample pairs, a Dunn-Bonferroni correction (Dunn, 1961) and a Dunn-Bonferroni multiple comparison (Dunn, 1964) are performed as post hoc tests.^{Footnote 6}

For controlling the type I error, the correction method basically adjusts the p-value used for post hoc analysis, where α_B = adjusted risk level, α = initial risk level, and k = number of groups (Corder & Foreman, 2014, p. 119):

$$\alpha_{B} = \frac{\alpha }{k}$$

For identifying the significant sample pairs, the multiple comparison assesses stochastic dominance between certain sample groups using the z-statistic. To draw statistically significant conclusions, the test examines standardised absolute mean differences of the group ranks (exemplified by group A and B) (Dinno, 2015, p. 298):

$$Z_{i} = \frac{{y_{i} }}{{\sigma_{i} }}$$

The nominator $Y_{i}$ with $\overline{R}_{A}$ and $\overline{R}_{B}$ corresponds to the mean ranking from the previous Kruskal-Wallis results, where R_A or R_B = sum of group sample ranks and n_A or n_B = sample size for group A or B.

$$Y_{i} = \overline{R}_{A} - \overline{R}_{B}$$

$$\overline{R}_{A} = \frac{{R_{A} }}{{n_{A} }}$$

$$\overline{R}_{B} = \frac{{R_{B} }}{{n_{B} }}$$

The denominator $\sigma_{i}$ relates to the standard deviation of $Y_{i}$, where N = number of observations across groups, T = number of tie values, and n_A or n_B = sample size for group A or B.

$$\sigma_{i} = \sqrt {\left\{ {\frac{{N\left( {N + 1} \right)}}{12} - \frac{{\sum \left( {T^{3} - T} \right)}}{{12\left( {N - 1} \right)}}} \right\}} \left( {\frac{1}{{n_{A} }} + \frac{1}{{n_{B} }}} \right)$$

In order to evaluate the statistically significant group differences observed, its effect size is also calculated. For that reason, using Pearson's correlation coefficient (r) is suitable, as its effect size is always between 0 (minimum effect) and 1 (maximum effect) (Bell et al., 2019, p. 323). By applying Cohen's (2013, p. 82) definition for the r-value, the strength of the effect can be further categorised into a weak (r = 0.10), medium (r = 0.30), and strong effect (r = 0.50), where z_i = Dunn-Bonferroni multiple comparison z-statistic results and n_A or n_B = sample size for group A or B:

$$r_{i} = \left| {\frac{{z_{i} }}{{\sqrt {n_{A} + n_{B} } }}} \right|$$

Overall, to support and develop the descriptive and non-parametric statistics explained, proven survey (Unipark) and statistical software tools (SPSS) were used.

7.2 Description of the Survey Sample

The core of the overall survey sample studied are the 80 participants. As the survey analysis is influenced by its participants, it is necessary to explain their characteristics accordingly − before the analysis. In order to approach that appropriately, the sample is descriptively divided into person-related and organisation-related demographics.

7.2.1 Person-related Demographics

From the 80 final participants, the majority of respondents were male (87%). This indicates a seven times higher proportion of men than women, who accounted for 13% of the total sample, and emphasises the ongoing discussion on gender diversity, including the chair position (Figure 7.1).

In terms of age, not a single participant was younger than 40. Approximately 4% fell into the category between 40 – 49 years, 40% into the range between 50 – 59 years, and the majority with 56% were 60 years and older. On the one hand, the fairly high age is advantageous as seniority indicates experience in processing succession planning, but on the other hand, it can also be disadvantageous, as it can represent the paradigm of the old boys’ network (Allemand et al., 2022, p. 786).

Most survey participants were chairpersons (61%), followed by nomination committee chairs (28%), and persons simultaneously serving as board chair and nomination committee chair (5%). In the category ‘others’, participants indicated a variety of options (and/or): designated chair, vice-chair, nomination committee member, strategy committee member, audit committee chair, and ordinary board member. Overall, the overview of functions emphasises that all respondents held at least one function on a board of directors (Figure 7.2).

Overall, cumulative professional board experience was high. People with less than 5 years of experience were the least represented group in the sample with 14%, whereas the most represented group with 29% ranged anywhere between 6 – 10 years of experience. Those with 11 – 15 years (19%), 16 – 20 years (18%), and more than 20 years (21%) were equally distributed. Overall, it can be summarised that the group of individuals in the sample is experienced. Professional board experience can thus be considered as highly relevant for the survey’s target group. It is worth noting that cumulative board experience refers exclusively to years of experience in listed organisations (Figure 7.3).

7.2.2 Organisation-related Demographics

For industry affiliation of the 80 respondents (Figure 7.4), industrials (25%), financials (24%), and healthcare (20%) are among the top three represented industries in the sample, adding up to more than two-thirds of the overall sample (69%). They are followed by real estate (8%), energy (6%), information technology (5%), and consumer discretionary (5%). Less distinct and below the 5% trigger level are materials (3%) and communication services (3%). Despite the strong influence of the top three, the sample can be considered diverse and representative of the overall economy. With regard to financials (24%) and healthcare (20%), it should also be noted that these two sectors are more closely monitored by the regulator compared to others.

An overview of market capitalisation in Figure 7.5 reveals that the lion’s share of the sample organisations were small-caps (35%), mid-caps (29%), and large-caps (19%). Less represented were micro-caps (9%), mega-caps (8%), and nano-caps (1%). The high occurrence of mega-cap bodies in relation to the absolute number of organisations present in Switzerland can be seen as an indicator that chair succession is of practical relevance. Moreover, the sample distribution of market capitalisation shows that chair succession also matters for small- and mid-sized organisations.

The demographics of ownership structure show that almost two-thirds (63%) of all sample organisations have block holdings with an equity share of more than 20% (Figure 7.6). The high percentage of block holdings is a typical representation of the Swiss business environment, which has historically been characterised by a low level of dispersed ownership (Section 4.3.2.4). Further holdings are 1 – 5% (16%), 6 – 10% (10%), and 16 – 20% (6%). Less emphasised are the categories 11 – 15% (4%) and less than 1% (1%). It is important for further analysis to keep in mind that the sample represents a high level of block holdings, which may differ when comparing results from a dispersed ownership environment.

Focusing on the type of shareholders in the sample, it shows that individuals with 49%, institutional investors with 30%, and the government with 18% are the three main shareholders. It is interesting to see that (pension) funds (3%) and activists (1%) have low representation in the sample. Considering the high level of block holding as shown in Figure 7.6, it is no surprise that individuals and institutional investors are the most represented in the sample, as especially the latter have gained influence over the last twenty years (Section 4.3.2.4). The balanced ownership structure will make it possible to highlight the different stakeholder views on succession in the further course of this dissertation (Figure 7.7).

The high ratio of ownership is also reflected in board representation (Figure 7.8). Two-thirds of all sample organisations have stakeholder representatives at board level while one-third do not. In terms of formally establishing a nomination committee, 92% of all organisations have one in place, with only 8% choosing otherwise. It can thus be presumed that organisations that have already established a formal nomination committee have greater interest in the proper implementation of succession planning.

7.3 Introductory Results

Beyond the pure purpose of analysing the three research questions, it is important to recognise and understand how professionally succession planning is developed/acknowledged in Swiss boards. At the beginning of the survey, participants were thus asked to provide further opinions specifically on the chair and current best practice background to check whether they are generally accepted/neglected. Also pursuing this aspect allowed for a clearer distinction between ordinary board member and chairperson succession.

7.3.1 Descriptive Statistics

To begin with, survey contributors were asked to reflect on their last chairperson succession process and whether they felt it was managed adequately (Figure 7.9). Overall, 87% of all experiences were positive, with only 13% of respondents saying that improvements were necessary. It is interesting to see that critical answers are distributed relatively equally with 12% of chairs (6 vs. 43) and 14% of nomination committee chairs (3 vs. 19).^{Footnote 7} From that viewpoint, the result serves as a first indication that there is potential for development and that the survey is free of functional bias (chairperson vs. nomination committee chair).

When it comes to the question of who owns the succession process, opinions differ in theory and practice. Best practice principles call for a strict separation, i.e. that the chair should not lead his or her own succession. However, practice shows that the percentage of organisations where the chair still manages the succession process is quite high, at 30% (Section 2.1.3). A similar tenor confirms the survey result (Figure 7.10): Despite the fact that the majority (52%) voted for the head of the nomination committee, thus confirming the governance principles that the process owner should be unbiased, almost a third (32%) of all participants still favoured the chairperson as process leader. Although the election of the vice-chair is seen as an attractive option, especially when there is no formal nominating committee, only few voted for the vice-chair (4%). Other opinions on the choice of process owner included: all members of the nomination committee (6%), all members of the board (3%), and others (3%), with the latter referring to the senior independent director, the main shareholder, and the government council.

Since the professionalism of chair succession depends on organisational integration, participants were asked to assess the institutionalisation of succession planning in their organisation. Although the formal establishment of a nomination committee is the main indicator for an adequate organisational integration (Section 2.1.3), agenda setting, strategic long-term planning, and systematic profiling play an equally significant role.

Overall, the descriptive results in Figure 7.11 show that the majority of organisations (+48%) strongly agree that succession planning is always included on the nomination committee agenda. First, confirming modern team production theory, organisations are aware that human resource planning at top board level is key and needs to be addressed primarily long-term, yet without neglecting short-term changes. As recent examples with Doris Russi (Helvetia), Gert de Winter (Baloise), and Patrick Frost (Swiss Life) have demonstrated (not necessarily just chair-related), it is also possible for top performers there to resign/take a break for private/health reasons on short notice (Gerber, 2022). Second, the importance of human resource planning is also supported by the fact that only organisations that do not have a nomination committee strongly disagree (4%) with the inclusion of the succession issue in the agenda (which is a logical conclusion as then, no planning can take place at committee level). The result thus underscores the need for accurate planning. Third, pointing towards The Competing Value Framework, survey participants confirmed that systematic profiling/calibration of competences is necessary to meet today’s multidimensional demands (55% for ‘strongly agree’ and 26% for ‘agree’). Organisations are thus eager to minimise subjective, network-driven processing.

With respect to the scope of succession (Figure 7.12, first two questions) and the origin of the candidate (Figure 7.12, last two questions), the literature review and the interviews identified several different practices for chairperson succession. Referring to the comprehensiveness of the succession process based on survey responses, the majority of participants ‘strongly agree’ (64%) and ‘agree’ (25%) that chair succession is a comprehensive process. From the 89% of views, one can infer that chair succession requires a higher level of attention as opposed to ordinary member succession (Maitlis & Christianson, 2014, p. 80). Reasons for this are manifold, but the chair is assumed to have a high representative function, the duty to provide future strategic direction, and a key influence in determining overall board composition (Withers & Fitza, 2017, p. 1345). The second question is closely related to the first. In order to meet the multidimensional tasks (Sections 3.2 and 4.2.1), the chair must have a broader range of competences than ordinary board members (a total of 80% for ‘strongly agree’ and ‘agree’). It can thus be concluded that it is insufficient for the chair to be specialised in only one single competence dimension (e.g. the financial expertise of the chair of the audit committee). The role of the chair requires a variety of competences.

The last two questions in Figure 7.12 related to the origin of the chairperson. Under normal business conditions (no crisis but strategic continuity), participants slightly favour an internal board member as chair successor (a total of 52% ‘strongly agree’ and ‘agree’). The results with 26% ‘neutral’, 14% ‘disagree’, and 9% ‘strongly disagree’ show different opinions but are aligned with the qualitative findings (Section 6.3). Further, when organisations have decided against an internal candidate and thus favour an external candidate, the majority of participants first advocate a transition year as an ordinary board member before taking over the chair (36% ‘strongly agree’ and 31% ‘agree’). Only a few express a different opinion with 1% ‘strongly disagree’ and 11% ‘disagree’, confirming the emerging trend towards a transition year, long-term planning, and best practices for succession.

For substantiating the results, the answers on the origin of the candidacy were additionally checked for the influence of industry affiliation (Figure 7.13). For internal origin (left part of the graph), participants from the financial, health care, and industrial sectors agreed most strongly with an internal nomination (between 20 – 25% for ‘strongly agree’). For external origin/transition period (right part of the graph), the strongest support also stems from the same three industries, confirming the priority for internal origin (between 10 – 25% for ‘strongly agree’). From this point of view, in accordance with Hillman et al. (2000, p. 243) and Roberts (2002, p. 449), it seems that sensitive, complex, and more regulated industries are more likely to ensure that the person gets to know the organisation, the board, and the internal procedures thoroughly before taking full responsibility. However, results are not conclusive and should be interpreted with caution as the three industries also represent a majority in the sample (Section 7.2.2).

To deepen the understanding of the sample, participants were further asked to provide insights on the roles that the chairperson has to perform. Understanding the roles beyond those of the full board enhances the need to derive the board chair’s competences and involvement in succession planning. The questions for the answers in Figure 7.14 were based on Sections 3.2 and 4.2.1. Validating the strategic visionary, coaching, and initiating roles from prior studies on Swiss chairs (e.g. Lorenz Koller, 2010; Wunderer, 1995), the survey participants prioritised (1) entrepreneurial, goal-oriented strategic tasks (compete), (2) coaching and peer exchange (collaborate), (3) supervision and information distribution (control/govern), and (4) organisational development (create).

7.3.2 Inferential Statistics

Beyond the descriptive statistics, there are likely to be benefits/disadvantages of the group sample that affect the responses differently (as highlighted in Figure 7.13, for example). Overall, the non-parametric results for Kruskal-Wallis (KW) and Dunn-Bonferroni (DB) in Table 7.2 show few statistical differences in the structure of group responses related to market capitalisation, shareholder structure, and board experience. On the positive side, there is little discrepancy between participants, indicating that succession planning is being approached professionally.

Reminder: group definitions for the Dunn-Bonferroni (DB) multiple comparison post hoc test
*Market capitalisation*	*Shareholder structure*	*Board experience*
Gr1: < CHF 2,001m	Gr1 : Individual(s)/family	Gr1: < 10 years
Gr2 : CHF 2,001 – 10,000m	Gr2 : Inst. Investor/activist	Gr2: 10 – 20 years
Gr3: > CHF 10,000m	Gr3 : Government/(pension) fund	Gr3: > 20 years

Table 7.2 Survey – Introductory Non-parametric Results

Full size table

Market Capitalisation

Market capitalisation shows significant group differences for candidate profiling (3) and role collaboration (10). First, small-cap organisations do not seem to consider the candidate profile as important as mid-cap organisations (H = 4.695, group 1 vs. 2, p < 0.10, r = 0.28), whereby the correlation between the two groups is small with a significance level of only 10%. One reason for this result could be the limited resources of small-caps to address the multidimensionality of the chair role adequately (Olson & Adams, 2004, p. 442). Compared to mid-caps, small-caps have only limited financial and human capacities to thoroughly set up a candidate profile (e.g. time and support). Second, assuming that small-caps are less in the spotlight (Gallego Alvarez et al., 2008, p. 606), they seem to perceive the role of challenging management and exchanging ideas with peers to be less important compared to mid- and large-caps (H = 10.449, group 1 vs. 2, p < 0.05, r = 0.36; H = 10.449, group 1 vs. 3, p < 0.05, r = 0.34). This implies that chairpersons of organisations with a market capitalisation of below CHF 2bn are less expected to engage with stakeholders and represent the organisation on key platforms (interviews, conference keynotes, etc.).

Shareholder Structure

With regard to shareholder structure, the only discrepancy between the groups was seen in candidacy profiling (3). There, assuming that usually more parties are involved and thus require more transparency and rigor (Schepker et al., 2018, p. 525), state-controlled organisations emphasised the need for systematic profiling more strongly (H = 5.976, group 2 vs. 3, p < 0.05, r = 0.38). Although it was expected that the demand for adequacy and professionalism would be intensified by activistic tendencies of major shareholders (Huse, 2007, p. 72), institutional investors rated systematics in profiling and assessment lowest among the three groups under study.

Board Experience

In terms of board experience, the inferential statistics showed that participants with less than 10 years compared to participants with 10 – 20 of board experience were less likely to agree that a chairperson needs a wider range of competence categories (H = 11.584, group 1 vs. 2, p < 0.01, r = 0.43). Looking at the role vs. group fit perspective, which requires a chair having complementary experience to challenge board members (Section 4.1.2), the statistical discrepancy can be interpreted as an indicator that the less senior board members have not yet experienced all the cycles and dimensions of a board chair position (Helfat & Martin, 2015, p. 1286). This argumentation would emphasise the generally rather high seniority of chairpersons, also reflected in the 39% of survey respondents who have more than 16 years of board experience in listed organisations (Figure 7.3, Section 7.2.1).

7.4 RQ1 Results

To answer RQ1 (input-process-output), survey participants were asked to either rank or (dis)agree with the underlying competence dimension. By pursuing the twofold approach, (positive) excess response biases should be intercepted and thus competence classifications made accordingly (see Section 7.7). In line with the methodological procedure (Section 7.1.4), first the descriptive statistics and then the inferential statistics will be highlighted.

7.4.1 Descriptive Statistics

To begin with, Figure 7.15 reflects the overall results of the dyadic relationship between the competences of a chairperson (Section 4.2.2). The first graph (above) shows the ranking of the five competence dimensions when participants were explicitly asked to rank/prioritise them (from highest to lowest).^{Footnote 8} The second graph (below) reflects the aggregated responses regarding all individual competence items. As summarised in the following, the results for the ranking (graph above) partially contradict the theoretical hierarchical pyramid structure established by Garavan and McGuire (2001, pp. 151–152) and Spencer and Spencer (1993, p. 11):

(1) Personal competence has the highest priority among all aggregated individual aspects (98% ‘very important’ and ‘important’), but comes second when respondents were asked for a concrete ranking (rank #2); (2) social competence seems to be at the bottom of the pyramid rather than at the top, both in ranking (rank #4) and in aggregated dimensions (77% ‘very important’ and ‘important’); (3) leadership competences are ranked as the most important competence by 39 out of 80 participants (rank #1), but overall only as the third most important among the aggregated items (92% ‘very important’ and ‘important’); (4) business competences are the second most important in the aggregated dimensions (95% ‘very important’ and ‘important’) and ranked #3 by the respondents when asked specifically; and (5) technical competences play the least important role (rank #5) and are the lowest aggregated dimension (45% ‘very important’ and ‘important’).

A closer look at the components of the aggregated dimensions revealed that personal competences are deemed very important for the chair (Figure 7.16). This is shown by the meaningful result where only 3% of the participants chose ‘of average importance’ for personal competences (Figure 7.15, below). Strikingly important is integrity, which confirms societal discussions about the need for boards to behave with more integrity (Baches, 2021). Of all participants, 94% considered it ‘very important’. The second and third most important dimensions were authenticity (75% ‘very important’ and 24% ‘important’) and engagement (55% ‘very important’ and 45% ‘important’). For participants, as for others in this field, two important key competences for living up to the chair role are that the chairperson gives his or her best (authenticity) and goes the extra mile (engagement). Of all four competences, curiosity is the least pronounced, with the majority identifying it as ‘important’ (60%).

In terms of social competences (Figure 7.17), the overall low number of ‘not important’ answers illustrates the value of the chair’s social skills for performing representative functions (Pasaribu, 2015, p. 32). First, cognitive empathy rated highest with 39% ‘very important’ and 49% ‘important’, confirming that it is an important factor for the quality of stakeholder exchange (Mahsud et al., 2010, p. 570). Second and third in social competences in terms of communication expertise and cultural affinity are verbal eloquence (82% ‘very important’ and ‘important’) and cultural awareness (79% ‘very important’ and ‘important’). Network orientation as the fourth dimension, on the other hand, played a minor role despite its importance for access to resources (Section 4.2.1). Here, the majority of participants chose ‘important’ (45%) and ‘of average importance’ (35%).

When it comes to leadership competences (Figure 7.18), the results subject to the chairperson suggest that steering and controlling skills are a top priority for a chairperson. In accordance with Gabrielsson et al. (2007, p. 31), a look at the ranking of leadership competences highlights the critical questioning attitude (stamina, 59% ‘very important) and the ability to collaborate, but to lead in an inclusive way (stewardship, 58% ‘very important’). With regard to collaboration and consensus building in teams and identifying and seizing visions for the organisation, team play (48% ‘very important’ and 46% ‘important’) and visionary thinking (44% ‘very important’ and 41% ‘important’) are ranked as the third and fourth leadership dimensions. The latter is the only variable that respondents rated as ‘not important’ (3%).

In terms of business competences (Figure 7.19), survey partakers gave a high rating to strategic thinking (79% ‘very important’ and 21% ‘important’) and analytical thinking (46% ‘very important’ and 50% ‘important’). Both belong to the core tasks of the chair and the board (Section 2.1.2). As the 60% ‘very important’ and the 31% ‘important’ demonstrate, the chairperson’s crucial meta-cognitive skills for decision-making depend strongly on board literacy (Nordhaug, 1998, p. 11). The 3% ‘not important’ viewpoints symbolically indicate the crucial role board literacy plays for the chair. Somewhat less pronounced is disturbance handling, referred to as conflict and solution management skills, with 45% each for ‘very important’ and ‘important’.

For technical competences (Figure 7.20), the characteristics are more spread across the 5-point Likert scale and less pronounced overall. Here, human resources (63% ‘very important’ and ‘important’), audit and finance (62% ‘very important’ and ‘important’), and ESG (58% ‘very important’ and ‘important’) are the three top-rated functional skills of the chair. Less important are technology (39% ‘very important’ and ‘important’), legal (36% ‘very important’ and ‘important’), and marketing (18% ‘very important’ and ‘important’).

As the analysis indicates, respondents are somewhat critical in their assessment of legal competences (4% ‘not important’ and 25% ‘of little importance’). This is probably related to the fact that the nomination committee can count on the company secretary’s support in administrative and legal affairs for succession planning on the board (Section 4.3.2.2).

As the position of the chair is often occupied by senior individuals, the survey additionally assessed whether participants consider previous experience necessary and if they do, what kind of experience (Figure 7.21). Confirming the view that experience in an organisation with multiple business units is valuable for qualifying as a chairperson and results in higher announcement returns (Meyerinck et al., 2016, pp. 209–210; Spencer Stuart, 2011, section 2), the survey participants ranked board experience first (87% ‘very important’ and ‘important’), management experience second (80% ‘very important’ and ‘important’), board committee experience third (73% ‘very important’ and ‘important’), chairperson experience fourth (65% ‘very important’ and ‘important), and industry experience fifth (59% ‘very important’ and ‘important’).

Reflecting on the results, three points are striking. First, contrary to the results for the top two variables (board experience and management experience), the results showed that the majority of organisations in the sample refrained from insisting on previous experience as a chairperson (16% ‘very important’). It is said that overemphasis of previous experience would further reduce the limited number of available candidates (Brown, 2007, p. 304).^{Footnote 9} Second, the strong preference for management experience for the chair function is also evident in Fortune 50 organisations where 90% of persons elected were previously CEO, CFO, or COO (Heidrick & Struggles, 2021, p. 3). However, the importance of prior management leadership experience for the chair position is likely to diminish over time as boards seek to diversify themselves. Third, consistent with insights reported by Frey’s (2019, p. 91) Swiss board study, participants disagree about the importance of industry expertise. Some consider it indispensable (14% ‘very important’), others ‘of little importance’ (6%) or ‘of average importance’ (35%).^{Footnote 10}

7.4.2 Inferential Statistics

Now that the descriptive findings for RQ1’s chairperson competences have been presented, it is time to check for statistical variation of the responses in a non-parametric analysis according to Kruskal-Wallis (KW) and Dunn-Bonferroni (DB). For market capitalisation, shareholder structure, and board experience, an overview of the statistical results in Table 7.3 shows that significant response discrepancies primarily occur for shareholder structure, but occasionally also for market capitalisation and board experience.

Reminder: group definitions for the Dunn-Bonferroni (DB) multiple comparison post hoc test
*Market capitalisation*	*Shareholder structure*	*Board experience*
Gr1: < CHF 2,001m	Gr1 : Individual(s)/family	Gr1: < 10 years
Gr2 : CHF 2,001 – 10,000m	Gr2 : Inst. Investor/activist	Gr2: 10 – 20 years
Gr3: > CHF 10,000m	Gr3 : Government/(pension) fund	Gr3: > 20 years

Table 7.3 Survey – RQ1 Non-parametric Results

Full size table

Market Capitalisation

For market capitalisation, statistically significant discrepancies occur for (17) visionary thinking and (21) strategic thinking. First, since a visionary personality can have a positive effect on the governance of the often cumbersome large(r) organisations (Den Hartog, 2004, p. 175), visionary thinking is less pronounced in small-caps than mid-caps (H = 5.800, group 1 vs. 2, p < 0.05, r = 0.31). While there are significant differences at the 5% level, it is important to remember that they occur at a moderate correlation level (r = 0.31) and therefore need to be put into context when interpreting them. Second, mid-caps emphasise the importance of strategic thinking skills more than large-caps (H = 7.414, group 2 vs. 3, p < 0.05, r = 0.41). Of all participants from mid-caps, only one chose ‘important’, all others 'very important'. Despite the 5% significance level, however, it should be mentioned here that large-caps also give high priority to strategic capabilities (median = 5.00).

Shareholder Structure

For shareholder structure, there are eight significant group comparison results for (1) personal competences, (3) leadership competences, (12) networking, (22) audit and finance, (23) ESG, (25) legal, (28) chair experience, and (32) industry experience. To start with, the analysis of the ranking of the five competence dimensions shows, on the one hand, that organisations with institutional ownership place a higher value on personal competences than organisations with government ownership (H = 5.555, group 2 vs. 3, p < 0.10, r = 0.37). On the other hand, family-owned organisations prioritise leadership competences significantly more when comparing the results with those of institutional investor/activist ownership (H = 6.228, group 1 vs. 2, p < 0.05, r = 0.30).^{Footnote 11} In more detail, the sample shows that family- and state-owned organisations clearly favour leadership skills over all other dimensions (median 5.00). This result is somewhat surprising, however, as it was expected that organisations owned by individuals or the government would rank personal competences higher due to family/political lobbying relationships than organisations owned by institutional block holders (Kim & Cannella, 2008, p. 288).

Second, even more significant among the three group variables is networking (H = 12.566, group 1 vs. 3, p < 0.01, r = 0.48; H = 12.566, group 2 vs. 3, p < 0.10, r = 0.37). The statistical differences for family- vs. state-owned (1% significance level) and institutional investor- vs. state-owned organisations (10% significance level) highlight the need for the chairperson to interact with key stakeholders. Thereby, the individual(s)/family dimension seems to play the least role (median = 3.00). One can infer that the 1% significance level for the group comparison between family- and state-owned organisations has to do with the fact that a large number of family-owned organisations have shareholder representatives on the board (31 out of 39). It can thus be assumed that representation on the board reduces the need for extensive ownership interactions for the participants in the underlying sample.

Third, there are discrepancies in technical competences between audit and finance (H = 9.600, group 1 vs. 3, p < 0.05, r = 0.37; H = 9.600, group 2 vs. 3, p < 0.05 and r = 0.45), ESG (5.203, group 1 vs. 3, p < 0.10, r = 0.31), and legal (H = 9.749, group 1 vs. 3, p < 0.01, r = 0.41; H = 9.749, group 1 vs. 2, p < 0.05, r = 0.39). The results for the three technical components are driven by high conformity of state-owned organisations, while organisations owned by individuals/families and institutional investors/activists have lower relevance. As highlighted in the descriptive statistics of the introductory results (Section 7.3.1), this could be due to the fact that boards from organisations owned by individuals/families and institutional investors/activists are more diversified and that functional competences play a stronger role with regard to ordinary board members.

For the last block, which deals with candidates’ track record, significant differences exist between family- and state-owned organisations in terms of chair experience (H = 8.408, group 1 vs. 3, p < 0.05, r = 0.39) and industry experience (H = 8.012, group 1 vs. 3, p < 0.05, r = 0.38). In this context, experience is understood as a tool for coping with complexity (Helfat & Martin, 2015, p. 1286). To counteract organisational complexity, as one of many reasons, state-owned organisations therefore value practical experience for the chair position even more than technical competences (Aberg & Shen, 2020, p. 177; Owen & Kirchmaier, 2008, p. 205).

Board Experience

Regarding board experience, different statistical results are examined for (7) curiosity, (15) stewardship, and (31) management experience. First, survey participants with 10 – 20 years of experience are more likely to favour the curiosity components than participants with less than 10 years of experience (H = 5.527, group 1 vs. 2, p < 0.10, r = 0.28). Although at the 10% significance level, the results underline that it is more important for board members with intermediate experience to vote for a chairperson who keeps pace with economic and technological developments while fostering a collective learning and development process in the board (Morais & Kakabadse, 2013, pp. 82–83). As expected and illustrated by the weak Pearson coefficient (r = 0.28), curiosity is also not unimportant for less experienced participants, but is less emphasised in the underlying data structure.

Second, people with 10 – 20 years of experience consider the stewardship and facilitation competence to be crucial (H = 10.289, group 1 vs. 2, p < 0.01, r = 0.39). By contrast, people with less than 10 years of board experience see less benefit in the steering and goal-orientation skills of the chair. Given the results, it appears that board members with less than 10 years of experience recognise the joint and several liability (in German: Solidarhaftung) of the board of directors more strongly with respect to Article 716a revCO.

Third, another statistical difference relates to management experience (H = 5.839, group 2 vs. 3, p < 0.10, r = 0.35). In the group of individuals with 10 – 20 years and the group with more than 20 years of experience, the participants with medium experience most strongly prefer people at the top who have already held one or more senior management positions. Considering that the manager is a vital organisational resource and that it has been common in the past to move from CEO to chairperson (Buchholtz et al., 2003, p. 506; Spencer Stuart, 2014, p. 5), it was to be anticipated that the more experienced participants would emphasise management experience the most at the top (median 4.00). However, this phenomenon could not be confirmed in the underlying sample.

7.5 RQ2 Results

In examining RQ2 (input-process-output), survey participants were initially asked to what extent the key contingencies influence the search process. Also, they were asked to what extent they are willing to involve the key stakeholders in the succession of the chairperson. Again, in accordance with the methodological procedure described (Section 7.1.4), the results are first presented using descriptive statistics and then using inferential statistics.

7.5.1 Descriptive Statistics

To start with, Figure 7.22 looks at the aggregated responses of all contingency dimensions (business, environmental, governance, political). In line with societal discussions (e.g. Hill, 2021), the strongest effects come from governance- (65% ‘very important’ and ‘important’) and business-related factors (63% ‘very important’ and ‘important’). Somewhat less important or positive are environmental (33% ‘very important’ and ‘important’) and political contingencies (40% ‘very important’ and ‘important’). When looking at all four dimensions, it is remarkable that all ‘very low/negative’ and ‘low/negative’ combined are below 30%. This can be interpreted in such a way that there is a great opportunity to influence the process structure.^{Footnote 12}

In terms of business contingencies (Figure 7.23), which now specify the individual variables in each dimension, board structure and strategy appeared to be the two most important influencing factors. Overall, 81% of the participants voted ‘high’ and ‘very high’ for board structure and 74% for strategy. Life cycle ranked third with slightly more ‘average’ (31%) and fewer ‘very high’ (11%) responses. In contrast to the qualitative statements in the interviews (Section 6.5), respondents seem to attach less importance to the life cycle thematic (11% ‘very high’ and 49% ‘high’).

When applying a contingency perspective to the life cycle, as proposed by Gabrielsson and Huse (2004, p. 20), it is worth noting that the results should be interpreted with caution, as the advanced maturity stage of the underlying organisations influences the outcome (the more mature, the lower the importance of the life cycle). The high maturity of the chairpersons and head of nomination committees in the survey sample (Section 7.2.1) is the reason why the life cycle is probably given too little importance. Last but not least, the almost even distribution from ‘very low’ (23%) to ‘very high’ (19%) shows that the participants have divergent views on the influence of the owner in chair succession.^{Footnote 13}

Among environmental contingencies (Figure 7.24), industry (26% ‘very high’ and 48% ‘high’) and market dynamics (6% ‘very high’ and 40% ‘high’) have the strongest effect on chairperson succession. The influence of hard and soft legal regulation is slightly above average and third strongest (median 3.00). Approximately 36% of all participants see a ‘very low’ to ‘low’ influence on process planning. Interestingly, cross-checking regulation by sector distribution shows that regulatory influence primarily affects organisations in the financial, healthcare, and real estate sectors. Although anticipated to have a strong impact on the succession of the chair due to increasing activism in recent years (Section 4.3.3.2), the proxy advisor is the environmental dimension with the lowest impact with 20% ‘very low’ and 36% ‘low’ responses.

For governance contingencies (Figure 7.25), the items confirm the expected high scores. In particular, the three established factors of independence (51% ‘very high’), confidentiality (43% ‘very high’), and overboarding (35% ‘very high’) have an effect on the selection process of the board chair. With lower scores than the three most important governance variables, the influence of diversity (11% ‘very high’ and 39% ‘high’) and Swissness (5% ‘very high’ and 30% ‘high’) is moderate compared to expectations.

For political contingencies (Figure 7.26), as political games tend to be disadvantageous (Walther, Calabrò, & Morner, 2017, p. 353), it was assumed in the theory (Section 4.1.2) and the interviews (Section 6.2) that the nomination process follows objective rather than subjective criteria. It was therefore predicted that the political dimensions would be rather ‘(very) negative’ than ‘(very) positive’. However, the overall positive results of the underlying sample speak a different language. While network effects have the advantage of making contacts and gaining prior knowledge about the candidate (4% 'very positive’ and 44% ‘positive’), there is also a slightly above-average positive attitude towards board chair influence (9% 'very positive’ and 38% ‘positive’). In contrast, power, with a median of 3.00 and a mode of 2, appears slightly below ‘average’ (34%) and more in line with the literature review and qualitative insights (9% ‘very negative’ and 35% ‘negative’).

After dealing with the key contingencies, the descriptive results of the key stakeholders with respect to the succession process of the chairperson are analysed. Beginning with board involvement (Figure 7.27), survey participants overall agree that full board participation in succession is important. Of the four questions asked, the one question whether the board members must express an opinion on the candidate profile was most clearly answered with a total of 65% ‘strongly agree’, 24% ‘agree’, 5% ‘neutral’, and only 5% ‘disagree’. Less important and slightly controversial among survey participants are participation in board interviews (18% ‘strongly disagree’ and ‘disagree’) and approval of candidacy for the entire board (16% ‘strongly disagree’ and ‘disagree’). To further explore the question of whether the current or the departing chairperson should be involved, participants were specifically asked about their opinion on the active engagement of the chair and his or her involvement in the search for a successor. Following Owen and Kirchmaier (2008, p. 203), who see in a strong involvement a knowledge and advisory advantage for the nomination committee, the findings indicate that the majority of participants were also in favour of a strong involvement of the board chair (68% ‘strongly agree’ and ‘agree’).

For company secretary involvement (Figure 7.28), the overall results are critical. Participants see the company secretary’s role primarily as a process facilitator (31% ‘strongly agree’ and 18% ‘agree’). However, they see no influence on the distribution of information (76% ‘strongly disagree’ and ‘disagree’) or the need to conduct interviews (87% ‘strongly disagree’ and ‘disagree’). As there are assumingly differences in the role of the company secretary depending on the organisations one is talking to, some of the respondents give less importance to the supporting function (16% ‘strongly disagree’ and 20% ‘disagree’).

For CEO involvement (Figure 7.29), participation was rated positively overall. Analogous to Olson and Adam (2004, p. 449), the participants recognised the expression of opinion on the candidate’s profile, the alignment with strategic entrepreneurial philosophy, and the interview participation as insightful. However, as the values in Figure 7.29 show, interview participation (18% ‘strongly disagree’) and expressing an opinion on the profile of the chair profile (5% ‘strongly disagree’) are not considered uncritical. Following the results and theories (Sections 1.4.1 and 4.3.2.3), it can be concluded that in the Swiss governance environment, the CEO tends to have little influence on the choice of the chair.

With respect to board advisory involvement (Figure 7.30), the answers are distributed on average around the median value of the 5-Likert scale (3.00 = ‘neutral’). Following the median, the participants prioritise the gatekeeper function (49% ‘strongly agree’ and ‘agree’), the objectivity provided (37% ‘strongly agree’ and ‘agree’), and the need for process support (26% ‘strongly agree’ and ‘agree’) in the ranking order presented. As can be seen in the overall distribution in Figure 7.30, there are participants who are very positive about the support of a board advisor in chair succession, while others are rather sceptical. The tension described in the theory in Section 4.3.2.1 between the expansion of the candidate pool (positive) and the different strategically limited concept of fit approaches with which various board advisors work (negative) are each a reason for the partly positive, partly negative attitude (Doldor et al., 2012, p. 29).

For shareholder involvement (Figure 7.31), the results indicated that shareholders demand rigorous succession planning (10% ‘strongly agree’ and 40% ‘agree’) and an understanding of its process (4% ‘strongly agree’ and 31% ‘agree’). According to the results, organisations should therefore be concerned about providing appropriate information to meet shareholders’ expectations regarding the chairperson succession process. However, with 39% ‘disagree’ and 31% indicating ‘neutral’, participants see neither a strong positive nor a strong negative desire among shareholders to express expectations and opinions. Given the outcome, it seems that shareholders trust the nomination committee to conduct an objective and competent succession process.

7.5.2 Inferential Statistics

Section 7.5.2 focuses on the non-parametric statistics for RQ2 and highlights the results of the Kruskal-Wallis (KW) and Dunn-Bonferroni (DB) tests (Table 7.4). By outlining the inferential statistics, yet without intending to neglect the statistics for market capitalisation and board experience, particularly the strong response variations in terms of shareholder structure are highlighted. In accordance with Banerjee et al. (2020), the large discrepancy with respect to shareholder structure illustrates that the succession processes for board chairs in organisations that are owned by individuals/families or institutional investors/activists or governments/(pension) funds have unique dynamics and consequently differ.

Reminder: group definitions for the Dunn-Bonferroni (DB) multiple comparison post hoc test
*Market capitalisation*	*Shareholder structure*	*Board experience*
Gr1: < CHF 2,001m	Gr1 : Individual(s)/family	Gr1: < 10 years
Gr2 : CHF 2,001 – 10,000m	Gr2 : Inst. Investor/activist	Gr2: 10 – 20 years
Gr3: > CHF 10,000m	Gr3 : Government/(pension) fund	Gr3: > 20 years

Table 7.4 Survey – RQ2 Non-parametric Results

Full size table

Market Capitalisation

For market capitalisation, the statistically significant results for small- vs. large-caps relate to (3) ownership and for small- vs. mid- caps to (5) industry, (11) independence, and (20) active chairperson contribution. When looking at small- vs. large-caps variations, the first finding is that small-caps rate the influence of ownership structure more highly (H = 7.085, group 1 vs. 3, p < 0.05, r = 0.35). This relates to the fact that organisations with smaller market capitalisation have more substantial block holdings and consequently pay more attention to ownership (Credit Suisse Research Institute, 2017, p. 6). Based on this specific insight, small-caps in particular should seek to engage in proactive chair succession dialogue with their owners.

For mid-sized compared to smaller organisations, responses are stronger for industry (H = 4.646, group 1 vs. 2, p < 0.10, r = 0.28) and independence (H = 8.598, group 1 vs. 2, p < 0.05, r = 0.37), but weaker for active chair contribution (H = 6.457, group 1 vs. 2, p < 0.05, r = 0.33). Since larger organisations tend to be more in the focus of stakeholders (Deegan et al., 2002, p. 321), the examination of the results suggests that large-caps are more likely to implement processes according to best practice or regulatory criteria.

Shareholder Structure

For shareholder structure, statistically relevant group response variances arise for (3) ownership, (7) proxy activism, (8) regulation, (9) confidentiality, (10) diversity, (14) chairperson influence, (15) network, (16) power, (18) board interviews, (22) company secretary influence, and (23) company secretary interviews. Looking at the discrepancies by group, most of them relate to individual/family vs. government/(pension) fund (group 1 vs. 3) and institutional investor/activist vs. government/(pension) fund (group 2 vs. 3). They are the following:

With respect to business contingencies, ownership-related influences for organisations with predominantly institutional anchor shareholders are less important (H = 7.333, group 1 vs. 2, p < 0.05, r = 0.31; H = 7.333, group 2 vs. 3, p < 0.10, r = 0.35). As organisations owned by institutional investors often have a more diversified shareholder structure, greater personal anonymity, and consequently less voting power, participants opined that block holders from organisations owned by individuals/families and governments have more influence on the succession process (Molitor, 2010, p. 105).

In terms of environmental contingencies, due to the logical profit argumentations, the influence of proxy activism occurs more in organisations owned by institutional block holders but less in family-owned organisations (H = 8.309, group 1 vs. 2, p < 0.05, r = 0.32), with the effect of the Pearson correlation coefficient still moderate. Following the public-social legitimacy of stakeholder theory, compliance with regulations for chair succession as second environmental dimension is more crucial for government-owned in contrast to organisations owned by individuals/families and institutional investors/activists (H = 9.146, group 1 vs. 3, p < 0.01, r = 0.41; H = 9.146, group 2 vs. 3, p < 0.10, r = 0.34).

For governance-associated contingencies, statistically high discrepancies can be identified for confidentiality (H = 8.696, group 2 vs. 3, p < 0.05, r = 0.46) and diversity (H = 10.371, group 1 vs. 3, p < 0.01, r = 0.42; H = 10.371, group 2 vs. 3, p < 0.05, r = 0.42). Although legally the same for all organisations, state-led organisations show greater awareness for best practice activities than organisations owned by families and institutional investors (indicated by the Pearson coefficient). Thus, it can be concluded that state-owned organisations particularly adhere to (non-)mandatory governance regulations and go beyond mere minimum standards.

With regard to political contingencies, there are mainly significant group differences between institutional investor vs. state ownership. In contrast to the representatives of organisations owned by institutional investors/activists, the representatives of state-owned organisations acknowledge a stronger positive influence of chairperson influence (H = 5.368, group 2 vs. 3, p < 0.10, r = 0.36), networking effects (H = 12.619, group 2 vs. 3, p < 0.01, r = 0.54), and power relations (H = 6.344, group 2 vs. 3, p < 0.05, r = 0.39). Likely due to the multi-level context of social, political, and economic factors, as highlighted by Kozlowski and Ilgen (2006, p. 78), network effects seem to be a decisive factor in attracting candidates for state-owned organisations. This is also underlined by the discrepancy to the results for family-owned organisations (H = 12.619, group 1 vs. 3, p < 0.05, r = 0.40).

With respect to stakeholders, there is a statistical group difference between organisations with institutional investor- vs. state block holdings in terms of participation in board interviews (H = 6.986, group 2 vs. 3, p < 0.05, r = 0.38). Due to the tendency towards board interviews, organisations owned by institutional investors pay more attention to strategic board consensus (Mathieu et al., 2008, p. 430). On the other hand, government-owned organisations agree more with family-owned organisations on the influence of the company secretary (H = 6.013, group 1 vs. 3, p < 0.10, r = 0.32) and company secretary interview participation (H = 5.066, group 1 vs. 3, p < 0.10, r = 0.29). By involving the company secretary more, state-owned organisations assumingly use legal adequacy as a countermeasure to the rather strong influence of politics.

Board Experience

For board experience, the seven statistical results refer to (4) strategy, (10) diversity, (11) independence, (16) power, (21) company secretary support, (22) company secretary influence, and (32) shareholder expression. Despite a generally high rate of agreement for the underlying moderators of the succession process, participants with 10 – 20 years of board experience overall express certain contingencies stronger than those with over 20 years. Among them are strategy (H = 6.242, group 2 vs. 3, p < 0.05, r = 0.36), diversity (H = 11.248, group 2 vs. 3, p < 0.01, r = 0.47), independence (H = 7.452, group 2 vs. 3, p < 0.05, r = 0.36), and company secretary support (H = 5.079, group 2 vs. 3, p < 0.10, r = 0.33). Most of them relate to best practice attributes. The only exception was shareholder expression (H = 5.008, group 2 vs. 3, p < 0.10, r = 0.33) which was prioritised more by experienced participants.

Other statistical differences related to diversity (H = 11.248, group 1 vs. 3, p < 0.05, r = 0.39), power (H = 6.256, group 1 vs. 2, p < 0.05, r = 0.34), and company secretary influence (H = 8.313, group 1 vs. 2, p < 0.10, r = 0.28; H = 8.313, group 1 vs. 3, p < 0.05, r = 0.36). In contrast to members with low board experience (less than 10 years), the latter two were less emphasised by members with medium (10 – 20 years) and senior experience (more than 20 years). Especially the stronger integration of the company secretary is a possible consequence of the stricter regulation and legitimacy concerns in recent years, which is why it was emphasised by the younger generations (Luoma & Goodstein, 1999, p. 555).

7.6 RQ3 Results

To address RQ3 (input-process-output), survey participants were asked to state their reasons for (dis)agreement with the principles of voluntary disclosure in chair succession. Following the theoretical concepts (Section 4.4), questions were asked about the dissemination principles (why), scope and content (what), channel (where and by which means), and time/timing (when). In line with the previous systematics, the results were provided using descriptive and inferential statistical methods.

7.6.1 Descriptive Statistics

Based on the economic governance motives (Figure 7.32), respondents recognised the value that transparency has for decision-making. The 15% ‘strongly agree’ and 49% ‘agree’ indicate that organisations are aware that an adequate level of information is important for the nomination of the chairperson. In this sense, the organisations participating in the sample indirectly agree that it is also their point of view to provide an adequate level of information to shareholders (ICGN, 2018, p. 3). Given the organisation’s sense of accountability, survey respondents are confident that the organisations are achieving the disclosure targets they have set themselves (78% ‘strongly agree’ and ‘agree’). As a consequence, 60% of respondents (‘strongly agree’ and ‘agree’) see no further need to drastically expand the information level for chairperson succession. The few participants who disagree (15% ‘strongly disagree’ and ‘disagree’) emphasise the result.

Understanding in more detail how organisations deal with social governance motives (Figure 7.33), the partaking SPI organisations confirmed that they are eager and confident to meet shareholder and stakeholder expectations (10% ‘strongly agree’ and 51%’ ‘agree’). However, they are aware that uncertainty exists about the consequences the sensitive nature of disclosure may have. In line with the cautious attitude (Section 4.4.1.1), the majority of participants are hesitant about disclosing succession planning (14% ‘strongly agree’ and 43% ‘agree’). In addition, as 43% of the respondents with ‘disagree’ clarify, organisations rarely compare their concepts for succession reporting with those of their industry competitors. Rather, it seems that organisations set their strategy for voluntary disclosure after having fulfilled the mandatory obligations. The concept of mimetic isomorphism (Section 4.4.1.1) seems to have minimal effect in the underlying sample.

In terms of scope and content (Figure 7.34 and Figure 7.35), the participants continue to follow a rather passive tenor. Exceptions apply to person-related variables (‘strongly agree’ and ‘agree’): experience and knowledge (80%), role and time commitment (56%), reason/rationale for nomination (49%), and candidate profile (35%). As background information on the nominated personalities is somewhat standardised in voluntary disclosure, however, it was expected that the four variables would reach the highest agreement.

By contrast, participants are far more critical of the disclosure of information relating to the structure of the process (total for ‘strongly agree’ and ‘agree’): process owners (30%), process steps (23%), nomination committee succession and strategic planning (19%), assessment centre performance (16%), and board advisory involvement (12%). As the overall disagreement rates of 69% and 64% show (‘strongly disagree’ and ‘disagree’), the information on the board advisor and the assessment centre were particularly controversial.

Figure 7.36 addresses questions on channel and time/timing of voluntary disclosure. In terms of channel (left side of the graph), respondents favour the press release (48%) over the AGM (19%) or the annual report (15%). Below the 10% threshold are roadshows (9%), shareholder letters (6%), and other (4%). Investor days and social media do not seem to be a choice at all, as not one single participant selected them (although also offered as response option). In principle, the survey results confirm that traditional channels are favoured. Yet differences to theory exist (Section 4.4.1.3), with the annual report having the highest priority and roadshow and letter to shareholders the lowest.

As far as time/timing is concerned (right side of the graph), participants selected ad hoc publication (48%) as the most appropriate point in time as a consequence of the press release being the favoured channel. Although in principle, voluntary disclosure is addressed (beyond the mandatory scheme), the priority for ad hoc publication is strongly related to the existing stock exchange rules that require immediate disclosure shortly after the official vote on board nomination (Section 2.2). The temporal congruence between mandatory and voluntary disclosure can be justified by the fact that organisations try to publish the official nomination and background information on the election simultaneously. Beyond the ad hoc announcement, other participants specify the required period as 3 months (28%), 6 months (21%), and 12 months (4%) before the respective AGM. However, participants agreed that it should not be earlier than 12 months (although also offered as response option).

7.6.2 Inferential Statistics

Again, the inferential statistics of RQ3 are provided by Kruskal-Wallis (KW) and Dunn-Bonferroni (DB) tests. Checking the sample variations related to market capitalisation, shareholder structure, and board experience, as highlighted in Table 7.5, reveals that there is only a small number of statistically significant discrepancies between the groups’ answers. Basically, the low discrepancy confirms the rather hesitant and/or barely noticeable openness towards substantial disclosure with regard to board chair nomination (Parum, 2005, p. 704).

Reminder: group definitions for the Dunn-Bonferroni (DB) multiple comparison post hoc test
*Market capitalisation*	*Shareholder structure*	*Board experience*
Gr1: < CHF 2,001m	Gr1 : Individual(s)/family	Gr1: < 10 years
Gr2 : CHF 2,001 – 10,000m	Gr2 : Inst. Investor/activist	Gr2: 10 – 20 years
Gr3: > CHF 10,000m	Gr3 : Government/(pension) fund	Gr3: > 20 years

Table 7.5 Survey – RQ3 Non-parametric Results

Full size table

Market Capitalisation

For market capitalisation, the only statistically significant variation concerns (14) process steps. This shows that small-caps, in contrast to mid-caps, are less willing to report process steps publicly (H = 6.314, group 1 vs. 2, p < 0.05, r = 0.32). The fact that fewer small-caps have a formal nomination committee does not mean that there is no adequate process. However, the results emphasise that in organisations with a market capitalisation of less than CHF 2bn, the process is more likely to take place as a collaborative approach by the board (Walther, Calabrò, & Morner, 2017, p. 2201).

Shareholder Structure

For shareholder structure, significant discrepancies appear for (3) need for reporting and (9) overboarding. First, participants in government-owned organisations oppose expanding the level of information on succession more than family-owned organisations (H = 6.231, group 1 vs. 3, p < 0.05, r = 0.33). Board representatives from state-owned organisations are already strongly involved in social and political exchanges. It looks as if this limits the support for more extensive succession disclosure. Second, overboarding in voluntary disclosure seems to be rather a concern for institutional investor/activist- than for state-owned organisations (H = 5.640, group 2 vs. 3, p < 0.10, r = 0.34). There is a logical explanation for this. Although statistically at the 10% level, organisations owned by institutional block holders are more focused on time resources and independence than organisations owned by the government. As lack of time can be a decisive factor (Harris & Shimizu, 2004, p. 777), representatives of institutional block holders pay more attention to governance alignment.

Board Experience

With regard to board experience, a statistical discrepancy occurs for (3) the need for reporting. At the 10% level, members with less than 10 years of experience are more critical of disclosing information than members with 10 – 20 years of board experience (H = 5.716, group 1 vs. 2, p < 0.10, r = 0.27).^{Footnote 14} The increased attention from more 'inexperienced' board members is related to regulatory concerns that have emerged in the last few years (see also Section 7.5.2). In contrast to the two other groups (10 – 20 years and more than 20 years of experience), concerns about over-reporting are more pronounced among board members with less than 10 years of experience.

7.7 Limitations

Despite its great benefits, board survey data collection also faces limitations. Methodologically, as the online survey is self-administered, the researcher has no control over how the participants complete the survey and how they interpret the questions (Bernard, 2017, p. 258). Thus, the chairperson may have delegated the survey to the company secretary and/or the questions may not have been answered in a purposeful manner. Moreover, since chairperson succession is strongly oriented towards societal norms, there is a risk that participants may have avoided culturally unacceptable responses and thus circumvented conventional practices (Bernard, 2017, p. 260). As the survey also inquired about one’s own tasks (chairperson) and process responsibilities (head nomination committee), which requires a strong ability to self-reflect, there is a potential self-assessment bias in the results (Campbell & Lee, 1988, p. 308). In this respect, the latter two aspects suggest that, the results may contain best practice tendencies but distort practical realities. It is one reason, among others, why there is a gap between theory and practice (Section 8.1.1).

Moreover, conceptually, the results of the chair survey are subject to the specific underlying sample (Huse et al., 2011, p. 21). Therefore, it cannot be ruled out that the answers may diverge if other participants from the target group who did not participate were to present their opinions (e.g. influence of organisations, the majority of which do not have a nomination committee). This also concerns other types of stakeholders, for example, the inclusion of non-governmental organisations, media, and proxy advisors (beyond pure activism). Not to lose focus, however, the main attention was on the key stakeholders previously confirmed in theory and interviews. Likewise, there is a high temporal relevance. The survey was conducted in Q4 2021. If the survey results were to be repeated numerous times (multiple points in time) or postponed to a different point in time (single point in time), the response structure might rightly look different (Huse, 2009b, p. 380). Also, some organisations refused to participate as the election of the board chair was imminent and they did not want to make any prior statements. Apparently, despite personal assurances, confidentiality and anonymity were two other obstacles to a higher response rate (Bednar & Westphal, p. 38). From a corporate sociological point of view, chair succession appears to be a highly sensitive topic. Given the partial reluctance to engage in public dialogue, there may be other, as yet unknown, factors driving chairperson succession. However, since a thorough literature review and mixed-methods research were used, the probability that missing factors exist or have not been adequately considered is unlikely. To conclude, the descriptive statistics showed little response variance. With a few exceptions, survey participants often agreed on the relevant topics. However, since the inferential statistical analysis is based on the former, it consequently led to minor group-specific discrepancies.

Lastly, theoretically, the focus on the research questions indicates that there are also question-specific limitations: The non-homogeneous results of RQ1 show that chair competences are subject to unique dynamics. This is reflected in the only partially confirmed dyadic relationships and structural overlaps between the competence categories. According to Carter and Lorsch (2003, p. 115), one reason for this could be that competences are perceived differently depending on the functional positions that the persons occupy (e.g. chairperson vs. board member vs. senior management). Therefore, to counter functional despotism, the head nomination committee was involved. Furthermore, many competences seem to be important, as indicated by the strong emphasis on ‘strongly agree’ and ‘agree’. This could be related to the over-profiling effect (Section 8.1.1) or a methodological pitfall (e.g. ceiling effect).^{Footnote 15} The latter can be ruled out, however, as the dual approach with the ranking of categories and assessment of single variables attempted to take adequate measures to deal with such methodological effects.

For the process moderators of RQ2, governance plays a leading role and reflects its strong momentum in society/economy. However, despite widespread application, there is no uniform definition (Section 1.4.1). Participants may thus understand factors differently. To counter response bias, keywords for each variable were added in the survey (e.g. for independence: independence in thought and independence in appearance). In addition, participants in the succession process had a slightly positive relationship to political contingencies. While questions on business, governance, and environmental contingencies focused on their importance, questions on political contingencies inquired about positive/negative impacts. If participants completed the survey with too little attention, the results may have been misinterpreted due to the altered measurement scale, resulting in a potential over-preference (Collins, 2003, p. 230). To avoid response bias in the survey, however, the scale was explicitly repeated before each dimension.

Compared to the first two research questions, the results of RQ3 are less prominent overall (literature review, interviews, and survey). This is due to the intermediate research (academia) and the lack of organisational recognition (practice). Considering the input-process-output scheme (output building on input and process insights) and the sound survey methodology (10 – 15 minutes response time), the focus was placed on RQ1 and RQ2 questions. Otherwise, artificially expanding the survey with RQ3 questions would have increased the risk of a lower response rate (Bell et al., 2019, p. 234).

7.8 Review

The results of the quantitative survey revealed that chairperson succession planning is a multidimensional topic. Along the defined input-process-output approach, descriptive and inferential statistics provided insights on how SPI board chairs and SPI nomination committee heads identify competences, process moderators, and voluntary disclosure. In this context, in-depth descriptions not only linked the results to the theoretical background and exploratory findings, but also specifically addressed the defined research questions under study. By addressing RQ1, RQ2, and RQ3, it was possible to show which dynamics are at play and which organisation representatives, according to the three group sub-samples market capitalisation, shareholder structure, and board experience, perceive certain aspects differently. The following overview summarises the underlying findings:

The introductory results for board chair succession highlighted that this is a longer-term issue, that candidate profiling is crucial, and that overall planning is more profound than for ordinary board member succession. Despite the high complexity and multidimensionality, only 13% of sample participants felt dissatisfied with the quality of the last succession process.
For competences (RQ1), participants ranked leadership competences above personal, business, social, and technical competences. When the opinions on the individual competence items are aggregated, hower, personal competences are ranked higher than business, leadership, social, and technical competences. It can thus be concluded that boardroom behaviour (personal, leadership, business competences) is of greater importance than network (social competences) or function-specific skills (technical competences), albeit with slight statistical differences between mid-caps and large-caps (personal competences) as well as small-caps and mid-caps (leadership competences).
Among process moderators (RQ2), governance contingencies have the greatest influence, followed by business, environmental, and political contingencies. However, statistically significant differences emerged in the responses of small-caps vs. large-caps (independence), institutional investor- vs. state-owned organisations (confidentiality), and members with less than 10 vs. 10 – 20 years of board experience (diversity, independence). When stakeholders are involved, process owners prioritise dialogue with board members and the CEO, with less preference for the company secretary, shareholders, and the board advisor.
With regard to voluntary disclosure (RQ3), the sample participants saw the advantage of an adequate disclosure mechanism, but were rather hesitant as they wanted to minimise risks and not restrict the freedom to act. More precisely, analogous to the qualitative findings (Section 6.6), survey respondents also saw clear limits to what organisations are able and willing to provide. Hesitation was particularly evident in the low approval rate for new, non-standardised disclosure proposals (e.g. board advisory and assessment centres), with no discrepancies between the group variables under study.

Notes

1.
Walther and Morner (2014, p. 142) have already successfully used the tactic of employing a gatekeeper for chairs and vice-chairs in their study.
2.
Montana Aerospace and Sensirion have two persons simultaneously serving as chair. Also, according to the annual reports, seven organisations have entrusted other committees with nomination duties. Their committee heads were also included in the sample population.
3.
For Simms et al. (2019, p. 7), it is often a choice between simplicity and measurement (im)precision by the researcher.
4.
For a visual overview on non-parametric/parametric testing see, Fitzgerald et al. (2001, p. 289) and Corder and Foreman (2014, p. 4).
5.
This thesis refers to family-owned (individual(s)/family), institutional investor-owned (institutional investor/activist), and government/state-owned (government/(pension) fund).
6.
It is possible that the Kruskal-Wallis H test is significant and the Dunn-Bonferroni post hoc test is insignificant. Among other reasons, this may be due to the adjusted alpha risk level (type I error).
7.
Further responses came from ‘chairperson & chair NC’ holding both positions (4 vs. 4; 100%) and ‘others’ (1 vs. 4; 20%).
8.
In contrast to the survey, the ranking was reversed/adjusted to the 5-Likert scale level with highest (5) and lowest priority (1). Initially, participants were asked to rank from 1 (highest priority) to 5 (lowest priority).
9.
According to Spencer Stuart (2020a, p. 5), for example, half of all FTSE100 chairpersons appointed between 2016 and 2019 were persons who had never chaired the board of a listed organisation before.
10.
According to Fitzsimmons and Callan (2016), specific industry and/or prior management experience can be “valued by firms only where they are accrued in specific industry or organizational contexts” (p. 778).
11.
This finding also applies to state-owned organisations, but the result was not statistically significant.
12.
This is confirmed by Barth (2013, p. 40), who speaks of comprehensive process characteristics in this context (in German: übergreifende Prozessmerkmale).
13.
The discrepancy is most likely related to the structure of block holdings (individual/family vs. institutional investor/activist vs. state/(pension) fund) affecting the sample (Chapter 7.2.2).
14.
In principle, for (3) the need for reporting, the values between group 1 (individual/family) and group 3 (state/(pension) fund) differ more. Since the results of the comparison between group 1 (individual/family) and group 3 (state/(pension) fund) are adjusted for type 1 error by the Dunn-Bonferroni correction (p = 0.068, adjusted p = 0.204), only the comparison between group 1 (individual/family) and group 2 (institutional investor/activist) is shown (p = 0.033, adjusted p = 0.098).
15.
See Simkovic and Träuble (2019) for more specific information on the ceiling effect.

Author information

Authors and Affiliations

St. Gallen, Switzerland
Cornel Germann

Authors

Cornel Germann
View author publications
You can also search for this author in PubMed Google Scholar

Corresponding author

Correspondence to Cornel Germann .

7.1 Electronic supplementary material

Below is the link to the electronic supplementary material.

Supplementary file1 (DOCX 488 kb)

Rights and permissions

Open Access This chapter is licensed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

The images or other third party material in this chapter are included in the chapter's Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the chapter's Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.

Reprints and permissions

Copyright information

About this chapter

Cite this chapter

Germann, C. (2023). Quantitative Paradigm. In: Chairperson Succession. Springer Gabler, Wiesbaden. https://doi.org/10.1007/978-3-658-40817-6_7

Download citation

DOI: https://doi.org/10.1007/978-3-658-40817-6_7
Published: 19 April 2023
Publisher Name: Springer Gabler, Wiesbaden
Print ISBN: 978-3-658-40816-9
Online ISBN: 978-3-658-40817-6
eBook Packages: Social Science and Law (German Language)

Publish with us

Policies and ethics

Quantitative Paradigm

Abstract

7.1 Research Instrument

7.1.1 Web-Based Survey

7.1.2 Quantitative Application

7.1.3 Quantitative Reliability and Validity

7.1.4 Methodological Analysis

7.2 Description of the Survey Sample

7.2.1 Person-related Demographics

7.2.2 Organisation-related Demographics

7.3 Introductory Results

7.3.1 Descriptive Statistics

7.3.2 Inferential Statistics

7.4 RQ1 Results

7.4.1 Descriptive Statistics

7.4.2 Inferential Statistics

7.5 RQ2 Results

7.5.1 Descriptive Statistics

7.5.2 Inferential Statistics

7.6 RQ3 Results

7.6.1 Descriptive Statistics

7.6.2 Inferential Statistics

7.7 Limitations

7.8 Review

Notes

Author information

Authors and Affiliations

Corresponding author

7.1 Electronic supplementary material

Supplementary file1 (DOCX 488 kb)

Rights and permissions

Copyright information

About this chapter

Cite this chapter

Download citation

Share this chapter

Publish with us

Search

Navigation