Management turnover expectations: a variable to explain company readiness to engage in continuous management training

Abstract

What makes companies give their support to Continuous Management Training (CMT) and why, on occasion, do they block its progress? If a company thinks its managers might leave in the near future, would there be reasons for it to provide continuous training of these professionals? Do companies interpret all likely resignations in the same way? We attempt to provide an answer to these questions, by analysing the behaviour of companies' vis-à-vis continuous management training in terms of their expectations of the appropriability of the benefits that stem from such training. Appropriability is in turn assessed with regard to firms' expectations that their managers will leave their organisation without it having had the chance to recover the outlay invested in training. The analysis of business behaviour also demands an understanding of how these turnover expectations are produced and, more specifically, of the variables on which turnover expectations depend or might depend. The empirical study we carried out with over 300 Spanish companies revealed that company projections for voluntary turnover of management employees are positively connected with earlier experiences of turnover, with markets prone to change, and with business risk situations, and negatively associated with developed social management networks, satisfied managers and their degree of specificity in relation to the company. Nevertheless, no significant relationship was observed between expected turnover and the intensity of continuous management training. In general, companies make their management training investment decisions independently of the overall turnover expectations they might have. Different causes might exist for this absence of a relationship, centred in particular on the need for management training and an associated sense of urgency, which takes precedence over the risk of not recovering the investment when the business decision is taken, and on the nature of the management turnover perceived by the company, which is more complex than a strictly economic one.

Keywords:

• appropriability
• Continuous Management Training
• internal labour markets
• managers
• moral hazard
• readiness
• turnover

Notes

 1. According to European Commission data (2004), in 2001 only 1.6% of Spanish managers had taken part in continuous training schemes during the month prior to the reference survey, a percentage considerably under the EU-15 average (10.6%) and the percentages for Spanish professionals and technical engineers (9.4% and 6.2%, respectively).

 2. Continuous Management Training as a means for creating and maintaining competitive advantage is an activity of unquestionable relevance in today's companies. Managers themselves constitute a group whose special relevance for business competitiveness has been pointed out by many authors (Penrose 1959; Andrews 1987; Child 1972; Katz 1974; Anderson and Paine 1975; Hambrick and Mason 1984; Castanias and Helfat 1991, 2001; Lado and Wilson 1994; Pickett 1998). The particular importance of managers implies that they should be given special academic attention and that organisations must make a major effort so that they have the qualified, involved management team they need to meet their purposes at all times. This means undertaking management administration policies to guarantee the recruitment, retention and, of course, continuous development of managers, through proper management of their ongoing training.

 3. This study forms part of a more extensive University-Business research project financed by the Basque Government concerning company behaviour vis-à-vis continuous management training, in which different variables that may condition this behaviour were investigated and modelled (Araujo, Barrutia, Hoyos, Landeta and Ibañez 2006).

 4. Some data that make it possible to back up such an extrapolation are: i.- In terms of real GDP, the Basque Country grew in 2005 and 2006 at rates of 3.9% and 4.2% respectively, against 1.6% or 2.9% in the entirety of the EU-25 or 3.5% in the USA; ii.- In the Basque Country, the percentage of expenditure on R+D in relation to GDP reached 1.43% in 2005, slightly below the 1.86% that constituted the average in the EU-25; iii.- The Synthetic European Innovation Indicator (SII), an index that measures the degree to which the different economies have the capacity to develop innovation activities, rose in the Basque Country in 2006 to 0.35 (over a maximum of 1) against the 0.45 obtained by the EU-25 for the same year; iv.- The exporting tendency of the Basque Country (Exports/ GDP) rose in 2005 to 27.32% against 26.10%, 24.83% and 24.10%, in the cases of France, Italy and the United Kingdom respectively (Eustat 2006; Eurostat 2006)

 5. Our analysis did not take into consideration aspects directly connected with the personal characteristics of managers, whose influence in the likelihood of turnover has been dealt with by other studies (De Luis et al. 2002), since we intended to concentrate our research on internal and external aspects of the organisation that might potentially affect the overall rate of turnover from the firm's point of view.

 6. The analysis of correlations performed enables the nature of the turnover studied here to be qualified – typified through the variables with significant correlation, – which provides further analysis instruments of use in construing the result that is obtained later on after analysing the impact caused by turnover expectations on readiness to engage in management training.

 7. Nevertheless, we previously obtained a correlations matrix (Table A2) in order to trace a priori an initial scenario and obtain an approximate idea of the relationship existing between both groups of variables. The results display an almost total absence of relationship between these correlations, which shows that turnover expectations do not a priori seem to have the explanatory capacity that was initially expected to determine the position adopted by organisations in respect of their investment in CMT.

 8. Specifically, we proceeded to make a mixed or two-stage analysis, by means of which, to begin with, a hierarchical type cluster analysis was made -through agglomeration and through Ward's algorithm- using squared Euclidean distance as a similarity measurement. This is how we identified the number of groups to be used for determining the initial seed set (centroids of groups) which will intervene later on in a non-hierarchical conglomerate analysis⁴ (k-means algorithm). This operation will enable us to maximise the homogeneity of each group and the heterogeneity between some conglomerates and others.

 9. The hierarchical cluster analysis shows that the optimum solution consists of 3 groups – the stopping rule used is the increase in distance within the dendogram – and this grouping solution enables us to draw out the initial centroids which we will use later on to confirm the classification by means of the k-means clustering method. The ANOVA variance analysis displays that there are significant results for the two variables included among the 3 groups formed through the cluster analysis. For this reason the three conglomerates presented are significantly different (Table A3).

10. The ANOVA variance analysis again shows that there are significant differences for the two variables included (EA₁ and EA₂) between the 2 groups formed through the cluster analysis. Consequently the two conglomerates presented are significantly different (Table A4).

11. This result is consistent with those obtained in other studies (Loewenstein and Spletzer 1997; Veum 1997; Krueger and Rouse 1998; Forrier and Sels 2003; Bassanini, Booth, Brunello, De Paola and Leuven 2005), where no significant relationship whatsoever is detected in the training-turnover binomial and stress is laid on the ambiguity found in the empirical literature which has attempted to look into the nature of this relationship. One should emphasise that while such research has identified investment in training as an explanatory variable for its analysis rather than turnover expectations – as occurs in this article – we consider that the result obtained in our study agrees with and is representative of the lack of significativity displayed by a good number of the analyses which have attempted to give such a relation empirical backing.

12. We consider that the percentage of companies that state that they have moderate expected manager turnover (25.4%) is quite important, taking into account the nature of the variable and the context studied.

13. Although on the whole the chi-squared comparison indicates that there is no significant dependence, analysis of the typified residues shows us that the behaviour of companies with a high degree of inclination towards management training is indeed not visibly affected by turnover expectations, but that, nevertheless, there does seem to exist a negative relation in companies whose readiness is low (they perceive more risk of turnover). This means that the percentage of companies that invest less in management training and have moderate expectations of turnover is higher than the percentage of companies with a medium or high degree of readiness and whose turnover expectations are moderate.

14. The average mark of the group most inclined to adopt managerial training when posed the questions ‘Covering of management jobs through internal promotion’ and ‘Company managers have taken on their first responsibility in this company’ was 3.52 and 3.59 points (out of 5), while the mark of the group with the least inclination in this direction was 2.83 and 3.03, respectively. When the Student T equality test for independent sample averages was performed, it showed that the averages for both groups are significantly different. Analysis of the correlations of the variables for the existence of an ILM with the readiness indicators also shows positive and significant results.