Implementing process innovations: the benefits of combining delegative-participative with consultative-advisory leadership.

Second, both leadership patterns are complementary. Leading others by providing advice and orientation reduces the specific risks associated with delegative-participative leadership, which consist of a subordinate's mis- or reinterpretations of the process innovation. Conversely, delegative-participative leadership attenuates the specific disadvantages connected with consultative-advisory leadership, which comprise displays of reactance on the part of the subordinate. Because each of these two leadership facets can compensate the risks of the respective other set of leadership behaviors, it is reasonable to assume that with respect to implementing innovations, these leadership styles will have the greatest beneficial effect when they are combined. By contrast, if only one of these patterns is used, the results will most likely be suboptimal. Concomitantly employing both delegative-participative leadership and consultative-advisory leadership brings to fruition the full positive potential of each of these leadership styles. Thus, these behavioral patterns can be viewed as functionally equivalent. We therefore posit,

Hypothesis 3: Consultative-advisory leadership moderates the relationship between delegative-participative leadership and the implementation success of product innovations. When levels of consultative-advisory leadership are high, this relationship is positive, whereas when levels of consultative-advisory leadership are low, this relationship is negative.

Method

Sample and Procedure

Managers (N = 388) from German organizations of different sizes and sectors were surveyed. The process innovations were treated as critical incidents (Flanagan, 1954), which has the advantage of a higher context specificity (see Krause & Kearney, 2006) and thus more valid answers in comparison to other methods. The managers were requested to recall a specific process innovation in their work unit and to describe this innovation in a qualitative and quantitative manner. They were then asked to answer questions about how they as managers were led by their immediate superior during the innovation process and to rate the degree to which the implementation was successful.

The managers were recruited for the study in three different ways. First, letters were sent to randomly selected persons in leadership positions whose names and addresses are contained in a German catalogue (called Hoppenstedt) that features 55,700 managers. Second, at airports, exhibitions, conferences, and leadership training courses, managers were approached in person and asked for their participation directly. Third, questionnaires were handed to managers by the heads of the personnel departments of various organizations. We had business-related contacts with those heads of the personnel departments and asked them if they would be willing to distribute the questionnaire to the managers working in their departments.

After a brief personal communication concerning topic and goals of the study, the questionnaire and attached letter were either handed to the participants directly along with a self-addressed stamped envelope or sent via regular mail or e-mail. The questionnaires were returned anonymously to our university via regular mail in a sealed envelope. Participation in the study was voluntary, and no remuneration was offered. In return for answering the questionnaire, those participants who were interested were supplied with aggregated and anonymous information regarding the descriptive results of the study. The strict anonymity policy that prevented us from sending reminders to the contacted managers who had not yet responded and the length of the questionnaire militated against obtaining a high response rate. However, our final response rate of 24% might indicate that selection effects of the queried managers had occurred. Hence, we examined the possibility of a response bias with respect to industry sector and organization size. Contrary to what would have been expected in the case of a response bias, our analysis revealed that our participants constituted a sufficiently representative sample of German managers. Our results can therefore be generalized to managers at different hierarchical levels and in different industries and fields of specialization.

The respondents ranged in age from 22 to 64 years (M = 39 years, 6 months; SD = 9 years, 9 months) and represented several levels of hierarchy (21% group leaders, 23% department heads, 12% division heads, 23% area heads, 16% general managers, 5% members of the managing board) and areas of expertise (27% from marketing and sales, 22% from business administration, 21% from personnel and organization design departments, 9% from production, 7% from research and development, 6% from technical support, and--because of missing data--8% from unknown areas). Of the respondents, 82% were men. This overrepresentation of men reflects the current gender disparity in German organizations. The sectorial distribution of the managers in the sample was broad (19% in banks and insurance companies; 18% in telecommunications, data processing, or the media; 16% in services; 8% in construction; 8% in mechanical engineering and the automotive industry; 8% in trade; 5% in the chemical industry; 5% in utility companies; 2% in the food industry; and--again because of missing data--11% from unknown industries). The distribution of the managers with respect to the size of their respective organization showed that most of the managers (81%) worked either in medium-sized companies (with up to 500 employees) or large companies (with 501 to 5,000 employees).

Measures

Measurement of leadership. The instruction in the part of the questionnaire pertaining to leadership was, "Please evaluate the extent to which your superior used the following leadership strategies during the innovation process." Leadership was assessed using a 6-item scale (7-point Likert scale, ranging from 1 = strongly agree to 7 = strongly disagree), whose construct and criterion validity have been shown to be satisfactory (Krause, 2004). The leadership scale was developed in three steps. First, we developed a pool of 82 leadership items pertaining to process innovations. In a pretest, we calculated item characteristics and scale dimensionality and reliability. Based on the results of the pretest, we selected 18 items for our final leadership scale. Second, we tested the structure of this scale in a different sample and found five leadership components (Krause, 2004). Third, we evaluated these five leadership facets with respect to their relevance concerning our criterion implementation success. We identified two leadership patterns that we consider to be most important in regard to implementation success, namely, delegative-participative leadership and consultative-advisory leadership.

We checked the factor structure of the employed leadership scale by means of a principal components analysis with varimax rotation. The results confirmed our assumption of two leadership factors, which together explained 71% of the variance. The items of the first factor were "During the innovation process, my superior gave me many opportunities to contribute to shaping this innovation in my area of responsibility;" "During the innovation process, my superior granted me autonomy, degrees of freedom, and decision-making authority;" "During the innovation process, my superior presented me with a fait accompli (reversed)." This factor is interpretable as delegative-participative leadership (Cronbach's [alpha] = .82). The items of the second factor were "During the innovation process, my superior helped me in solving complicated issues;" "During the innovation process, my superior shared with me his or her professional ideas;" "During the innovation process, my superior provided me with all the important information." This factor can be construed as consultative-advisory leadership ([alpha] = .70).

Measurement of implementation success. Implementation success was measured with four newly developed items on a 7-point Likert scale, ranging from 1 = unsuccessful to 7 = successful. The items were "How would you rate the overall success of the implementation of the process innovation?" "How would you rate the success of the implementation of the process innovation with respect to the effectiveness and efficiency of your work unit?" "How would you rate the success of the implementation of the process innovation with respect to your initial hopes and fears?" "How would you rate the success of the implementation of the process innovation with respect to unexpected side effects?" All of these items loaded on one factor ([alpha] = .91) that explained 79% of the variance.

Control variables. We included job tenure (which ranged from 1 year to more than 10 years) and degree of innovation as control variables. The respondents' job tenure can be interpreted as experience in their jobs, which may have an impact on the level of implementation success. Furthermore, the degree of innovation needs to be controlled because there is reason to believe that implementation barriers rise with increasing degree of innovation. The degree of innovation (7-point Likert scales) was operationalized by three criteria: scope ("The chosen new process differs strongly from the processes employed by my organization in the past"), initiative ("My work unit was the first to develop and/or implement this novelty"), and ramifications ("Through this novelty, the extant structures of power, control, and competencies were altered significantly"). These items loaded on one factor that explained 45% of the variance. The reliability of this scale was modest ([alpha] = .60).

Results