The Romance of LEADERSHIP SCALE Cross Cultural Testing and Refinement
The Romance of LEADERSHIP SCALE Cross Cultural Testing and Refinement
The Romance of LEADERSHIP SCALE Cross Cultural Testing and Refinement
Published by:
http://www.sagepublications.com
Additional services and information for Leadership can be found at: Email Alerts: http://lea.sagepub.com/cgi/alerts Subscriptions: http://lea.sagepub.com/subscriptions Reprints: http://www.sagepub.com/journalsReprints.nav Permissions: http://www.sagepub.co.uk/journalsPermissions.nav Citations http://lea.sagepub.com/cgi/content/refs/3/1/29
Leadership
Abstract The Romance of Leadership Scale (RLS) has been used in various studies in different countries and contexts. However, to date, the structure of the scale has been a subject of discussion, making it difcult to compare results over different studies. In this study, using student as well as organization samples from two countries, we want to clarify the factor structure of the RLS. In order to do so, we used a hypothetical factor matrix into which we rotated our data. Although this matrix ts some of the data quite well, the results argue for the use of one core factor. The factor solutions are, however, still ambiguous and we therefore recommend doing more research on a core factor of the Romance of Leadership scale. Keywords culture; leadership; romance of leadership; organizational context
Introduction
Leadership is a complex matter that can be seen from several different perspectives. The focus of research has predominantly been on the leader and his/her behaviour or his/her traits. Recently, researchers have become more interested in the followers perspective of leadership. In this article, we are focusing on an example of followercentred research, namely, the Romance of Leadership (e.g. Meindl et al., 1985). In this approach, leadership is regarded from a social constructivist point of view (Meindl, 1998a) and is dened as an experience undergone by followers (Meindl, 1993: 97). This means that individuals are actively involved in constructing leadership rather than that leadership being simply what a leader does. The Romance of Leadership approach predicts that in times of good or poor company performance, people will attribute high levels responsibility to leaders thereby ignoring other possible inuencing factors, such as the general economic situation. For organizations, this romantic view of leaders can have negative consequences, for example, if leaders are replaced in bad times without any change in the company performance resulting from that (Meindl, 1990, 1993; to regard this in a broader context, imagine football or soccer coaches being replaced without the performance of the team improving).1
Copyright 2007 SAGE Publications (London, Thousand Oaks, CA and New Delhi) Vol 3(1): 2946 DOI: 10.1177/1742715007073063 www.sagepublications.com
Downloaded from http://lea.sagepub.com by Tomislav Bunjevac on May 4, 2009
Leadership
3(1) Articles
To assess this phenomenon, Meindl and Ehrlich (1988) established a measure of assessment for the Romance of Leadership. Since then, the instrument has been widely used in different contexts (experimental research with student samples versus eld studies with employees) and countries (e.g. the USA: Awamleh & Gardner, 1999; Germany: Felfe, 2005). However, in different studies, different factor structures emerged (e.g. Awamleh & Gardner, 1999; Cho & Meindl, no date). The focus of our article is, therefore, to analyse the factor structure of the Romance of Leadership scale on both theoretical and empirical grounds and then to examine this structure in different countries and different contexts.
30
Downloaded from http://lea.sagepub.com by Tomislav Bunjevac on May 4, 2009
Leadership
the construction of leadership because different styles of leadership lead to different social constructions. Similarly, Schneider (1998) emphasizes that shared perceptions are really grounded in the attributes and behaviours of leaders. Without taking the leader into account, in his view, the stimulus or basis of social construction (p. 311) is missing. He argues that followers experience the same substantive issues (p. 313) at work and, therefore, social construction without taking into account these circumstances is insufcient. In addition, Schneider maintains that in organizations, people with similar personalities interact, and through this interaction come to shared meanings. To summarize, Romance of Leadership has been criticized for not taking into account the leader and the situation in which followers act. Therefore, it has been regarded as too narrow an approach to leadership. In a response to this critique, Meindl (1998b) stresses that this approach focuses on processes not directly connected to actual leader behaviour and characteristics. He underlines that Romance of Leadership is simply an alternative way of looking at leadership. Rather than narrowing the view on leadership, he claims to broaden it.
Empirical evidence
In the following, we provide a short overview of studies on the Romance of Leadership. In their initial examination of the approach, Meindl et al. (1985) conducted several studies to show the tendency to romanticize leadership in several areas of society. Analysing business journals, they found that leaders are often highlighted when their company is doing well. Setting Romance of Leadership in the context of nationwide economic tendencies, they found that dissertations, as well as business periodicals, focus more strongly on leaders when the economy is doing badly (dissertations) or well (business periodicals).2 This implies that leaders are seen as having the ability to control and inuence the fates of the organizations they run, whereas other factors, such as the economic situation, are more or less neglected. In 1987, Meindl and Ehrlich conducted two studies that showed the effect of attributions to leadership on outcome evaluations. When participants received an explanation for company performance that referred to leadership (as opposed to employees, markets or government policies) as being responsible for company performance, they evaluated the performance of that company as more protable and less risky. In a second study, Meindl and Ehrlich (1987) found that participants rated stability, internality and controllability higher for the leader and employee explanation of company success than those explanations related to markets or government policies. In addition, in the leadership explanation, attributions of company performance to the top management were higher than in the other conditions. Attributions to the companys environment or to chance or luck were higher in the markets or the government policies condition than when the participants received an explanation of company success referring to leaders or employees. This means that participants differentiated between company internal and external factors in relation to their inuence over companies performance. With these two initial studies, one might say that the phenomenon of Romance of Leadership was proven. Research from here on mainly concentrated on the effects of Romance of Leadership rather than the phenomenon itself.
31
Downloaded from http://lea.sagepub.com by Tomislav Bunjevac on May 4, 2009
Leadership
3(1) Articles
Ehrlich et al. (1990) conducted a rst study on the effect of Romance of Leadership on the perception of specic leadership styles. They examined the effect of Romance of Leadership on the perception of charisma and leader behaviour as assessed using the LBDQ (Stogdill, 1963). Although they did not nd any signicant relationship between Romance of Leadership and charisma, Romance of Leadership and leader behaviour were signicantly related. In addition, Romance of Leadership and charisma overlapped in explaining the perception of leader behaviour. In line with Meindls (1990) assumption that leadership, especially charismatic leadership, is an output of social psychological forces operating among followers, subordinates and observers, rather than arising directly out of the interactions between follower and leaders (p. 188), other researchers have focused on the effect of Romance of Leadership on the perception of actual leaders (Al-Dmour & Awamleh, 2002; Awamleh, 2003; Awamleh & Gardner, 1999; Meindl, 1990; Schyns et al., forthcoming; Schyns & Sanders, 2004). The aim of these studies was to examine the extent to which a followers individual degree of Romance of Leadership impacts on the perception of transformational/charismatic leadership. Transformational/charismatic leadership was the focus of these studies, as Meindl (1990: 182) claimed that transformational/charismatic leadership is itself a type of hyper-romanticism. The results of these studies are mixed. Some authors (Al-Dmour & Awamleh, 2002; Awamleh, 2003; Meindl, 1990) found a positive relationship between Romance of Leadership and the perception of transformational/charismatic leadership. Shamir (1992) found that the attribution of inuence to the leader, as well as general agreement on the importance of leadership, is positively related to the attribution of charisma to a leader. However, others could not conrm this relationship (Awamleh & Gardner, 1999; Schyns & Sanders, 2004). These mixed results may require more complex analyses. Haslam et al. (2001) found that the attribution of company performance to a leader is not only dependent upon the type of company performance (crisis turnaround, stable prot, stable loss, or crisis decline) but also on the way leaders behave in terms of afrming or negating the identity of the group they lead. Bligh et al. (2005) conducted a study on the 2003 California recall election and how charismatic and effective the candidates (the present governor, the challenger from his own party and an outside challenger from another party) were seen to be.3 As a result of their study, they go one step further than previous research, in that they enhance our knowledge of the conditions under which Romance of Leadership is related to the perception of charisma. Bligh and colleagues found that for the outside challenger Romance of Leadership was indeed related to the perception of charisma, though this was not the case for the other candidates. They found no relationship between Romance of Leadership and expected effectiveness. With respect to the interaction between Romance of Leadership and a perceived crisis in California, they found that for the outside challenger there was indeed a higher correlation between Romance of Leadership and the perception of charisma among those participants who perceived California to be in a crisis. The opposite was true for the challenger from the party of the governor: for those participants who perceived California to be in a crisis, the correlation between Romance of Leadership and the perception of charisma was negative.
32
Downloaded from http://lea.sagepub.com by Tomislav Bunjevac on May 4, 2009
Leadership
33
Downloaded from http://lea.sagepub.com by Tomislav Bunjevac on May 4, 2009
Leadership
3(1) Articles
of a company as the quality of its top-level leaders and When the top leaders are good, the organization does well; when the top leaders are bad, the organization does poorly. The two items loading highest on the second factor (four items) are: Sooner or later, bad leadership at the top will show up in decreased organizational performance and The great amount of time and energy devoted to choosing a leader is justied, because of the important inuence that person is likely to have. The last factor consists of three items, such as The process by which the leaders are selected is extremely important and Its probably a good thing to nd something out about the quality of the top-level leader before investing in a rm. Recently, Bligh et al. (2005) used the Romance of Leadership in the context of a study on the election in California. They found a two-factor solution for the Romance of Leadership Scale, with the reverse-coded items loading on a separate factor. Considering these mixed results, it would appear to be very useful to establish a factor solution that is stable across different contexts. In a rst endeavour to clarify the structure of the Romance of Leadership Scale, Schyns et al. (2004) constructed a hypothetical factor matrix on the basis of the content of the 32 items contained in the Romance of Leadership Scale. They were able to differentiate three factors. The rst or core factor, referred to as the inuence of a leader, is the one that best represents the theory. It contains items reecting beliefs concerning the extent to which a leader is able to affect organizational outcomes. The second factor emphasizes the interchangeability of leaders. Although the label given to this second factor suggests that it is nothing more than the opposite of the rst factor, some qualitative differences have to be highlighted. It is true that the rst factor focuses on one leader who is considered to be completely responsible for everything that happens in an organization, whereas the second factor emphasizes the exchangeability of leaders. However, it must be noted that the belief in the exchangeability of a leader does not necessarily preclude the belief that the leader is still very inuential. The third factor concentrates on the signicance of the inuence of other factors, in addition to leadership, that may impact on an organizations performance (see Appendix for the grouping of the items). In order to examine the structure of the Romance of Leadership Scale, Schyns et al. (2004) drew three samples of German participants: a student sample (N = 146), a heterogeneous organizational sample (N = 104) and a homogeneous organizational sample (N = 202) (for a detailed description see Schyns et al., 2004). The results of their analyses showed that, although there were some differences in the number of items that could be grouped to the factors, the three-factor solution continued to t the data better than could be expected by chance. However, this prior research had focused only on the German translation (an overview of the results of prior studies using the Romance of Leadership Scale is depicted in Table 1). Further tests are needed to examine other language versions, especially the original English instrument. In this study, we will apply the same hypothetical matrix to Dutch and US data in order to examine further the extent to which these factors can be used for future research.
34
Downloaded from http://lea.sagepub.com by Tomislav Bunjevac on May 4, 2009
Leadership
Table 1 Overview of the factor found for the Romance of Leadership Scale in prior research
Sample Jordan, employees Jordan, employees 202 3 (32) 1) Different effects for different leaders 2) Absolute effects of leaders 3) Relative effects of leaders 1) There is nothing as crucial to the bottom-line performance of a company as the quality of its top-level leaders and When the top leaders are good, the organization does well; when the top leaders are bad, the organization does poorly. 2) Sooner or later, bad leadership at the top will show up in decreased organizational performance and The great amount of time and energy devoted to choosing a leader is justied, because of the important inuence that person is likely to have. 3) The process by which the leaders are selected is extremely important and Its probably a good thing to nd something out about the quality of the top-level leader before investing in a rm. 1) Inuence of a leader 2) Interchangeability of a leader 3) Inuence of other factors 1) Positive-coded items 2) Reverse-coded items 1 (11) 1 (11) N of factors (items) Name of factor/example items*
Authors
Year
Topic
2002
Transformational leadership
Awameh
2003
Transformational leadership
undated
Work ethics
1999
Charisma
USA, students
3 (11)
Schyns et al.
2004
Factor structure
3 (32)
Bligh et al.
2005
Charisma
*Note. Awamleh and Gardner (1999) did not name the factors they found. Therefore, to better understand the contents of the factors, example items are given.
35
Leadership
3(1) Articles
Method
Samples and procedures
The Netherlands Prior to the collection of data, a Dutch and an American scientist made independent Dutch translations of the RLS. Minimal deviations were discussed with James R. Meindl, who nally approved the translation.4 Organization sample (heterogeneous; Sample 1). Students from a course on research methods received course credit for gathering the data for this sample from their acquaintances. The questionnaire was constructed identically for all students/ participants and contained several instruments on leadership. Only the data referring to Romance of Leadership are reported here. Two-hundred and ninety-two respondents from different companies lled in the questionnaire. Of these, 145 were male, 147 female. The mean age was 36.5 years (SD = 12.9). On average, the respondents had worked for 9.4 (SD = 10.4) years in their company. Overall 29.8 per cent of the participants have or had previously been working as a manager. Organization sample (homogeneous; Sample 2). The participants were personally approached in the workplace. After being asked if they wanted to take part in the research, the participants were handed the questionnaire. The surveys were returned personally or sent back via mail. In the questionnaire, participants were asked for demographic data concerning age, sex, discipline, and work experience before they were requested to ll in the rest of the questionnaire. The 191 participants were employees of an accountancy agency in The Netherlands (54% accountancy, 18% support, 9% tax advice, 7% consultancy, 5% other and 7% didnt indicate their discipline). Of the participants, 47 per cent were men and 52 per cent were women (1% did not indicate their gender). The mean age was 34.3 (SD = 10.0). The mean years of work experience was 14.9 (SD = 10.4). Twenty-ve per cent of the participants had had work experience as a manager. Student sample (Sample 3). The participants were asked during lectures whether or not they wanted to take part in the study. The questionnaire was handed to them and they were asked to give it back within two weeks. In the questionnaire, they were asked for demographic data concerning age, sex, discipline, and work experience before they were requested to ll in the rest of the questionnaire. The questionnaires were returned personally or via internal mail (in the case of psychology students, in exchange for course credits). The 145 participants were students of the following subjects: psychology students (n = 84), human resource students (n = 56), and students of other majors (n = 5). Most of them (117) were women (22 men; 6 did not indicate their gender). The mean age was 21.3 (SD = 4.0). Of the students, 97.8 per cent had had work experience in different lines of businesses (prot and non-prot).
36
Downloaded from http://lea.sagepub.com by Tomislav Bunjevac on May 4, 2009
Leadership
USA Student sample (Sample 4). Three-hundred and sixty-four US students lled in the questionnaire. Of the participants, 214 were male and 149 female. The mean age was 21.4 years (SD = 2.5). Students were majoring in different subjects, namely, accounting (21.2%), management science (20.9%) marketing (18.7%), nance (14.6%), business administration (12.9%), human resources studies/organizational behaviour (9.3%), and others (2.5%).
Analysis
Treatment of missing values The samples contained few missing values. These were replaced by an estimate derived from an appropriate regression analysis of the variable in which the missing value occurred on the remaining items in the RLS. This regression imputation method was carried out by means of the Missing Value Analysis procedure provided by SPSS 11.5. A randomly selected regression residual was added to the estimates of the missing values (see Schafer, 1997). Factor extraction and rotation In order to examine the extent to which the data t our hypothetical matrix, we determined a three-factor principal axis solution for each sample. In all analyses, the number of common factors was chosen on the basis of a Parallel Analysis, as described by Humphreys and Montanelli (1975). We then orthogonally rotated each principal factor solution into the hypothetical matrix so that maximal agreement occurred, meaning that the items load high on the factor they are supposed to load on and zero on the other factors. We then tested the congruence of the rotated factor matrix with the hypothetical matrix, using Tuckers factor congruence coefcient (Korth & Tucker, 1976). In order to test the signicance of our solution, we rotated each factor solution into matrices that consisted of random permutations of the rows of the original hypothetical matrix and compared the congruence coefcients to our solution. The probability level of the coefcient was determined under a normal distribution with the mean and standard deviation of the simulated distribution. Although we could have carried out conrmatory analyses, we decided to restrict ourselves to explanatory factor analyses, as we were also interested in describing the differences between the factor solutions obtained in the different subgroups.
Results
Testing the signicance of the agreement of factors with the hypothetical factor solution
Table 2 contains the results of the congruence tests. This table shows the observed congruence coefcient along with its signicance under the random permutation distribution for each factor in each sample. For Samples 1, 2, and 3, the congruence coefcients for the rst two factors are signicantly higher than would be expected by chance. In all three samples, the congruence coefcient for the third factor does
37
Downloaded from http://lea.sagepub.com by Tomislav Bunjevac on May 4, 2009
Leadership
3(1) Articles Table 2 Observed congruence coefcients and the results of the permutation tests
Factor 1 Sample 1 M SD z p Sample 2 M SD z p Sample 3 M SD z p Sample 4 M SD z p .946 .602 .077 4.456 .000 .828 .403 .101 4.191 .000 .796 .283 .110 4.647 .000 .871 .464 .107 3.806 .000 .726 .344 .119 3.207 .001 .438 .254 .111 1.658 .049 .903 .582 .089 3.633 .000 .751 .397 .102 3.482 .000 .306 .278 .111 .252 .401 .923 .606 .079 4.013 .000 .789 .410 .103 3.702 .001 .445 .284 .113 1.420 .077 Factor 2 Factor 3
not reach signicance. This means that in relation to the data for the rst two factors the hypothetical matrix t is signicantly better than would be the case in a chance matrix. However, this is not true for the third factor. In the fourth sample, the congruence coefcients for all three factors are signicantly higher than would be expected by chance.5 This means that, according to the overall agreement test, the solution for the US students ts best. In the next section, we will take a closer look at the item structure of the respective solutions.
38
Downloaded from http://lea.sagepub.com by Tomislav Bunjevac on May 4, 2009
Leadership
The Romance of Leadership Scale Schyns et al. Table 3 Optimally rotated factor solutions
Sample 1 Item 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 1 .487 .007 .118 .552 .402 .510 .131 .530 .499 .272 .258 .481 .242 .442 .314 .327 .146 .234 .363 .501 .375 .028 .233 .606 .592 .026 .049 .324 .145 .203 .067 .300 2 .088 .323 .463 .041 .365 .313 .407 .015 .210 .353 .477 .121 .577 .110 .021 .074 .428 .491 .232 .054 .344 .405 .415 .191 .029 .447 .183 .108 .157 .079 .570 .198 3 .021 .078 .227 .251 .257 .299 .127 .003 .030 .051 .071 .145 .155 .118 .111 .004 .238 .195 .118 .064 .102 .292 .385 .280 .262 .292 .259 .279 .252 .284 .012 .181 1 .395 .136 .182 .360 .323 .310 .093 .505 .520 .213 .315 .564 .049 .596 .475 .260 .011 .079 .487 .631 .466 .046 .379 .541 .619 .082 .121 .433 .306 .382 .109 .403 Sample 2 2 .014 .261 .335 .110 .404 .328 .182 .013 .104 .141 .485 .021 .502 .096 .026 .227 .465 .525 .156 .022 .324 .199 .431 .118 .074 .255 .350 .157 .171 .187 .487 .170 3 .304 .056 .014 .105 .098 .209 .042 .119 .264 .003 .128 .282 .002 .112 .058 .106 .244 .254 .404 .043 .190 .158 .218 .398 .469 .199 .328 .110 .325 .184 .064 .061 1 .387 .014 .041 .425 .369 .111 .023 .334 .538 .091 .245 .520 .082 .460 .378 .381 .396 .174 .124 .499 .482 .091 .311 .638 .610 .016 .127 .356 .318 .325 .098 .395 Sample 3 2 .076 .291 .151 .006 .449 .139 .185 .078 .036 .143 .637 .042 .520 .079 .070 .097 .434 .532 .438 .208 .292 .310 .321 .118 .140 .446 .209 .069 .333 .167 .378 .191 3 .284 .396 .488 .009 .117 .172 .008 .104 .369 .164 .068 .078 .081 .298 .150 .015 .272 .008 .195 .205 .015 .117 .367 .139 .289 .291 .146 .249 .237 .221 .192 .026 1 .339 .056 .145 .265 .369 .426 .001 .315 .427 .063 .088 .442 .010 .384 .392 .437 .159 .085 .333 .463 .350 .045 .026 .567 .552 .040 .020 .250 .053 .043 .057 .186 Sample 4 2 .014 .390 .465 .060 .245 .238 .353 .159 .158 .285 .422 .023 .528 .019 .142 .076 .559 .428 .319 .034 .287 .501 .463 .193 .185 .582 .507 .092 .067 .098 .593 .090 3 .061 .318 .096 .056 .161 .150 .304 .087 .018 .004 .119 .191 .082 .136 .020 .003 .091 .082 .056 .040 .138 .138 .306 .070 .074 .153 .026 .084 .367 .524 .278 .119
Note. Numbers in italics indicate the highest factor loading of this item. Bold indicates the factor on which the item should theoretically have the highest loading. Consequently, bold-italics indicate when an item has the highest loading on the factor it hypothetically belongs to.
factor solution to a more detailed item-wise investigation of the differences and similarities between the factor solutions in the four samples. For the latter strategy the risk of capitalizing on chance uctuations may be rather high. However, some differences can be highlighted.
39
Downloaded from http://lea.sagepub.com by Tomislav Bunjevac on May 4, 2009
Leadership
3(1) Articles
Regarding factor 1, two items load especially high in all samples: item 24 (When the top leaders are good, the organization does well; when the top leaders are bad, the organization does poorly) and item 25 (Theres nothing as critical to the bottom line performance of a company as the quality of its top-level leaders). In addition, item 8 (High- versus low-quality leadership has a bigger impact on a rm than a favorable versus unfavorable business environment) and item 20 (When a company is doing poorly, the rst place one should look to is its leaders) load especially high in the employee samples. A closer look at items loading on the second factor shows that this factor could have been interpreted somewhat differently in the four samples. Whereas in the rst and second sample, interchangeability and irrelevance are emphasized, in the third sample this factor consists of items referring to attributing no inuence to leaders, as well as to denying that leaders deserve the attention they get (leaders selection, salary and investment). In the fourth sample, in addition to items referring to interchangeability, items referring to leaders having little impact are included in this factor. For the third factor, similar interpretation differences apply: in the rst and second sample, only one or two items load on the factor referring to lack of inuence and chance aspects. In the third sample, items reecting a kind of resignation with respect to company performance, as well as the aspect of chance, dene the third factor. In the fourth sample, the only two items making up the third factor refer to the fact that no responsibility for company performance should be attributed to leaders.
40
Downloaded from http://lea.sagepub.com by Tomislav Bunjevac on May 4, 2009
Leadership
situations where they cannot rely on their followers to back them (here, think again of sport teams). Additionally some differences emerged between the student and employee samples on factor 1. This might hint at structural differences between employee and student samples, although further research is needed to conrm this result, before engaging in a stronger interpretation of our ndings. Future research could focus on the relationship between the two main factors (inuence of a leader and interchangeability of a leader). We can assume that the relationship between these factors will be different for different individuals and groups of followers. Four groups of followers might be differentiated depending on their values on the two factors: The rst group (high inuence/high interchangeability) contains of people who assume that leaders assumed to have a high impact in the organisation. However, this impact is independent of a specic person. Rather leaders in general have a high impact, maybe due to their position. We could call this collective Romance of Leadership. The second group (high inuence/ low interchangeability) could be called the classical romanticizers: they attribute high responsibility for company performance to a specic leader, in that sense that they assume that this specic person makes a strong difference to the company he/she works in. We can call this individualized Romance of Leadership. The third group (low inuence/high interchangeability) does not show Romance of Leadership at all. They assume that leaders do not impact on organizational performance and also that leaders are interchangeable. The fourth group (low inuence/low interchangeability) again does not show Romance of Leadership. However, this maybe due to a specic person, meaning that not all leaders lack inuence but specic ones do. When differentiating between these groups, an interesting line of research would be to examine the different attitudes the group members have towards their actual leader. These considerations mean that both directions of the relationship between the two factors make sense. However, it would be interesting to examine the nature of Romance of Leadership in organizations depending on the relationship between the two factors. We can assume that followers who attribute a lot of responsibility to leaders but still think that one leader is like any other may show little loyalty to their current leader. On the other hand, followers who attribute a lot of responsibility to leaders may just as likely think that these leaders are unique and that they cannot be exchanged for one another. This could lead to followers placing importance on being able to work with a particular leader. More research is needed to extend the third factor (inuence of other factors). In our study, it turned out to be the least stable factor. Nevertheless, it covers a theoretically interesting aspect of leadership. Meindl and Ehrlich (1987) already examined the degree to which people attribute responsibility for company performance to other factors rather than leadership when stimulated to do so. The third factor contains ideas that tap into a certain idea, namely the attribution of company success to other factors besides leadership. Strictly speaking, the attribution of responsibility to other factors than leadership may not be Romance of Leadership. However, when people explicitly deny the inuence of other factors, their romanticization of leaders is indeed strong. For future research, it would be useful to have a better assessment of this aspect of Romance of Leadership. In addition, it could be interesting for future research on Romance of Leadership
41
Downloaded from http://lea.sagepub.com by Tomislav Bunjevac on May 4, 2009
Leadership
3(1) Articles
to give people a choice between a range of factors that might explain company success, for example, leadership, the state of the economy or HR policies and so on. Doing so could shed light on the question whether people attribute externally (in the sense of not feeling responsible for the companys success themselves but making other factors responsible) or to leaders in particular. Another line of research should involve further examinations of the structure and comparability of measurement of the Romance of Leadership Scale in other countries and cultures. Although, to date, a total of seven samples have been included in the testing for the factor structure of the RLS, all three of the countries involved were western. For future research, samples from other countries should be drawn from countries that differ from those considered here with respect to Hofstedes (2001) cultural dimensions. In addition to expecting different degrees in the romanticizing of leaders, we can assume the structure of the instrument to be different in, for example, high power-distance cultures. Interchangeability of leaders may be less prominent in these countries, given that they are individualistic as well. A strong sense of hierarchy connected with emphasis on the individual can turn Romance of Leadership into having high expectations of ones actual leader. In high power distance, high collectivist countries, however, interchangeability may be more strongly related to the attribution of responsibility, meaning that emphasis is put on the top, rather than on one specic person. Although we did not test for cultural differences and could not do so given the nature of our samples (all from western countries), we still found that the second factor comprises slightly different items in the samples. This may be hinting at different aspects included in interchangeability, namely, that it contains romanticizing leadership (no matter who it is, all leaders are highly inuential on company performance) or imply a negative notion of leaders impact (it makes no difference who is at the top, leaders have no inuence). Only when having a close look at the different items that comprise this factor in different samples and in relationship to the rst factor can we really determine the degree of Romance of Leadership. Future research could focus on uncovering the extent to which these different aspects of interchangeability are culturally determined.
Acknowledgements
Requests for reprints should be addressed to the rst author. An earlier version of this article was presented at the First Conference on Cross-cultural Leadership and Management Studies, Seoul, South Korea, 1012 June 2004. The authors would like to thank the anonymous reviewers for their helpful comments and Jule Epp and Gill Christy for proofreading.
Notes
1. We thank Hartmut Blank for this very illustrative example. 2. Interestingly, we can nd a nice example in the German press: in the last years, growth in Germany was well below expectations and, during this time, a major weekly newspaper (Die Zeit) put forward two different series on leaders in Germany, one called Moments of Decision (February 2003 to June 2004) and What Drives . . . (October 2004 to today). 3. Only the results for Romance of Leadership are reported here.
42
Downloaded from http://lea.sagepub.com by Tomislav Bunjevac on May 4, 2009
Leadership
The Romance of Leadership Scale Schyns et al. 4. To give an example of differences: the English original organization has the connotation of referring to a very big company both in Dutch and German. Therefore, we preferred to use the term company throughout the questionnaire. 5. Note however that although many congruence coefcients reach an acceptable signicance level, the values they attain are not so high as to indicate perfect agreement with the hypothetical factors.
References
Al-Dmour, H., & Awamleh, R. A. (2002) Effects of Transactional and Transformational Leadership Styles of Sales Managers on Job Satisfaction and Self-perceived Performance of Sales People: A Study of Jordanian Manufactoring Public Shareholding Companies, Dirasat: Administrative Sciences 29: 24761. Awamleh, R. A. (2003) A Test of the Transformational Leadership Model: The Case of Jordanian Banks. Paper presented at the Academy of International Business (UK Chapter), Leicester, UK, 1112 April. Awamleh, R. A., & Gardner, W. L. (1999) Perceptions of Leader Charisma and Effectiveness: The Effects of Vision Content, Delivery, and Organizational Performance. Leadership Quarterly 10: 34573. Bligh, M. C., Kohles, J. C., & Pillai, R. (2005) Crisis and Charisma in the California Recall Election, Leadership 1: 32352. Cho, G.-S., & Meindl, J. R. (no date) The Romance of Leadership and Work Ethics: Occidental and Oriental Variations. Unpublished manuscript. Day, D. V., & Lord, R. G. (1988) Executive Leadership and Organizational Performance: Suggestions for a New Theory and Methodology, Journal of Management 14: 45364. Ehrlich, S. B. (1998) Leader-centric and Follower-centric Research at Multiple Levels of Analysis: Toward a Balanced Perspective, in F. Dansereau & F. J. Yammarino (eds) Leadership: The Multiple-level Approaches Part B: Contemporary and Alternative, pp. 3039. Stamford, CT: Jai Press. Ehrlich, S. B., Meindl, J. R., & Viellieu, B. (1990) The Charismatic Appeal of a Transformational Leader: An Empirical Case Study of a Small, High-technology Contractor, Leadership Quarterly 1: 22948. Felfe, J. (2005) Personality and Romance of Leadership, in B. Schyns & J. R. Meindl (eds) The Leadership Horizon Series Implicit Leadership Theories: Essays and Explorations, pp. 199225. Greenwich, CT: Information Age Publishing. Haslam, S. A., Platow, M. J., Turner, J. C., Reynolds, K. J., McGarty, C., Oakes, P. J., Johnson, S., Ryan, M. K., & Veenstra, K. (2001) Social Identity and the Romance of Leadership: The Importance of Being Seen to Be Doing It for Us, Group Processes and Intergroup Relations 4: 191205. Hofstede, G. (2001) Cultures Consequences. Thousand Oaks, CA: SAGE. Humphreys, L. G., & Montanelli, R. G. (1975) An Investigation of the Parallel Analysis Criterion for Determining the Number of Common Factors, Multivariate Behavioral Research 10: 193206. Korth, B., & Tucker, R. L. (1976) Procrustes Matching by Congruence Coefcients, Psychometrika 41: 5315. Kroon, B. (2005) The Role of Implicit Organization Theory in the Start-up Phase of New Firms, in B. Schyns & J. R. Meindl (eds) The Leadership Horizon Series Implicit Leadership Theories: Essays and Explorations, pp. 33348. Greenwich, CT: Information Age Publishing. Meindl, J. R. (1990) On Leadership: An Alternative to the Conventional Wisdom, Research in Organizational Behavior 12: 159203.
43
Downloaded from http://lea.sagepub.com by Tomislav Bunjevac on May 4, 2009
Leadership
3(1) Articles Meindl, J. R. (1993) Reinventing Leadership: A Radical, Social Psychological Approach, in J. K. Murnighan (ed.) Social Psychology in Organizations, pp. 89118. Englewood Cliffs, NJ: Prentice-Hall. Meindl, J. R. (1998a) The Romance of Leadership as a Follower-centric Theory: A Social Construction Approach, in F. Dansereau & F. J. Yammarino (eds) Leadership: The Multiple-level Approaches Part B: Contemporary and Alternative, pp. 28598. Stamford, CT: JAI Press. Meindl, J. R. (1998b) Thanks And Let Me Try Again, in F. Dansereau & F. J. Yammarino (eds) Leadership: The Multiple-level Approaches Part B: Contemporary and Alternative, pp. 3216. Stamford, CT: JAI Press. Meindl, J. R. (1998c) Appendix: Measures and Assessments for the Romance of Leadership Approach, in F. Dansereau & F. J. Yammarino (eds) Leadership: The Multiple-level Approaches Part B: Contemporary and Alternative, pp. 199302. Stamford, CT: JAI Press. Meindl, J. R., & Ehrlich, S. B. (1987) On Leadership: An Alternative to the Conventional Wisdom, Research in Organizational Behavior 12: 159203. Meindl, J. R., & Ehrlich, S. B. (1988) Developing a Romance of Leadership Scale. Proceedings of the Eastern Academy of Management, 1335. Meindl, J. R., Ehrlich, S. B., & Dukerich, J. M. (1985) The Romance of Leadership, Administrative Science Quarterly 30: 78102. Schneider, B. (1998) There Is Some There, in F. Dansereau & F. J. Yammarino (eds) Leadership: The Multiple-level Approaches Part B: Contemporary and Alternative, pp. 31119. Stamford, CT: JAI Press. Schyns, B., & Sanders, K. (2004) Impliciete Leiderschapstheorien en de Perceptie van Transformationeel Leiderschap: Een Replicatie van Duits Onderzoek (Implicit Leadership Theories and the Perception of Transformational Leadership: A Replication of a German study), Gedrag en Organisatie 17: 14354. Schyns, B., Felfe, J., & Blank, H. (forthcoming) Is Charisma Hyper-Romanticism? Empirical Evidence from New Data and a Meta-analysis, Applied Psychology: An International Review. Schyns, B., Meindl, J. R., & Croon, M. A. (2004) Romance of Leadership: berprfung eines US-Amerikanischen Konzepts im Deutschen Sprachraum (The Romance of Leadership: Examination of a US-American Concept in Germany). University of Tilburg, The Netherlands: Unpublished working paper. Shamir, B. (1992) Attribution of Inuence and Charisma to the Leader: The Romance of Leadership Revisited, Journal of Applied Social Psychology 22: 386407. Stogdill, R. M. (1963) Manual of the Leader Behavior Description Questionnaire Form XII. Columbus, OH: Bureau of Business Research, Ohio State University.
Appendix
Item 1) When it comes right down to it, the quality of leadership is the single most important inuence on the functioning of an organization. 4) Anybody who occupies the top-level leadership positions in an organization has the power to make or break the organization. 5) The great amount of time and energy devoted to choosing a leader is justied, because of the important inuence that person is likely to have. 1 x 2 3
x x
44
Downloaded from http://lea.sagepub.com by Tomislav Bunjevac on May 4, 2009
Leadership
Item 6) Sooner or later, bad leadership at the top will show up in decreased organizational performance. 8) High versus low quality leadership has a bigger impact on a rm than a favourable versus unfavourable business environment. 9) It is impossible for an organization to do well unless it has high quality leadership at the top. 12) A company is only as good or as bad as its leaders. 14) With a truly excellent leader, there is almost nothing that an organization cant accomplish. 15) Even in a bad economy, a good leader can prevent a company from doing poorly. 16) Top-level leaders make life-and-death decisions about their organizations. 19) Its probably a good idea to nd something out about the quality of top-level leaders before investing in a rm. 20) When a company is doing poorly, the rst place one should look to is its leaders. 21) The process by which leaders are selected is extremely important. 24) When the top leaders are good, the organization does well; when the top leaders are bad, the organization does poorly. 25) Theres nothing as critical to the bottom-line performance of a company as the quality of its top-level leaders. 28) Leadership qualities are among the most highly prized personal traits I can think of. 32) No expense should be spared when searching for and selecting a leader. 3) Most things in an organization have very little to do with the decisions and activities of its leaders. 10) When faced with the same situation, even different top-level leaders would end up making the same decisions. 11) Many times, it doesnt matter who is running the show at the top; the fate of an organization is not in the hands of its leaders. 13) You might as well toss a coin when trying to choose a leader. 17) The connection between leadership and overall company performance is often a weak one. 18) Many times, organizational leaders are nothing more than gureheads like the King and Queen of England. 22) So what if the organization is doing well; people who occupy the top-level leadership positions rarely deserve their high salaries. 26) In many cases, candidates for a given leadership position are pretty much interchangeable with one another. 27) The President of the United States can do very little to shape the course of our country. 31) One leader is as good or as bad as the next. 2) The majority of business failures and poor organizational performances are due to factors that are beyond the control of even the best leaders.
1 x x x x x x x x x x x x x x
x x x x x x x
x x x x
45
Downloaded from http://lea.sagepub.com by Tomislav Bunjevac on May 4, 2009
Leadership
3(1) Articles
Item 7) Luck has a lot to do with whether or not business leaders are successful in making their rms protable. 23) In comparison to external forces such as the economy, government regulations, etc., a companys leaders can have only a small impact on a rms performance. 29) Leaders should not be held totally responsible for what happens to a rms performance. 30) There are many factors inuencing an organizations performance that simply cannot be controlled by even the best of leaders.
3 x x
x x
Note. 1 = the inuence of a leader; 2 = the interchangeability of a leader; 3 = the signicance of the inuence of other factors; the numbers before the items refer to the original item numbers as given in Meindl (1998b). Reprinted from Leadership: The Multiple-level Approaches, Vol. 2, James R. Meindl, Appendix: Measures and Assessment for the Romance of Leadership Approach, pp. 199302, 1998, with permission from Elsevier.
Birgit Schyns received her PhD at the University of Leipzig, Germany. She worked as an assistant professor in The Netherlands where she continued her work on leadership, preparedness to change, and self-efcacy. Her research focuses on leadership, especially the perception of leadership and implicit leadership theories. She is currently a reader at the Portsmouth Business School, UK. [email: [email protected]] James R. Meindl (deceased) was the Donald S. Carmichael Professor of organization and human resources in the Department of Organization and Human Resources, in the School of Management, State University of New York at Buffalo, USA. He was the director of the Schools Center for International Leadership. He received his BA at the University of Rochester, his MA and PhD, University of Waterloo, and was a post-doctoral fellow, University of Illinois at Champaign-Urbana. James Meindl authored and/or edited numerous articles, chapters and books on the behavioral aspects of organizing and managing. He is best known for his work on the romance of leadership, a unique perspective on charisma and other popular forms of leadership. Marcel A. Croon is associate professor at the Methodology and Statistics Department of the University of Tilburg (The Netherlands). His research interests are in the development of latent variable models for the analysis of data in sociaal and behavioural research. He also teaches courses in research methodology, factor analysis, and structural equations modelling.
46
Downloaded from http://lea.sagepub.com by Tomislav Bunjevac on May 4, 2009