Working in the Sphere of Politics-Administration Dichotomy: A Study of the Legislative Liaison Agents Belief System with Q-methodology in Taiwan +

Transcription

1 Working in the Sphere of Politics-Administration Dichotomy: A Study of the Legislative Liaison Agents Belief System with Q-methodology in Taiwan + Don-yun Chen * Tun-Wen Sun ** Szu-yu Kuo *** Abstract The politics-administration dichotomy is an important but controversial concept in the filed of public administration. Scholars have been debating it since the inauguration of the field by Woodrow Wilson. However, most literatures on this subject are at best verbal in nature and seldom is there discussing through empirical investigation. This research intends to investigate the dichotomy from the viewpoints of legislative liaisons agents (LLAs) at the Taipei City Government in Taiwan. As the executive-legislative division is one of the critical governing mechanisms which is related to the issue of politics-administration dichotomy, it is an adequate effort to explore the dichotomy by examining LLAs belief system. For this reason, author adopted an inductive research technique called Q-methodology to explore the belief systems of 11 Taipei City Government s LLAs concerning their works. There are three major findings in this research. First, results show that these legislative liaisons perceive two distinct attitudes. Author labeled the two type as Agency-interest oriented LLAs and Professional-service oriented LLAs. Second, concerning managing the executive-legislative relation smoothly, there are four basic elements to follow: trust-building, law-binding, professionalism, and following executive leadership. Finally, this research uncovers two important practical proverbs concerning the fusion of politics/administration for bureaucrats in political arena: political connection is a convenient path of administrative operation, and administrative professionalism is an important political asset. Keywords Politics-administration Dichotomy, Executive-legislative Relations, Legislative Liaison Agents (LLAs), Belief System, Q-methodology + Earlier version of this paper was presented at the International Conference for Governance across Boundaries, Taipei: Taiwan, 2005/7/15. This is a draft, please do not cite without the permission from the authors. * Corresponding author ( 陳敦源 ): Associate Professor, Department of Public Administration, National Chengchi University, 64, Sec. 2, Zhi-nan Rd., Wenshan, Taipei 116, Taiwan, Republic of China. donc@nccu.edu.tw. ** Professor ( 孫同文 ), Department of Public Administration and Public Policy, National Chi-Nan University. *** She ( 郭思禹 ) has a Master degree in Public Policy and Management from the Department of Public Policy and Management, Shih Hsin University in Taipei, Taiwan.

2 Introduction the field of administration is a field of business. It is removed from the hurry and strife of politics. - Woodrow Wilson (1887) Every knows that policy making and administration should be kept distinct. At the same time, everyone knows that policy making and administration cannot be kept distinct and that the distinction itself is difficult to make precise. - March and Olsen (1989: 141) Dichotomy is a term to describe a classification into two opposite parts of subclasses. Since Woodrow Wilson s first publishing The Study of Administration about one century ago, scholars in the field of public administration have being facing the fact that they are working on politics-administration dichotomy where Wilson vacillates between the two poles of thought regarding the separability and inseparability of administration from politics. (Stillman, 1973: 586) 1 Wilson s essay does not offer us a blueprint for a proper relationship between the administrative and political realms in all polities, but it does pinpoint a critical and interesting research realm for future scholars to explore. As a result, the question of how to locate public administration in relation to politics bears critical implications for both the intellectual identity and practical development of public administration. However, few are there efforts to explore this dichotomy in practical world to understand the relationship. In this paper, authors try to explore the relationship between the political and administrative realms by examining a special group of civil servants called legislative liaison agents (LLAs) who works as boundary spanners across legislative (political) and executive (administrative) realms. By using Q methodology, authors try to answer the following question: In LLAs belief system, are they working through the executive-legislative relationship in a political, administrative or mixed way? Or, how political and administrative realms coexist in LLAs working realm? And what this coexistence can teach us about the legislative-executive relationship in specific and the 1 As quoted in Levine, Peters, and Thompson (1990: ). 2

3 politics-administration dichotomy in general? Literature Review Researches on the Dichotomy Since The Study of Public Administration, published in 1887, the politics-administration dichotomy has become one of the founding concept of public administration researches. (Demir and Nyhan, 2008) It is related to other important concept in public administration such as neutrality, responsibility and administrative ethics in a critical way. (Overeem, 2005; Svara, 2006; Yang and Holzer, 2005) Following this Wilsonian progressive mentality, Frank Goodnow restates the dichotomy with a structural-functional tune in his renowned book Politics and Administration. Although there are many different interpretations about Wilson-Goodnow s efforts on reforming government, they do open up an important research path for public administration scholars to explore. (Rabin and Bowman, 1984; Stillman, 1973; Fry, 1998) During the past one century or so, scholars basically discuss the dichotomy in three different ways. First, the dichotomy has become the foundation of many theoretical explorations. For example, in a more direct way, James H. Svara (1998, 2001) argues for a complementary relationship between the two. He stats that (c)omplementarity entails ongoing interaction, reciprocal influence, and mutual deference between elected officials and administrators. (Svara, 2001: 180) In another article, Svara (1998) even calls the dichotomy an aberration in the history of idea in public administration. Beth E. Warner (2001) reinterprets the dichotomy through the lens of John Stuart Mill s theory of bureaucracy within representative government where bureaucratic competence is a necessary condition for a successful participatory polity. Also, Kenneth J. Meier (1997) argues the case for more bureaucracy and less democracy to follow Wilson-Goodnow s progressive spirit in finding a proper balance between political and administrative realms in American democracy. These theoretical statements can also be linked to the Blacksburg Manifesto issued by Gary Wamsley and colleagues who argue for an independent and legitimate role for bureaucracy in a democratic polity. (Wamsley and others, 1990) Within the first route of studying the dichotomy, 3

4 scholars are searching for proper arguments of the relationship between politics and administration. The method they use to do the job is usually argumentative rather than empirical. The study of council-manager government (CMG) in the U. S. is another area with important connection to the dichotomy. The idea of CMG was born in the progressive era and its institutional realization mushroomed after World War Two in the U. S. In other words, local government ruled by an appointed professional administrator and elected council inherits the Wilson-Goodnow s progressive spirit. As a result, understanding of the continuity and change of the CMG in the U. S. is a way to understand (1) whether the dichotomy is still relevant nowadays; (2) what is the meaning of the continuity and change of the CMG to the contemporary researchers who are interested in the dichotomy. For Montjoy and Watson (1995), although the dichotomy can not serve as practical guidance for running the CMG, it can be a professional standard for drawing a clear line for institutional ban on particularism from the representative government. Svara (1999; 2009) has found that the line between the council and the city manager is increasingly overlapping. It means that both sides of the dichotomy are getting more and more entangled with each other. Research methods used in this route mostly are case study or multiple case studies. The last research route on the dichotomy is explored by Aberbach, Putnam, and Rockman (1981) by empirically comparing the belief system in bureaucrats and politicians in Western democracies (Aberbach, Chesuey, and Rockman, 1975; Aberbach and Rockman, 1976, 1988; Dogan, 1976). In general, they find that bureaucrats and politicians play different roles, which bring distinctive perspectives and competencies to policy making. The rationale for their research, as Putnam puts in the following words, is to bring up bureaucrats influences to our attention in the democratic policy making process. He states that: Granted that the permanent civil service has a strategic role in the policy-making process, the central question for political scientists is: How responsive is the bureaucracy to changing social needs and political demands, and why? (Putnam, 1976: 87) To conclude the researches in the field, it is obvious that we need more empirical works to explore not only the existence (or non-existence) of the 4

5 dichotomy, but how the political and administrative realms coexist with each other in real world context. For example, Sung-Don Hwang (1999) utilizes statistical tools such as Q-methodology to explore the compatibility between the two realms by way of surveying bureaucrats. In an article on discussion the dichotomy, John Nalbandian (1994) correctly points out that the minds of political practitioners rather than their behaviors should be the place to look for answers. He uses the CMG as a case and pinpoints the role that the chief administrative officer played as a translator to connected the elected officials and professional staffs. In this paper, we follow this empirical route to explore the dichotomy. By targeting the LLAs in Taipie City Government, we use Q-methodology to categoricalize their experience and thinking about their works to understand the dichotomy. Legislative-executive Relationship and the Dichotomy From an institutional viewpoint, modern democracies are all in some way a system of check and balance, with which the politics-administration dichotomy is conceptually intertwined. For example, the legislative-executive separation is one important aspect of the system of check and balance. The separation of the two branches is a form of politics-administration dichotomy where the legislature represents the political and the executive (mostly the bureaucracy) represents the administrative side on the constitutional power map. Cunningham and Olshfski (1986:104)once states that (t)he relationship between legislature and administration highlights the tension between the requirement of separation of powers and the necessity for sharing policy responsibility. However, within the executive branch, there is also a dichotomous relationship between elected officials (together with political appointees) and career civil servants. In a narrower sense, the civil servants, who are at the same time accountable to both legislators and executive leaders, are the precise representation of administrative apparatus in the dichotomous relationship. As a result, there are two aspects of the politics-administration dichotomy can be realized in the legislative-executive relationship; one is the legislative-bureaucracy relationship and the other is the relationship between executive leaders and their bureaucratic subordinates. It is the two-clients 5

6 situation described by Holtzman (1970: 77) on studying the White House legislative liaison agents in the U. S. In our paper, these two aspects of the dichotomy can be empirically comprehended by studying legislative liaison agents (LLAs) in the Taipei City Government (TCG). Our argument is as follows. Since LLAs are mostly permanent civil servants and there are asked to work in political realm, we can know that whether the administrative and political realms are mixed or separated, we can understand it from the mind of the LLAs who are the boundary-spanners between administrative and political realms. If the two realms are mixed, the mixture can best be seen in the LLAs thinking about their works. Or, if the two realms are coexisted in real world, the coexistence can best be understood in the LLAs belief system concerning the dichotomy. In the next section, we will do the background introduction of the Legislative Liaison Mechanism in Taipei City Government. Legislative Liaison Mechanism in Taipei City Government Government is an organization, the organization interacts with and is dependent upon it environment. Boundary spanner individual is a representative to the environment and the environment to the organization whilst serving as a critical path by which salient information can be gathered and passed onto organizational managers. (Katz and Kahn 1967) Strategically speaking, the boundary spanner is constantly engaged in attempting to shape the environment so that the contingencies faced by the organization can be more effectively managed. Williams (2002) sees the boundary spanner in governmental organization as a policy entrepreneur who owns several important personal attributes to work for the organization s best interest. In certain way, LLA is the illustrative example of the boundary spanner in governmental organization. The boundary they are spanning on is the line between the administrative and political realms drawn by the theory of politics-administration dichotomy. In 1978, the executive of the Taipei City Government issued a decree on Guidelines for Strengthening Legislative Liaison Works. In 1990, the executive issued an internal decree of the Guidelines for Appointment of Departmental Legislative Liaison to manage the legislative liaison mechanism as a whole in the 6

7 Taipei City Government. The organization of the legislative liaison mechanism is as follows. (1) Every department or sub-department can assign one liaison as a representative to the city council. If there are more than five sub-departments within a department, that department can assign one more liaison. Or, if there is a special need for sending a representative to the city council, approval should be granted by the mayor. For example, the Taipei City Bank and the Metro-Rapid-Transit Company are sending their own liaison to the city council. (2) All legislative liaison agents are in an ad-hoc position. They still belong to their original departments for performance evaluation or compensation. The selection process of LLA is a highly decentralized and personalized process involving around the department heads. The total number of LLA is around seventy to eighty persons. (3) There is a chief liaison officer (the vice city mayor usually takes the position) and three vice liaison officers at the top of the mechanism. Because the diversity of the work contents, there is not much command-and-control activities between the chief officer and departmental liaison agents, unless there raises an issue required coordination among liaison agents, such as natural disaster. (4) It is interesting to know that although the legislative liaison mechanism is ad hoc at best in the city government organization, the city council usually passes its budget without any delay. We can see that the city council needs the mechanism to exist as much as the executive needs to establish it. (5) Lastly, according to Chen and Kuo (2003), LLAs usually engage in four major works on the job. First, they have to response to information needs from the council members. Second, they need to utilize their expertise to help council members doing caseworks for their constituents. Third, they have to lobby for policies proposed by their departments. Forth, they need to detect the possible 7

8 questions council members will ask to their departmental heads during the interpellating sessions. In this paper, we use the LLAs in Taipei City Government as the target population to understand their boundary spanning role in the legislative-executive relationships. In the next section, we will introduce the Q-methodology and justify its use in this paper. Methodology and Data Q Methodology in Public Administration William Stephenson invented Q-methodology in It is a research technique to examine human subjectivity. It has an idea of manipulating data matrix which is opposed to traditional survey method. Stephenson (1935) distinguished between the traditional R-methodology referred to a selected population of n individuals each of whom has been measured in m tests, and that Q-methodology referred to a population of n different tests (or essays, picture, traits or other measurable material), each of which is measured or scaled by m individuals. (Brown, 1997) In Q-methodology, respondent is the test and statements are like respondents in survey method. Q-methodology uses a series of items to test respondents. The perspective of a respondent is the subject of this test. Respondents are asked to sort statements about a topic according to how strongly they agree or disagree with each statement. By ranking the statements, respondents express their perspectives about the topic. The result of the ranking is called Q-sort. (Brewer, Selden & FacerⅡ, 2000) Then researcher can put all Q-sorts into the statistic software to correlate and factor-analyzing them. 2 Factors indicate clusters of respondents who have ranked the statements in essentially the same distribution. According to the result, we can identify different types of respondents idea about certain issue. In other words, respondents in the same type share the attitudes or perspectives on the issue under quary. (Durning & Osuna, 1994) 2 The software we are using is called PQMethod which is free of charge and can be downloaded from the following internet address: 8

9 Since Stephenson first used Q-methodology in studying the change of a patient s personality in 1935, Q-methodology has been applied to many fields, such as psychology, sociology and political science. In the field of public administration, Q-methodology has been applied to research the topics about policymaking, the roles of policy analysts and government officials, or to reconstruct some vague concepts as democratic theory (Dryzek & Berejikian, 1993; Durning & Osuna, 1994; Steelman & Maguire, 1998; Selden, Brewer & Brudney, 1999). Q-methodology is used to deal with the complex and multidimensional questions. It is useful to clarify an issue where there can be as many perspectives as there are stakeholders. (Donner, 2001) It is also used as a method to categoricalize individual s perspective, value, attitude, or ethic. The politics-administration dichotomy is a vague concept to grasp in the filed of public administration. However, the attitudes of LLA systematically reveal the dichotomy in the real world. Therefore, Q-methodology is a good technique to understand the dichotomy by studying the belief system of LLAs. The Q-statements and P-Sample In order to collect statements to formulate the Q-statements in this research, we select 36 statements from a pool of documents concerning LLA s work, including books, newspapers articles, journal articles, and an one-hundred-thousand-words interviewing records collected by Chen and Kuo (2003). 3 Because the subject in Q-methodology is the viewpoints of respondents, the population of this research means various opinions about the dichotomy via the works of LLAs. In order to assure the completeness of the Q-statements, we structure these statements from the two dimensions below: (1) Three dimensions of the public-administration relationship: responsiveness, responsibility, and accountability (Levine, Peters & Thompson, 1990). (2) Four major activities for LLAs: offering written materials, helping legislators to resolve their constituent caseworks, lobbying on department s budgets and acts, and detecting the contents of legislators interpellation (Chen and Kuo, 2003). 3 There are totally thirty persons interviewed by Chen and Kuo (2003). Only a half of them are LLAs. Others are those who are related to LLA s works, such as departmental head, city council members and their assistances, partisan coordinators in the council, and media reporters. 9

10 The interplay of the above two dimensions gives us an 3 4 matrix outlined in Table One. Using the factorial design, we collected 36 Q-statements according to the 12 cells. (See appendix A) In others words, all possible categories connecting to LLA s works and the dichotomy is revealed in the structure. Table One: The Structure of Q-Statements Major Activities Dimensions (1) Responsiveness (A) Offering Information A1(3) 16,18,22 (B) Constituent Caseworks B1(5) 9,12,17,31,35 (C) Lobbying C1(3) 5,33,36 (D) Prepare Interpellation D1(1) 2 (2) Responsibility A2(3) 1,8,26 B2(3) 7,27,34 C2(3) 3,13,21 D2(3) 11,23,28 (3) Accountability Total A3(3) 4,10,20 B3(1) 32 C3(3) 14,25,30 D3(5) 6,15,19,24,29 Total (12) (12) (12) (36) Source: This research Unlike the traditional survey research, Q-methodology does not require a large number of respondents. We don t have to sample the LLA population systematically, but we still have to encounter as different LLA as possible according to his or her background. In the beginning, we interview three LLAs who were interviewed by Chen and Kuo (2003). Then, we ask them to introduce someone who has as different background and personality from them as possible. As the result of rolling snowballs from June to August in 2004, we collect a P-sample included eleven LLAs from the Taipei City Government. Six of them are female. Two of them are political appointees and the rest are all permanent civil servants. Mostly are them owns at least university degree. Their average age is forty-seven years old. They on average already spend about twenty-one years in public service and five years in working as LLAs. (9) (9) (9) (9) 10

11 The Process of Using Q-methodology To gather the Q-sorts from each respondent, we have to have the instrument with four components: chips, guide bar, answer sheet, and instructions. (Donner, 2001) We typed one Q-statement on each chip. The 36 chips are what respondents actually sort. We print the answer sheet (see Table Two) on a whiteboard. The range of agreement is from +4 (the mostly agreed) to -4 (the mostly disagreed). Finally, we type a short instruction for respondents to follow to finish the sorting. The four-step-procedure for Q-sort is as follows: (1) Respondent is asked to review the 36 statements: Q-methodology requires respondents to evaluate each item in relation to the other items. The placement of one item will affect the placement of the other item. (Brewer, Selden & FacerⅡ, 2000) We prepare an uncluttered table, so that respondents can review and compare each statement at the same time. (2) Preliminarily sorting the 36 statements into three piles: agree, disagree, and neutral. (3) Ranking 36 statements in a fixed distribution: Ranking statements according to how strongly you agree or disagree with them. Respondents are asked to rank statements according to a fixed distribution (see Table Two). Repeat three times until filling all the blanks on the answer sheet. (4) Checking the placement: Finally, each respondent is asked to review all statements placing on the board for the last time. Look over them to see whether there are any statements the respondent would like to move. If not, fill in the answer sheet by writing the number of each statement in the appropriate block on the answer sheet. Table Two: Q-methodology Answer Sheet Mostly disagreed Mostly agreed

12 Source: This Research. After the sorting is completed, we conduct a short interview with the respondent. We asked for his (or her) opinion about the dichotomy, and how and why he or she assigns certain statement to one place rather others. Also we collect information on the respondent s age, educational background, and years of experience as a civil servant as well as a legislative liaison. Data Analysis In this research, a statistical package-pqmethod-is used for calculating the collected data. We put 11 Q-sorts into PQMethod and then conduct a factor analysis of the data with principal components method. After doing varimax rotation, we find that there are two factors identified in the procedure. According to the factor loadings of each Q-sorts for the two factors (see Table Three), 5 people loaded on factor A and 6 people loaded on factor B. It means that respondents who are placed significantly on a factor hold similar perspectives on their LLA in specific and on the dichotomy in general. We examine the Q-sorts for each factor focusing on the statements where individual strongly agreed or disagreed (see Q-factor scores at appendix A). We also compare the factor with the other factor. In conclusion, we find that there are two different types of LLAs. We give each type of LLA a name which is Agency-interest Oriented LLA and Professional-service Oriented LLA respectively. We then discuss them one by one in the following sections. Type A: Agency-Interest Oriented LLAs People belong to the type A think that a LLA should try his best to obtain the support on department s budgets and acts from legislators (statement 30). So the will try to maintain a warm relationship with city councilmen or councilwomen for the purpose of protecting agency s interest. They believe that the relationship is helpful 12

13 for passing their department s budgets and policies (statement 24, 33). Another important job for Type A LLAs is to protect their departmental leaders. They understand the political perspectives of their leaders (statement 14). And they will do their best to detect the contents of legislators interpellation in order to avoid direct conflicts between the legislator and their leaders (statement 2, 6). Furthermore, if they can, they will not offer insider s information to legislators for keeping legislators away from understanding the various activities in their departments (statement 8, 21). They will be very selective on giving any written materials to legislators in order to eliminate any possible harm to their leaders or agencies (statement 20, 26). As a result, the agency-interest oriented LLAs serve as gatekeepers to protect their agency interest as well as their departmental leaders. Because they don t believe that the representative institution can truly represents the public interest (statement 9), they low attention on responsiveness. Table Three: Respondent s Factor Loadings and Demographic Variables Code Type A Type B Sex Age Seniority a Experience b Education Status R F university Civil Servant R F college Political Appointee c R F master Civil Servant R M master Civil Servant R M university Civil Servant R M university Civil Servant R F university Civil Servant R M university Civil Servant R M master Civil Servant R F master Political Appointee c R F university Civil Servant a It is the years of joining the civil service. 13

14 b It is the years of joining the legislative liaison. c Political appointees are those LLAs who come with their departmental heads and do not belong to the civil servants. Type B: Professional-Service Oriented LLAs The LLAs loading on type B emphasize their professionalism. They think that their ability to give council members satisfactory services is from their long-term experience and professional knowledge (statement 13, 27, 17, 29). They need not to break rules to service the council members (statement 7, 31, 22, 8, 32). And it shows that they fulfill their bureaucratic responsibility through professionalism. People in type B think that a LLA should gain the trust and support from his department head in order to do the job effectively. He (or she) has to know what is on his (or her) leader s mind (statement 14). Also, he needs directing power to call for help from the staffs in his department (statement 34). Leader should not blame on the LLA on the face of scolding from council members (statement 15, 10). One major difference between the Type A and Type B LLAs is that the Type B LLAs are more willingly to provide services to council members in particular and to citizens in general. They think that government should be as transparent as possible via the Internet (statement 18). Basically, they are more responsive than the Type A LLAs. Another difference is that comparing with the Type A, Type B LLAs are serving the council members via their professional abilities rather than personal relationships. The finding counteracts our original thinking that for a bureaucrat to be politically active or responsive, he should be open for negotiation in any case and work under the table regardless of his professional background. However, this finding confirms the statement made by Kearney and Sinha (1988) that professionalism is the key to establish a responsive bureaucracy. Table Four: A Comparison of Type A and B Dimensions Type A Type B Protecting agency Protecting head Major Statements Building personal relationships Distrust the council Professionalism Problem-solving for council members Need trust from head Responsiveness Common Statement Obeying the law and following the head s will Responsiveness Low Median Responsibility Median High Accountability High Median 14

15 Dimensions Type A Type B Base for Interaction Personal connection Professionalism Major Concerns Agency s Interest Responsive to Council Role Playing Gatekeepers Representatives Source: This Research. Basically, the Type A stand on a firm ground of protecting the agency s interest. The purpose of maintain a relationship with councilmen and councilwomen are promoting departmental budget and policies. The Type B think that the professional competence is the key for doing the LLA job well. They also need the trust from their departmental leaders to help the council members to solve their problems. Both type LLAs agree that to obey the laws and follow departmental leader s will are two basic requirements for serving as LLAs. We summarize our finding in the following Table Four. Summary and Conclusions In this paper, we begin with a question concerning the politics-administration dichotomy in the real world. We choose to examine the belief system of LLAs in the Taipei City Government by utilizing Q-methodology to explore the systematic variations and commonalities. Statistically, we find that there are two types of LLAs. We label them as agency-interest oriented LLAs and professional-service oriented LLAs respectively. To summarize our findings, we can conclude from two parts, one is concerning the legislative-executive in particular and another one is concerning the politics-administration dichotomy in general. In the first place, from our analysis, we can distract four major elements for doing LLA s job in the real world. They are (1) trust-building, (2) law-binding, (3) professionalism, and (4) following executive leadership. Different type of LLA might emphasis different elements to do their job but there are common elements across the types. As a boundary spanning across political and administrative realms, LLAs basically is still binding by the executive branch more than the legislative branch. In the second place, we can see from above findings that the connecting role a LLA played is a translator between the executive and legislative branches where information transmitting and communicating activities are the nature of the 15

16 job. However, they are still loyal to their sending organizations for the simple reason of resource-dependence. Under the theory of politics-administration dichotomy, this dependence makes the LLAs difference in attitude only a difference in degree within the administrative realm rather than a difference across the political and administrative realms. Finally, this research uncovers two important practical proverbs concerning the boundary spanning across the politics-administration dichotomy for bureaucrats in political arena. They are: Political connection is a convenient path of administrative operation, and Administrative professionalism is an important political asset. 16

17 Reference Aberbach, J. D., J. D. Chesney and B. A. Rockman. (1975) Exploring Elite Political Attitudes: Some Methodological Lessons. Political Methodology, pp Aberbach, J. D., R. D. Putnam and B. A. Rockman. (1981) Bureaucrats and Politicians in Western Democracies. Cambridge, MA: Harvard University Press. Aberbach, J. D. and B. A. Rockman. (1976) Clashing Beliefs Within the Executive Branch: The Nixon Administration Bureaucracy. The American Political Science Review 70: (1988) Mandates or Mandarins? Control and Discretion in the Modern Administrative State. Public Administration Review 48(2): Brewer, G. A., S. C. Selden and R. L. FacerⅡ. (2000). Individual Conceptions of Public Service Motivation. Public Administration Review, 60(3): Brown, S. R. (1997). The History and Principles of Q Methodology in Psychology and the Social Science. Paper presented at the conference on Celebration of the Life and Work of William Stephenson ( ). University of Durham, December, pp Chen, Don-yun and Cheng-wei Kuo. (2003) Information Transactions in Democratic Governance: A Case Study on Taipei City Government s Legislative Liaison Mechanism. Paper presented at the Second Annual Meeting of the TASPAA, Taipei, Taiwan. (in Chinese) Cunningham, R. and D. Olshfski. (1986) Interpreting State Administrator-Legislator Relationships. The Western Political Quarterly 39(1): Demir, Tansu and Ronald C. Nyhan. (2008) The Politics-Administration Dichotomy: An Empirical Search for Correspondence between Theory and Practice. Public Administration Review, 68(1): Dogan, M. (1976) The Mandarins of Western Europe: The Political Role of Top Civil Servants. John Wiley & Sons, Inc. Donner, J. C. (2001). Using Q-Sorts in Participatory Processes: An Introduction 17

18 to the Methodology. Social Development Papers: Social Analysis Selected Tools and Techniques, 36: Durning, D. and W. Osuna. (1994). Policy Analysts Roles and Value Orientation: An Empirical Investigation Using Q Methodology. Journal of Policy Analysis and Management, 13(4): Fry, B. R. (1989) Mastering Public Administration: From Max Weber to Dwight Waldo. Chatham, NJ: Chatham House Publishers. Goodnow, F. (1990) Politics and Administration. Macmillan. Holtzman, A. (1970) Legislative Liaison: Executive Leadership in Congress. Chicago: Rand McNally & Company. Hwang, Sung-Don. (1999) Bureaucracy vs. Democracy in the Minds of Bureaucrats: To What Extent Are These Ideologies Compatible with One Another? Grove/Atlantic. Katz, D and Kahn, R.L (1966) The Social Psychology of Organizations. New York: John Wiley and Sons Kearney, R. C. and C. Sinha. (1988) Professionalism and Bureaucratic Responsiveness: Conflict or Compatibility? Public Administration Review 48(1): Levine, C. H., B. G. Peters, and F. J. Thompson. (1990) Public Administration: Challenges, Choices, Consequences. Glenview, IL: Scott, Foresman/Little, Brown Higher Education. March, J. G. and J. P. Olsen. (1989) Rediscovering Institutions: The organizational basis of Politics. New York: Free Press. Meier, K. J. (1997) Bureaucracy and Democracy: The Case for More Bureaucracy and Less Democracy. Public Administration Review 57(3): Montjoy, R. S. and D. J. Watson. (1995) A Case for Reinterpreted Dichotomy of Politics and Administration as a Professional Standard in Council-Manger Government. Public Administration Review 55(3): Nalbandian, J. (1994) Reflections of a Pracademic on the Logic of Politics and Administration. Public Administration Review 54(6): Overeem, Patrick. (2005) The Value of the Dichotomy: Politics, Administration, and the Political Neutrality of Administrators. Administrative Theory & Praxis, 27(2):

19 Putnam, R. D. (1976) The Political Attitudes of Senior Civil Servants in Britain, Germany, and Italy. In The Mandarins of Western Europe: The Political Role of Top Civil Servants, M. Dogan, ed. John Wiley & Sons, Inc. Rabin, J. and J. S. Bowman. (1984) Politics and Administration: Woodrow Wilson and American Public Administration (Public Administration and Public Policy /22). New York: Marcel Dekker, Inc. Selden, S. C., G. A. Brewer, and J. L. Brudney. (1999). Reconciling Competing Values in Public Administration: Understanding the Administrative Role Concept. Administration and Society, 31(2): Steelman, T. A., and L. A. Maguire. (1998). Understanding Participant Perspectives: Q-methodology in National Forest Management. Journal of Policy Analysis and Management, 18(3): Stephenson, W. (1935). Technique of Factor Analysis. Nature, 136: 297. Stillman, R. J., II. (1973) Woodrow Wilson and The Study of Administration : A New Look at an Old Essay. American Political Science Review 67: Sun, T-W and J. J. Gargan. (1999). Public Administration Scholar-Practitioner Differences: A Q Study of Theory-Practice Connection in Taiwan. In Handbook of Comparative Public Administration in the Asia-Pacific Basin, edited by H-K Wong, H. S. Chan. New York: Marcel Dekker, pp Svara, J. H. (1998) The Politics-Administration Dichotomy Model as Aberration. Public Administration Review 58(1): (1999) The Shifting Boundary between Elected Officials and City Managers in Large Council-Manager Cities. Public Administration Review, 59(1): (2001) The Myth of the Dichotomy: Complementarity of Politics and Administration in the Past and Future of Public Administration. Public Administration Review 61(2): (2006) Complexity in Political-Administrative Relations and the Limits of the Dichotomy Concept Administrative Theory & Praxis, 28(1): (2009) The Facilitative Leader in City Hall: Reexamining the Scope and Contribution. New York: CRC Press. Yang, Kaifeng and Marc Holzer. (2005) Re-Approaching the Politics/ Administration Dichotomy and Its Impact on Administrative Ethics Public 19

20 Integrity, 7(2): Warmsley, G. L. et al. (1990) Refounding Public Administration. Newbury Park, CA: Sage Publications. Williams, P. (2002) The Competent Boundary Spanner. Public Administration 80(1): Wilson, W. (1887) The Study of Administration. Political Science Quarterly, 2(2): Administrative Theory & Praxis Issue: Volume 28, Number 1 / 2006 Pages: URL: Linking Options Forum: James H. Svara A1 20

21 Appendix A-Q statements and Factor scores No. Statements Type A Type B 16 A B C D1 2 I should help legislators to acquire proper 0 1 information quickly. (-0.140)(0.526) Administrative agencies should upload as much 0 2 information as possible to the Internet. (0.360) (1.047) When several legislators ask for the material at the -3-4 same time, I would offer to one who has closer (-2.138)(-2.258) friendship with me. Providing service for legislators is equal to -3-1 responding to the demand of the citizens. (-1.971)(-0.247) If an administrative agency could directly serve the -1 0 citizens well, it could reduce helping legislators to do (-0.551)(0.247) their constituent caseworks. I should help legislators to get in touch with the 1 3 correct unit of our agency quickly. (0.671) (1.066) I should not refuse the demands from legislators right 0-2 away, even if they look unattainable. (-0.175)(-0.804) When the request of legislators is the problem at the -2 0 gray area of law, I would help them as possible as I (-0.936)(-0.096) can in order to maintain the relationship. I must report the demands of legislators to my 1 0 agency director in a complete honest fashion. (0.389) (0.318) Keeping in touch with legislators offices more 2-1 regularly could help my agency to pass the act and (1.230) (-0.382) budget more easily. It is easier to seek supports from the legislators of the -1-1 ruling party than the legislators from the opposition (-0.509)(-0.507) party. Keeping in touch with legislators offices more 2 0 regularly could help me to know the direction of (1.096) (0.314) interpellation in advance. Responsiveness I will carefully review all written materials before 7 0 transferring to legislators. (-0.085)(-0.001) A2 8 In the name of public interest, I will offer -4-3 unpublished inside information to legislators. (-2.150)(-1.989) 26 I will transfer any information to legislators except

22 No. Statements Type A Type B 7 B C D those being classified as secret in the agency. (-0.717)(-0.407) I help legislators to resolve their constituent 1 2 casework within the realm of legal boundary. (0.827) (1.034) The legislative liaison officer must be a professional 1 3 senior in the agency. (0.811) (1.526) I do not give direct orders to the staffs in my agency Actually, I would ask my agency director to give the (-0.596)(-1.216) orders. I am usually proactive on problem-solving, so that I 0 1 can explain the reasons behind the problems to (0.319) (0.444) legislators on time and reduce misunderstandings. I usually communicate with legislators based on my 1 4 professional judgments. (0.463) (1.817) I usually let legislators know how things are going in -3 0 our agency in order to earn their understandings. (-1.337)(0.262) I can deduce the possible direction of interpellation 0 2 from the written materials that legislators asked for. (0.271) (0.762) When a legislator pays attention to a certain issue, I -1 1 could quickly determine what his (or her) purpose is. (-0.461)(0.352) I would pay attention to news all the time, because 1 1 the mass media is a critical rout to ascertain the (0.815) (0.599) questions of legislators interpellation. Responsibility A B C I would send all written materials to my leader to 0 1 check before offering to legislators. (0.004) (0.428) If my leader is scolded in the legislator s -2-3 interpellation, he will blame on me. (-1.020)(-1.678) I should sort through or clean up the written materials offered to the legislators in order to reduce 2-1 the possible harm to my agency director or the (0.971) (-0.445) agency. I usually take special care to those caseworks -1-2 which are handed over from the legislators directly to (-0.509)(-1.143) my agency director. In my work, I usually take the political perspective of 3 3 my agency director into consideration. (1.414) (1.765) In facing sensitive problems, I usually seek solutions -2 0 with legislators under table. (-0.755)(0.253) Lobbying legislators to support our agency s budget 4-2 is the most important mission of the legislative (1.529) (-0.937) 22

23 No. Statements Type A Type B 6 15 D liaison officers. To prepare the interpellation for my agency director, 3 0 I will let him or her know which written material had (1.265) (0.138) been offered to legislators. In the eyes of my agency director, a good liaison officer is the one who can help his or her director -1-3 smoothly passing through the legislator s (-0.485)(-1.652) interpellation. My agency director usually asks me to acquire the 0-1 interpellant questions from legislators in advance. (-0.191)(-0.356) I would introduce my agency director to legislators at 3 1 the right moment in order to keep a constructive (1.389) (0.509) relationship between them. Letting my agency director to know in advance the 2 2 possible problems of legislators interested in is the (0.880) (0.758) effective way to solve problems. Accountability