can be easily retrieved from a drawer like a pencil, a calculator or a written law. However ...... street-level bureaucrat or call-center receptionist. Therefore, this ...
VRIJE UNIVERSITEIT
VALUE SOLIDITY Differences, Similarities and Conflicts Between the Organizational Values of Government and Business
ACADEMISCH PROEFSCHRIFT ter verkrijging van de graad Doctor aan de Vrije Universiteit Amsterdam, op gezag van de rector magnificus prof.dr. L.M. Bouter, in het openbaar te verdedigen ten overstaan van de promotiecommissie van de faculteit der Sociale Wetenschappen op donderdag 19 juni 2008 om 15.45 uur in de aula van de universiteit, De Boelelaan 1105
door Zeger van der Wal geboren te Amsterdam
“The reason a writer writes a book is to forget a book and the reason a reader reads one is to remember it.” Tom Wolfe
Preface and Acknowledgements In many ways, writing a preface to a dissertation resembles the process of writing the dissertation itself. You have no idea where to start, what to include in which order (and perhaps even more important, what not to include) and when to finally end. And, just as with the actual dissertation, only after the last letter has been written down, it becomes clear what it was you wanted to make clear, making that you want to start the whole process all over again and this time, write a really good book. Yes, I know it is a terrible cliché, but bringing a Ph.D. project to an end is often compared to making a long and sometimes burdensome journey. But to be honest, in many ways it didn’t feel that long and certainly was not burdensome, except perhaps for the last few months, or as we say in Dutch, de laatste loodjes, yet another terrible cliché. That is not to say that a lot of hard work wasn’t put into this book, or that there weren’t times when I wanted to run away to start a beach resort on a tropical island, or wondered why I hadn’t chosen a career as a DJ or a drummer in a rock band… But seriously, most of the times the journey was pleasant, challenging and self-fulfilling, and that has had a lot to do with my main tour guides Leo Huberts and Gjalt de Graaf. Guiding and supervising me in “optimally bounded freedom,” always believing in me while being very – yet constructively – critical, they have made that I experienced this journey to be like an intensive holiday rather than a case of “doing time” in between the walls of academia, which is how outsiders often seem to perceive the process of obtaining a Ph.D. I know for sure that in the future, no matter whether the three of us will physically be in the same building, we will continue to work fruitfully together, pursuing academic excellence domestically, transatlantically and globally. Next to my main tour guides I wish to thank a number of people that have been my traveling partners during the last years and have been of influence on my academic thinking in general and probably will continue to be of influence for a long time. First of all, I acknowledge my direct colleagues at the research group “Integrity of Governance,” and my other colleagues at our department, for their inspiration, comradeship and in a number of cases the friendship that has grown over the years. You know who you are. I thank all the different “office roomies” that I have had since those early days in the autumn of 2003 for sustaining me and my black sense of humor. This goes in particular for Kim van Nieuwaal. I will never forget the last six months in N-323 in which we both had to finish our books, often staying at the metropolitan building until 23 PM, making crazy jokes and having unhealthy evening meals together, keeping each v
other going. Many thanks for being my partner in crime. In the national and international arena, many people have not only been of influence over the years or became co-authors, reviewers or editors (and thereby improvers) of my work, but also appeared to be enjoyable company during dinners and drinks – sometimes many – on conferences in places all over the world. There are too many of you to mention here but you know who you are. I hope you will continue to work with me and improve my scientific work. Second, I would like to thank a number of people that made this project possible. Mirjam Vermeulen and Marcel Haarhuis, thanks for writing both your master theses on the subject of organizational values, being a prelude to as well as the start of a framework for this study. I thank Robbert Braak, Jacques Mulders and Mirko Noordegraaf for enabling me to invite the members of the Senior Civil Service (ABD) to participate in this study, and Willem F. Hollander, Kitty van Mierlo and Robert de Koning, to do so for the Dutch Centre of Executive and Non-executive Directors (NCD). I would also like to acknowledge Henk Bakkerode from the Ministry of Health, Welfare and Sports, Boudewijn Peters from the Ministry of Finance, Gerben Biermann from the Provincial Government of Noord-Holland, Toon Molleman from DJI and Gert Siemons from Organon for assisting in arranging interview meetings with respondents from their organizations and of course all the other participants in this study. Julia Angelica (Dr. J), thanks for editing and greatly improving my texts and for the interesting and sometimes hilarious e-mail conversations in the middle of the night. And last but surely not least, Frank Anechiarico, for letting me stay in your apartment in Washington D.C. to write my first and my last chapters. The months I spend there were rewarding, effective and above all, very enjoyable. Without all of you this book wouldn’t be there as it is there now. Third, on a more personal note, I would like to thank my parents, my brother and sister, all my other family members and all the people that matter. Mar, thanks for keeping me sharp and for enjoying life together for over a decade. Now go finish your own doctoral journey, do it well and with excellence (of course you will)! All my friends, thanks for supporting me, being critical of me, keeping me sharp, partying and having fun together, producing music that I like and that keeps me going (hip hop, minimal and marnimal), and sustaining me, or more concisely: for being there and being who you are! To conclude, I would like to acknowledge my grandma, Oma van der Wal, you are such a strong individual and I am very happy that you are still here and able to hold this book in your hands. Unfortunately, this does not go for my late grandfather, Zeger van der Wal sr., who was an example of personal value solidity and personification of everything that is so beautiful about what we in the Netherlands call the “polder.” I am proud to live forth his name and I dedicate this book to him. Zeger van der Wal, Washington D.C./Amsterdam December 17, 2007 vi
CONTENTS 1
The Morals of Government and Business
1
1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9
Introduction: Two conflicting moral syndromes NPM: Business values in government organizations? CSR: Government values in business organizations? Value intermixing and moral and functional problems Research aims and research questions Central concepts defined Studying what is valued most Relevance Outline
1 3 5 6 8 9 11 13 14
Part I: Theory and Methodology 2
Conceptualizing the Organizational Value System
19
2.1 2.2 2.3 2.4 2.5 2.6 2.7 2.8
The V word: Widely used but seldom specified Conceptual, ontological and epistemological aspects of values Organizational values: A definition The significance of ‘publicness’ Origin and evolution of the organizational value system Organizational values, culture, and decision making The lack of comparative empirical studies Which values belong to which sector?
19 20 22 23 27 28 31 32
3
A Combined Methodological Approach
37
3.1
A combination of methodological perspectives to research values Clashes and contrasts between the quantitative and the qualitative worldview Multiple purposes of a combined approach in this study Three phases of data collection 3.4.1 Phase I: Desk research, expert interviews and a content analysis on values 3.4.2 Phase II: Self-administered mail survey 3.4.3 Phase III: Face-to-face in-depth structured interviews Data-analysis and reporting
37
3.2 3.3 3.4
3.5
vii
38 40 41 41 43 47 50
Part II: Empirical Results 4
A Public-Private Value Panorama
55
4.1 4.2 4.3 4.4 4.5
Organizational values in administrative ethics Organizational values in business ethics Constructing a public-private sector panorama A set of values as a survey research tool Value intermixing, and integrity and decision-making problems
55 59 62 65 66
5
Crucial Values in Public Sector Organizations
69
5.1
5.6
What is valued more and less in government 5.1.1 Relations and conflicts between public sector values What is ultimately valued most in government 5.2.1 What should be valued most in government Different types of decisions, different types of values? 5.3.1 Internal organizational decisions 5.3.2 Political/policy-related decisions 5.3.3 External organizational decisions 5.3.4 Budgetary/financial decisions 5.3.5 Different decisions sometimes imply different values How and when specific values are important 5.4.1 Responsiveness 5.4.2 Transparency and Accountability 5.4.3 Lawfulness 5.4.4 Consistency, Reliability and Incorruptibility 5.4.5 Efficiency and Effectiveness Complementarity, conflict, and context-dependency 5.5.1 Complementarity between public sector values 5.5.2 Conflicts between public sector values values 5.5.3 Context-dependency of public sector values Value intermixing, and moral and functional problems
69 72 73 74 75 77 77 78 78 78 80 80 82 85 89 91 94 95 95 97 98
6
Crucial Values in Private Sector Organizations
103
6.1
What is valued more and less in business 6.1.1 Conflicts and relations between business sector values What is ultimately valued most in business 6.2.1 What should be valued most in business Different types of decisions, different types of values? 6.3.1 Internal organizational decisions 6.3.2 Strategic/market-oriented decisions 6.3.3 External organizational decisions
103 105 106 108 108 110 110 111
5.2 5.3
5.4
5.5
6.2 6.3
viii
6.6
6.3.4 Financial/investment decisions How and when specific values are important 6.4.1 Responsiveness 6.4.2 Transparency and Accountability 6.4.3 Lawfulness 6.4.4 Consistency, Reliability and Incorruptibility 6.4.5 Efficiency and Effectiveness Complementarity, conflict, and context-dependency 6.5.1 Complementarity between values 6.5.2 Conflicts between values 6.5.3 Context-dependency of values Value intermixing, and moral and functional problems
111 112 113 115 118 121 124 125 127 128 130 131
7
Crucial Values In-Between Both Sectors
135
7.1
How and when specific values are important in contract agencies 7.2.1 Responsiveness 7.2.2 Transparency and Accountability 7.2.3 Lawfulness 7.2.4 Consistency, Reliability and Incorruptibility 7.2.5 Efficiency and Effectiveness How and when specific values are important in parapublic Organizations 7.3.1 Responsiveness 7.3.2 Transparency and Accountability 7.3.3 Lawfulness 7.3.4 Consistency, Reliability, and Incorruptibility 7.3.5 Efficiency and Effectiveness Intermixing, conflicts and problems in contract agencies Intermixing, conflicts and problems in parapublic organizations
135 135 137 138 141 143
6.4
6.5
7.2
7.3 7.4
145 146 148 150 151 152 154 156
Part III: Analysis, Conclusions and General Discussion 8.
The Public, the Private and the In-Between: A Comparative Analysis
8.1
Differences and similarities between public and private sector Values 8.1.1 The significance of organizational publicness Which values belong to which sector? Differences and similarities: Traditional distinctiveness, convergence or intermixing? How and when values are important in government and business
8.2 8.3 8.4
ix
161 161 163 165 168 169
8.4.1
8.6
How and when values are important in-between the sectors Integrating the quantitative and the qualitative: Adding contextuality and dimensionality Values and problems across the sectors
9.
Value solidity: General Discussion
179
9.1 9.2
Research results: Value solidity Strengths and limitations of the study 9.2.1 Issues meriting consideration in future research What does it mean anyway? Implications for academia and organizational practice 9.3.1 The academic debate on public and private sector values 9.3.2 Public and private sector organizations and managers
179 183 184
8.5
9.3
173 176 177
185 185 187
Notes
189
Appendices
197
Appendix A
197
Appendix B
205
Appendix C
207
Appendix D
211
Appendix E
215
Appendix F
216
References
217
Samenvatting (Summary in Dutch)
235
About the Author
243
x
“Hell hath no fury like a bureaucrat scorned.” Milton Friedman (1912-2006)
1.
The Morals of Government and Business
1.1
Introduction: Two conflicting moral syndromes1
Everybody knows there are large differences between the morals of government and the morals of business. For example, were government employees to ostentatiously invite clients to luxurious dinners and exotic trips abroad, scandals and public outcry would be the result and civil servants would be disciplined and fired. On the other hand, were business managers to donate the entire annual business revenue to welfare benefits for the unemployed, shareholders would be outraged, stock prices would decline, and board members would be fired and perhaps even prosecuted. Not only do acts such as these contradict the very nature of the public and the private spheres, they raise the question of whether government virtues are business vices and business virtues, government vices.2 Everybody knows that the primary interest of corporations is to make money and obtain market share in order to grow and satisfy investors. In contrast, government organizations focus primarily on maintaining order and stability, and in making sound policy, rules and regulations. Business pursues the private interests of shareholders, whereas government pursues the public interest of its citizens. Doing business involves entrepreneurship, courage, innovativeness and profitability, while government conduct embodies values like predictability, neutrality, obedience and lawfulness. It therefore appears to follow that because government and business organizations have different goals, tasks and roles to fulfill within society, they also embrace different morals and values. Or are these common perceptions wrong? According to anthropologist Jane Jacobs (1992), two radically distinct value systems characterize the public and private sector. In her seminal work Systems of Survival, she distinguishes between the commercial moral syndrome and the guardian moral syndrome, although not entirely synonymous, taken here to represent the value systems of government and business, which are themselves “a pair of contradictions” (1992: xi, xii): Because we possess … two radically different ways of dealing with our needs, we also have two radically different systems of morals and values – both systems valid and necessary. The human race has accomplished the two moral and value systems during millennia of experience with trading and producing, on the one hand, and with organizing and managing territories, on the other hand. For one thing, many of us have taken on casts of mind so skewed toward one set of morals and values that we have little understanding of the other, and little if any appreciation of its integrity too. Behavior that (randomly) picks and
1
chooses precepts from both syndromes creates monstrous moral hybrids; you can’t mix up such contradictory moral syndromes without opening up moral abysses and producing all kinds of functional messes.
Jacobs’ statement contains two strong normative claims. First, it argues that government and business possess radically different and contradictory morals and values; and second, it proposes that these morals and values should be kept as distinct as possible because intermixing is problematic and undesirable. To illustrate this latter, Jacobs (1992: 81) identifies particular types of “functional and moral quagmires” into which organizations and institutions sink when they confuse their own appropriate moral system with another; namely, inefficiency and waste in business organizations and corruption and unethical behavior in government organizations. Throughout history, the distinctiveness of these moral systems has not only been emphasized and legitimized but also ridiculed through powerful imagery and ideal types, both in positive and negative terms. The former is typified by Adam Smith’s [1776] (1937: 19) well-known description of the morals of the marketplace: “It is not from the benevolence of the butcher, the brewer, or the baker, that we expect our dinner, but from their regard to their own interest.” The latter is displayed by abundant stories in Western culture about businessmen who, like Dickens’ Scrooge or the landowners in the Grapes of Wrath, seemingly care nothing about the effects of their actions on the environment, the well-being or even the lives of ordinary people. At the other end of the scale lies the classical notion that what makes government – and government bureaucrats – different is that it should work in the public interest (Appleby 1945, 1952), and what government must do well is create value for society (Kirlin 1996). One widely used, but also misused, illustration of this principle is Max Weber’s [1921] (1968) rational, equal and unpartisan bureaucratic organization that is run by civil servants who are loyally and objectively serving the public interest, but, when perceived negatively, also suffers from impersonality, ‘red tape’ and trained incapacity and is bound by constraining rather than effective rules that are largely a product of the organization itself (e.g., Bozeman 2000; Guyot 1962; Merton 1968). Based on the above, the classical images of government and business morals are clear. However, the question can also be posed as to whether these images are grounded in reality. In other words, do they have anything to do with daily organizational life in the twenty-first century? Are government and business really that different in terms of the values that guide organizational conduct? What is the role and importance of specific organizational values, and how are they prioritized? How, when and to what extent are specific values like efficiency and accountability important in both sectors, and how important are the different contexts in which government and business conduct are embedded with regard to the differences in their actualization and application? Interestingly, despite all the recent academic attention toward changing and shifting values in the administrative sciences, little is actually known about the values that dominate public and private sector organizational decision making.
2
Although structural and managerial differences between government and business organizations have been addressed extensively in the literature (Allison 1979; Boyne 2002; Buchanan 1975; Coursey and Bozeman 1990; Knott 1993; Noordegraaf and Stewart 2000; Nutt 2006; Rainey and Perry 1988; Ross 1988), only a few comparative studies examine ethical differences between both sectors (Berman and West 1994; Buchanan 1976; Solomon 1986; Wittmer and Coursey 1996). In addition, when ethical issues do come into play, the academic debate immediately becomes far more heated and controversial, but also much less empirical. Moreover, statements on what government and business values are (not) and should (not) be are intermingled, making this a debate with an ideological twist (cf. Noordegraaf and Abma 2003: 857). Such ideological controversy has been fueled by contemporary debates on the influence of two recent paradigms: The new public management (NPM) and related philosophies – which concern the adaptation of business-like techniques and principles by government organizations – and corporate social responsibility (CSR) – which dictates that companies engage in actions that further some social good beyond their immediate interests and that which is required by law. In these debates, the question is often posed whether NPM and CSR may cause intermixing or even convergence between public and private sector value systems. Yet little is known about the actual extent of such intermixing or convergence. For example, would an empirical comparison result in fragmented or even contradictory portrayals? Or would one specific set of values predominate in each sector? In other words, will value solidity be the state of affairs, denoting that the values of government and business organizations have remained relatively stable and classical differences do still exist, or is value fluidity a term that fits better with contemporary organizational reality? Additionally, what would hold true for parapublic and hybrid organizations like hospitals and universities that lie somewhere in-between both sectors? Which values would dominate their decision-making processes? Pressing questions, in urgent need of empirical answers. 1.2
NPM: Business values in government organizations?
Since the 1980s, major changes – both in theory and practice – have taken place in the way public sectors are organized and operated, changes that have come to be generally known as public management reform (maintain, modernize, marketize, minimize; see Pollitt and Bouckaert (2004)) and ‘New Public Management’ (NPM) more specifically (Maesschalck 2004: 465). However, NPM has proven a highly contested concept and has thus been partly supplemented and replaced – but also somewhat cannibalized – by concepts like reinventing government (Osborne and Gaebler 1992), performance oriented management (Pollitt and Bouckaert 2004) and management by measurement (Noordegraaf and Abma 2003). In general, the term refers to reforms that have become popular in many, but particularly Western or OECD, countries, such as the introduction of performance management systems and performance-related
3
pay, more responsibility and accountability for public managers, more competition in the public sector and the introduction of quality management techniques (Pollitt and Bouckaert 2004; Hood 1991, 1995; in: Bozeman 2007). The concept initially emerged in the United Kingdom under the Thatcher government before being exported to many nations in Europe and other parts of the world. Conservative governments in New Zealand and Australia have become two of its most enthusiastic exponents (Bozeman 2007: 69).3 Within academia, the concept of NPM was first introduced by Christopher Hood (1991, 1995), who conceived of it as a specific set of administrative doctrines on how public administration should be conducted (Hood 1995: 97). Central to the NPM philosophy, as well as to the fiercest criticisms it has evoked, is the application of economic individualism to public sector conduct; that is, the pursuit of private interest rather than the pursuit of public interest that has always characterized government and distinguished it from business (cf. Appleby 1945; Bozeman 2007; Frederickson 2005). Bozeman (2007: 68) explains such application as follows: The ideas of economic individualism affect the design and analysis of public policy, but also its management and implementation. It seems indeed fair to say that the great preponderance of public management ‘reforms’ and innovations of the past two decades have primarily been ones based on market or quasimarket theories and concepts. Sometimes it is in the name of privatization or contracting out and sometimes it is more general notions of ‘managing government like a business.’
Such marketization of the state (Pierre 1995) may not only imply the adaptation of private sector tools and techniques but also its principles, philosophies and values (Bellone and Goerl 1992; DeLeon and Denhardt 2000). In other words, the classical Weberian government organization characterized by neutrality, predictability, obedience and lawfulness might have been significantly altered during the last 25 years by business-like values, some of which constitute the direct opposite of these ‘ideal’ traits (see Bovens et al., 2001, and van den Heuvel, Huberts and Verberk 2002 for the Dutch case, specifically). Indeed, the exact nature of the influence of NPM and its sister concepts on traditional public sector values like impartiality, fairness, and neutrality has been a recurrent and contested issue in recent administrative discourse (e.g., Eikenberry and Kluver 2004; Frederickson 2005; Kernaghan 2000, 2003). Scholars have suggested that NPM is characterized by a ‘managerialism’ (Pollitt 1993) that embraces traditional business values like efficiency, effectiveness, innovation, profit, competence and quality (Lane 1994; Tait 1997). Evoked by the fear that an overemphasis on business administration value assumptions has come at the expense of the unique value set necessary to service of the public interest (Maesschalck 2004), some authors have advocated a clear set of public service values (e.g., Van Wart 1998) to which other authors have responded (e.g., Kernaghan 2000, 2003). Still others have proposed new
4
models like new public service (Denhardt and Denhardt 2000) or, more recently, a public values-based management approach for the public sector, ‘managing publicness,’ as opposed to an NPM-based public management framework (Bozeman 2007: 175). From that perspective, NPM may – at least as a leading academic concept in high-income countries – already be past its peak in its purest manifestation (cf. Hood and Peters 2004). Nevertheless, as already argued, its core principles still seem to inspire many of the administrative reform agendas in practice, and the significance of the concept is far from vestigial. Indeed, Bozeman (2007: 77) suggests that “if NPM is no longer at the tongues of scholars and policy makers around the globe, neither is it yet relegated to the dustbin of managerial reform fads.” In the Netherlands, new public management techniques have become quite popular at the national as well as the local level, including performance measurement (De Bruijn 2002; Kickert 2004; Pollit and Bouckaert 2004; Ter Bogt 2004; Van Helden and Jansen 2003). Within for example the police sector, this stimulated fierce discussions whether new types of indicators and management would result in better performance or in strategic behavior by police officers and more corruption and fraud. A small number of scandals were explicitly related to developments in new public management and even very specifically to Osborne and Gaebler’s (1992) work (see Bovens 1996; and Yesilkagit and De Vries, 2002 about an overenthusiastic civil servant who’s secretive entrepreneurial activities caused large financial losses). 1.3
CSR: Public values in private sector organizations?
Almost concomitantly with the NPM movement, influential new ideas and philosophies on the moral responsibilities of corporations entered the business sector. Although the ever-increasing attention to ethics and values within business administration and economics during the last 25 years may seem to have been influenced by political and stakeholder pressures rather than structural convictions, it is somewhat unfair to accuse private sector academics and managers of applying a minimalist approach to ethics. For instance, Milton Friedman’s (1970: 122) notorious statement about the responsibility of business – “there is one and only one social responsibility of business and that is to increase its profits so long as it stays within the rules of the game” – has often been wrongly perceived as immoral or at least amoral, given that Friedman’s reasoning seemingly involves no moral elements. What should be stressed is that Friedman was defending his principle on moral grounds. That is, corporate managers are dependent on and responsible for the capital provided by shareholders and cannot spend all this money for the good of society. Either they must make a substantive profit or the capital may be withdrawn from the corporation, which would cause many employees to lose their jobs. Nevertheless, Friedman’s reasoning exemplifies a ‘traditional business morality’ in which thoughts on right and wrong are seemingly limited to the corporation and its direct environment, and profitability is always the leitmotif.
5
Since Freidman’s pronouncement, however, this traditional line of reasoning has clearly changed both in theory and practice.4 Business ethics as such has become a reasonably mature academic subdiscipline, with people, planet, profit (Elkington 1998), corporate social performance (CSP) and corporate social responsibility (CSR) as its central theoretical and practical concepts (Kaptein and Wempe 2002: 29–39).5 The concept of CSR (e.g., Bowie 1991; Caroll 1979, 1991, 1999; McWilliams and Siegel 2001) – considered here to be a ‘corporate engagement in actions that further some social good, beyond the direct interests of the corporation and that which is required by law’ – is in direct opposition to Friedman’s perspective that the only corporate responsibility is to maximize stockholder profits within the ‘rules of the game.’ Rather, it inserts into corporate responsibility the dictum of going beyond obedience to the law and exceeding the moral minimum (cf. McWilliams and Siegel 2001). One particularly influential development has been the paradigm shift from stockholder to stakeholder theory. Originating as a merely instrumental concept coined by Freeman (1984, 1994) and later further developed by other authors (e.g., Donaldson and Preston 1995; Mitchell, Agle and Wood 1997), the stakeholder concept contains inherent normative elements that were soon recognized. For example, this perspective describes stakeholders as “every group that has a stake in the corporation” and subscribes to the notion that “the corporation has a moral obligation to honor the legitimate expectations of the stakeholder” (Kaptein and Wempe 2002: 213). Following this, the attention to the ethics of business, and to integrity management and corporate social performance in particular, is increasingly characterized by a more structural and institutional approach (29–39). As a result, the number of academic publications on CSR and related concepts has proliferated at an astonishing rate over the last three decades.6 CSR has become rather popular in the Netherlands (Cramer, Van der Heijden and Jonker 2006; Graafland and Eijffinger 2004; Kaptein and Wempe 2002). This does not only concern the big multinational Dutch-based companies such as Shell that anticipated stakeholder expectations with the introduction of the now famous ‘Shell Report’ (Kaptein and Wempe 2002). A majority of the business sector is taking corporate social responsibility more seriously. The CSR development has also been supported by government (e.g., www.mvonederland.nl) which is in accordance with a tradition of consensus government and cooperation with civil society and business in this small Western European country with 16.3 million inhabitants (Andeweg and Irwin 2005). Real-life events like the notorious Brent Spar affair and the controversy surrounding Shell’s activities in Nigeria, as well as the accounting scandals within Enron, Worldcom, Parmalat and Ahold, have also influenced and contributed to this increased attention to business ethics. Indeed, this increase amounts to a heightened concern in the business world for public opinion, public responsibility and – most interesting but also most complex – fundamental questions concerning public accountability and liability obligations. How far
6
does private responsibility for public goods reach, and to whom (and to what extent) are businesses accountable? How can corporations make complex tradeoffs between different relevant and important stakeholders? And to what extent must corporations contribute to sustainability and social justice, and help to advance human rights? Classical public sector deliberations, taking place in modern day corporate boardrooms. Given this trend, it seems that not only have traditional business values been entering government organizations, but also that traditional public values have become increasingly important to corporations. As a result, the corporate social responsibility (CSR) approach now emphasizes traditional public sector values like sustainability, (social) responsibility, and a more far-reaching interpretation of accountability, as well as empathy, solidarity, reliability and fairness (Kaptein and Wempe 2002: 237–46). 1.4
Value intermixing and moral and functional problems
Without directly referring to the moral quagmires and monstrous moral hybrids posited by Jacobs (1992), many public administration scholars consider the intermixing or convergence of values to be problematic and undesirable (e.g., Bovens 1996; Frederickson 1997, 2005; Jos and Tompkins 2004; Schultz 2004). Moreover, they state that the introduction of businesslike methods and values into government can be expected to result in an increase in integrity violations among public servants (Frederickson 1997: 177–181, 2005). Most particularly, Frederickson (2005: 178) emphasizes potential conflicts between crucial values like fairness and efficiency: “The private market is designed to be efficient but not to be fair; democratic self-government is designed to at least try to be fair, and hope to be efficient.”7 Thus, Frederickson fears that corruption and unethical behavior are on the rise because government organizations are run as if they were businesses. Others particularly critique the overemphasis on output and performance that, in their eyes, has come at the expense of traditional democratic-constitutional and public service values and seems to have guided policy creation and marketing in many administrations in Western countries over the last two decades (e.g., Eikenberry and kluver 2004). Indeed, Rosenbloom (2007: 28) claims that “the impact statements and scorecards that are available today lack attention to the protection and promotion of democraticconstitutional values, including individual rights, constitutional integrity, transparency, and the rule of law.” In similar fashion – albeit in a far less heated manner – the alleged influence of public values on the ethical rules of business has raised both concerns and positive comments. As a result, discussions on the problematic aspects of value intermixing involve not only an increased appeal to market values in the public sector, but also, according to Schultz (2004: 293), its converse: Although many would laud the move to encourage corporate social responsibility and ethical behavior, especially in light of the recent Wall Street
7
scandals, the intermixing of public and private functions raises vexing ethical questions similar to those when governmental and nonprofit entities intermix. The result may be that no clear set of ethical rules dominates.
Nevertheless, not all administrative scholars are pessimistic about this alleged intermixing of value orientations. Indeed, some optimistically emphasize the similarities between social and organizational values, pointing to the positive aspects of intermixing and hybridism (In ‘t Veld 1997; Voth 1999) and suggesting that the same values (should) apply to all types of organizations (Caiden 1999; Kaptein 1998). Clearly, scholars have contradictory views on the differences and similarities in, as well as the relationship between, the values in both realms. Moreover, even though many administrative scholars claim that government and business organizational values should be different, whether they really are merits further empirical analysis. This question also applies to the assumed relations between ‘sector strange values’ and moral and functional problems in public and private sector organizations (specified here in accordance with Jacobs (1992)), corruption and unethical behavior (i.e., integrity violations) in government organizations, and inefficiency and waste (i.e., decision-making inefficiency) in the business sector. Thus, empirical contributions to this debate are urgently needed if an ideological stalemate between both schools of thought is to be prevented. This study attempts to make such a contribution. 1.5
Research aims and research questions
The goals of this project are to empirically determine the differences and similarities between the values in public and private sector organizational decision making and the types of problems that might result from intermixing sector-specific values. To this end, the study is guided by the following central research question: What are the most important values in public and private sector organizational decision making and to what extent are values related to perceived problems in both sectors? The following sub questions will act as both guidelines and themes throughout this book (the corresponding methodology is indicated in parentheses): 1. What are the most important values in public sector organizational decision making, as distinguished in the public administration theory, empirical research and public sector codes of conduct? (Literature review)
8
2. What are the most important values in private sector organizational decision making, as distinguished in business administration theory, empirical research and business sector codes of conduct? (Literature review) 3. What are the most important values in public sector organizational decision making, as distinguished by public sector managers? (Quantitative and qualitative empirical research) 4. What are the most important values in private sector organizational decision making, as distinguished by private sector managers? (Quantitative and qualitative empirical research) 5. Are public sector organizational values in private sector organizations related to the perception of decision-making problems? (Quantitative and qualitative empirical research) 6. Are private sector organizational values in public sector organizations related to the perception of integrity problems? (Quantitative and qualitative empirical research) 7. How and when are specific values important in public and private sector decision making, under which conditions and circumstances, and to what extent? (Qualitative empirical research) 8. What are the most significant differences and similarities between the most important values in public and private sector organizational decision making? (Quantitative and qualitative empirical research) To fully answer the research questions the study must also address institutions other than core public and private sector organizations (cf. Coursey and Bozeman 1990). After all, an important aspect of the debate on NPM and value intermixing is the increase over the last decades in the privatization and autonomization of many public services. In fact, even though privatization and NPM are obviously not identical, they “are closely interwoven, so much that the two share many values” (Bozeman 2007: 72). One of the consequences of such sharing is the emergence of “an organizational world in which sector blurring is becoming as much the rule as the exception and in which new organizational forms are emerging that are not easily classified by conventional labels of government-business” (Bozeman 2004: xii). In the presence of such blurring, the extent to which a particular organization is on the public or private side of the spectrum – that is, its ‘publicness’ (Bozeman and Bretschneider 1994; Coursey and Bozeman 1990) – might account for differences in value orientations (Boyne 2002). More specifically, different value systems that compete with each other and culminate (or not) in different types of problems might be expected in more autonomous contract agencies with a business-like financial governance regime (Agentschappen, in Dutch), but especially within parapublic organizations (Lyons et al. 2006) or quangos (quasi-autonomous nongovernmental
9
organizations; Van Thiel 2000). Therefore, such organizations will be included in the study. Being interdisciplinary, this project will combine knowledge and concepts from administrative ethics, business ethics and organization science. To date, administrative ethics and business ethics, as well as general ethics management within the field of organization science and economics, have been separate areas despite considerable similarities not only in the application of theory but also in the issues and problems they address. Such similarities increase the relevance of integrating their diverse research directions (e.g., Lawton and Doig 2006; Menzel 2005; Siebens 2005). 1.6
Central concepts
Before the main question and sub questions can be answered, it is necessary to stipulate the exact meaning of ‘value,’ the way in which it should be perceived and its ontological and epistemological status. Doing so, however, leads to an oft-mentioned problem in organizational ethics scholarship – the continuing disagreement on defining and measuring values (Posner and Schmidt 1986: 448), which has led to a “values literature confusion” (Agle and Caldwell 1999: 327). According to Connor and Becker (1994), this confusion has resulted from the constant use of different instruments and concepts, which makes it almost impossible to accumulate a coherent body of knowledge. Because the literature is rife with inconsistent definitions of the value concept, as well as in the distinctions between value and such related constructs as attitude, belief and norm (Wiener 1988), values are “essentially contested concepts” (Gallie 1955: 169). Even more problematic, many studies on values, particularly within the field of public administration, have offered no proper definition whatsoever (Schreurs 2005). Nevertheless, specific definitions for the concepts pertinent to this study can be derived from the common literature and previous research. Accordingly, values are here defined as ‘qualities and standards that have a certain weight in the choice of action.’ As a corollary, organizational values denote ‘important qualities and standards that have a certain weight in organizational decision making.’ Such values are therefore broader, more general and less situational than norms, which are ‘regulations prescribing the proper conduct in general as well as certain situations.’ Integrity, a concept variously defined and used in the literature (Huberts 2005) but often thought of as one specific value, is used here as the umbrella concept of ‘acting or being in accordance with all the moral values, norms and rules that are valid within the context in which one operates.’ Based on the above, morals are then defined as ‘the collection of values and norms that provides a framework for acting.’ Accordingly, ethics is the systematic reflection on these values and norms, or rather ‘the systematic reflection on morality.’ ‘Ethics’ refers to thinking about what constitutes good and bad and should or should not guide human conduct. Thus, in contrast to the more pragmatic use of the terms ‘ethics’ and ‘ethical’ found so often in the U.S.
10
literature, this work, while unable to sustain the usage when drawing on the considerable body of U.S. scholarship, endorses the classical philosophical definition of the concept.8 The above definitions of the central concepts are outlined below in Table 1.1. Table 1.1: Definitions of central concepts Central concepts
Definitions
Values
Qualities and standards that have a certain weight in the choice of action Qualities and standards that have a certain weight in organizational decision making Regulations prescribing the proper conduct in general as well as certain situations Acting or being in accordance with the moral values, norms and rules that are valid within the context in which one operates The collection of values and norms that provides a framework for acting Systematic reflection on morality
Organizational values Norms Integrity Morals Ethics
Many authors apply different adjectives or levels to the concept of value, or rather; they distinguish between different value systems. Examples include professional and democratic, but also new and old, values (Kernaghan 2000, 2003); individual, organizational and legal values (Van Wart 1998); bureaucratic and social values (Gortner 1994) and cultural and stakeholder values (Seeger 1997). Others make even more elaborate distinctions, such as between different internal and external organizational levels and government stakeholders (Beck Jørgensen and Bozeman 2007). Obviously, these distinctions between levels or systems are to some extent analytical and artificial. However, some authors argue convincingly that organizations have their own specific set of values encoded in their culture (e.g., Deal and Kennedy 1982; Peters and Waterman 2005; Schein 2004) and that employee enculturation involves abandoning individual morals and values as the basis of ethical judgment and replacing them with an organizationally based collective ethic (Jackall 1988). It has also been argued by a number of scholars that institutions have goals, values and knowledge that “exist” independent of the organizational constituents and determine in large part the decisions and behavior of people within the institutions; that is, organizations have their own dynamics (e.g., de Graaf, 2005; French 1984; Pruzan 2001). This present study focuses on those values that dominate the decision-making practices of the organization. To clarify, this investigation does not address the important and related a priori question that is the central locus of the debate on public service motivation (PSM), which can be
11
conceived as a conceptual follow-up to classical studies on differences between public and private sector managers and the distinctiveness of a public service ethic (e.g., Buchanan 1974, 1975). PSM (Houston 2000, 2006; Lewis and Frank 2004; Perry 1996, 1997, 2000) primarily concerns the individual values and motivations related to a preference for employment in the public or private sector .9 Rather, the primary interest of this study lies in organizational value systems, which “span a continuum from weak, in which key values are not broadly and intensely shared, to strong, in which they are” (Wiener 1988: 535). Indeed, the internal dynamics within organizations render methodological individualism hard to defend, which posits that organizational conduct is always reducible to the conduct of the individual members of the organization (March and Olsen 1989). This social theory, however, has generally been discredited because it simply does not explain many social phenomena around organizations. In addition, methodological individualism is inclined to see moral phenomena as the aggregate consequence of individual conduct. Nevertheless, internal dynamics in organizations transcend individual conduct and decisions. Thus, even though this study surveys and interviews individual managers in government and business, the object of analysis is ‘organizational values,’ those values that play a role in organizational decision making. In this context, executive managers are perceived as spokespersons for their organizations and overseers of strategic decision-making processes. As the above discussion has provided an overview of what will be studied, the next question then becomes how such organizational values should be studied. 1.7
Studying what is valued most
Conducting research on values is a highly contested endeavor, especially from a methodological and epistemological viewpoint. Indeed, positivistic scholars within public administration, such as Herbert Simon, argue that constructs like motives and values are in no way scientifically researchable (Rutgers 2004: 27). Others allude to the (predominantly ontological) problem of reification – treating an abstraction as if it had concrete or material existence – or point to the enormous conceptual and theoretical disagreements on the exact meaning and usage of the value construct. Such issues give rise to many questions about the content and nature of the concept, not least its ‘researchability.’ At present, the sparse empirical research on values is monodisciplinary (i.e., rooted only in public administration, business administration or social psychology) and predominantly quantitative (e.g., Agle and Caldwell 1999; Bardi and Schwartz 2003; Goss 2003; Kim 2001; Posner and Schmidt 1984; Vrangbaek 2006). Yet, despite all the criticisms directed at such an approach in the literature, few studies have attempted an alternative – that is, interdisciplinary or qualitative – method. This failure applies equally to comparative research. To compound the problem, such extant studies predominantly use general instruments like the Rokeach Value Survey (Rokeach
12
1973) or the England Personal Values Questionnaire (England 1967), which address individual and social values rather than the specific organizational values mentioned in the administrative literature. In addition, the majority of the recent literature on government and business values and value intermixing is prescriptive and ideological, even though the “debate with an ideological twist” (cf. Noordegraaf and Abma 2003: 857) does offer propositions on which values belong to government and which to business and which problems may result from intermixing classical sector-specific values. These propositions can be tested through a comparative survey study. Thus, as a first step, there remains an urgent need for baseline quantitative data on value differences between government and business that will fill the existing void. Drawing generic inferences about government and business sector organizations in the Netherlands, and thus for the overall Dutch public and private sector, requires a large sample and numeric data. A second step in filling the current research gap would be to complement the quantitative results obtained in the first step with qualitative data. Such data are also urgently needed because surprisingly little is known about the contextual factors that accompany and influence value preferences, and about the gradations and extent to which specific values like accountability or effectiveness are important in a government or a business context (see e.g., Koppel (2005) on the ‘pathology of accountability’ and Bozeman (2007) on the fuzzy and opportunistic use of this concept in NPM contexts). For example, is accountability valued very highly under the same conditions in both sectors, or do public and private sector executives consider their organizations to be accountable to different audiences, on different issues and at different stages in a decision-making process? Similar questions can be posed regarding the exact relationship between values and conduct and between action and decision making. Admittedly, some research has focused on the links between specific values and specific modes of behavior (e.g., de Vries 2002 in public administration and Bardi and Schwartz 2003 and Schwartz 1992 in organizational psychology); however, all such examples, while they did establish links, are quantitative. What is needed in addition is more ‘thick description’ (Geertz 1973), more explorative and interpretative stories about the role and shape of organizational values in public and corporate decision making, as well as the significance of context, that is “the set of circumstances and conditions” (Gershenson 2002: 1) which surround specific values. Therefore, to answer the central research questions, this study makes use of a combination of methods, for three specific reasons in particular. The interviews as an added component to the survey enable a triangulation in its classical meaning – discounting the effects of errors in answering the (same) question to identify which values are most important in which sector and thus validate the quantitative results with qualitative data – as well as a complementarity of results – identifying those aspects of the roles that specific values play in decision-making processes – that cannot be addressed through a survey alone. Chronologically, one methodology paves the way for the other,
13
which methodological facilitation (Hammersley 2004) constitutes a third reason for selecting the combined approach that is also known as between-method triangulation (Denzin 1970). Because this approach combines methods with very distinct philosophical backgrounds, assumptions and predispositions, it is of course not undisputed (cf. Hammersley 2004; King, Keohane and Verba 1994; Soss 2006). However, a defendable rationale underlies this combination, whose foundations are detailed in chapter three. 1.8
Relevance
The results of this study will hopefully prove of significant academic relevance. As the introductory sections have shown, little is actually known about which values guide public and private sector decision making, even though it is a heavily debated issue. Moreover, the relationship between certain values and certain problems implied by a number of public administration scholars remains an empirical blind spot. Therefore, shedding at least some light on this issue would be valuable. Perhaps as important academically is the evaluation of the study design and method. For example, are the values addressed in the survey and interview appropriately relevant to the research problem or are more advanced clustering techniques and search methods needed to produce a set or multiple sets of even more prominent and relevant values. Likewise, to what extent does the, frequently called for combination of quantitative and qualitative methods produce satisfying and feasible outcomes? From a societal perspective, the research outcomes might help clarify elements of the often fuzzy and fragmented, but ever real, debate on what should be the guiding values and norms in the public sector (e.g., the OECD and EU reports on public values (SIGMA 1999)10 and ethics infrastructure (2002, 2004), and the 2003 study on guiding norms and values for Dutch citizens and politicians of the Scientific Council for Government Policy WRR), as well as in the business sector (e.g., the Sarbanes-Oxley Act in the U.S.11 and the Tabaksblatt Committee code on corporate governance in the Netherlands). Indeed, this debate has recently been broadened even further by the major role that values seemingly played not only in the last two U.S. elections (e.g., Green, Rozell and Wilcox 2006) but also in the latest Dutch, German and French elections.12 All these factors increase the need for more understanding of what the term ‘values’ refers to and which values are actually important in different social spheres. Within organizations, the study results could assist a more empirically based prioritization of values that, even though present in the (sometimes obligatory) codes of conduct adopted by many modern organizations,13 tend to be so abstract as to seem minimally related to practical work dimensions.14 Thus, organizations formulating mission statements and codes of conduct will find this research helpful, especially when CSR and NPM developments might make it hard to define sector-specific characteristics. In addition, the findings may help prevent integrity violations or even diffuse moral dilemmas by increasing
14
knowledge of the specific decision-making processes in which organizational values or value systems compete and collide in both public and private sector organizations. 1.9
Outline
The remainder of the book is divided into three parts. Part I, Theory and Methodology, consisting of two chapters, outlines the conceptual, theoretical and methodological framework of the study. Accordingly, chapter two addresses the ontological and conceptual aspects of the construct ‘organizational value’ by first conceptualizing and defining it, and then presenting a research perspective on the distinction between public and private sector organizations, and on the relation between values and organizational decision making. This chapter also briefly reviews the debate on organizational values in the public administration and business administration literature. Chapter three then explains the research methodology, including the choice of a combined approach that first juxtaposes and then integrates quantitative and qualitative methodology. It also provides an extensive discussion of the different types of data collection and analysis that influenced this choice of method and details its implementation. Part II, Empirical Results, comprises four chapters that present and discuss the empirical results of the content analysis, the survey research and the in-depth interviews. First, chapter four outlines the construction of a public-private value panorama distilled from the administrative and business ethics literature, and presents a set of values to be used as a survey research tool. Chapters five and six then present and discuss the quantitative and qualitative results for the public and private sectors, respectively, to give both a general and a specific picture of what is valued most and why. Lastly, chapter seven reports comparable findings (based on in-depth interviews only) for three parapublic organizations (a university, a hospital and a professional association) and two contract agencies (agentschappen). Part III, Analysis, Conclusion and General Discussion, provides an overall synthesis of the study results and discusses their implications; that is, the conclusions that can be drawn from the empirical material and the directions these suggest for future research. Specifically, chapter eight provides a final analysis of public and private sector differences and similarities, while chapter nine presents a more general discussion of the research results; most particularly, a methodological reflection that includes the limitations of the chosen approach and research directions that merit future attention.
15
16
Part I: Theory and Methodology
17
18
“Reality is merely an illusion, albeit a very persistent one.” Albert Einstein (1897-1955)
2.
Conceptualizing the Organizational Value System
This chapter presents a perspective on the exact nature of values and how they can be situated conceptually, ontologically and epistemologically in relation to organizational culture and decision making. It also discusses how values originate and evolve within organizations, in which ways they relate to organizational decision making and whether it is possible to label values a priori as specifically belonging to the public or private sector. 2.1
The V word: Widely used but seldom specified
In the first years of the twenty-first century, ‘values’ have once again become a central issue in many societal and political discussions: “Everybody is using the ' V word'now”15. Indeed, after the 2004 U.S. presidential elections, CNN exit polls rated moral values as the most important issue before jobs and the economy.16 Similarly, in the 2006 midterm elections, evangelicals, the ‘values voters,’ walked away from the Republican Party in large numbers, again putting values at the centre of the electoral turnpike.17 Values also played an important role in the 2005 debates on the EU constitution and in a number of recent European parliamentary and presidential elections, including those in Germany18, France19 and the Netherlands.20 However, exact clarifications of this topic of debate are rare. Indeed, far too often the V word is used in a very general manner, devoid of any context. This fuzziness raises the questions of what exactly people mean when they discuss values, which specific values are part of the debate and toward which societal or organizational realm or domain such value discussions are directed. Usually, the answers to these queries remain unclear. Rather, because people agree on their importance in general, values are frequently used as a strategic smokescreen to sidestep painful issues (e.g., Stiglitz 2003). However, difficulties begin once individuals agree on a basic value but attribute very different norms to it, or in other words, actualize the value in different ways. For instance, whereas most citizens might value equity and social justice, no employee would be willing to pay an 80 percent income tax to help achieve these goals, and no employer would appoint an uneducated laborer to the board of directors simply so he or she could earn a large income. The contested concept of value not only plays an increasingly important role in political and social debates, but in the last few years, it has also become increasingly prominent in business sector discourse (e.g., Kaptein and Wempe 2002; Paine 2003). Moreover, just as in real-life debates, academic discussions
19
on values, whether in business administration, public administration, philosophy, psychology or any other discipline, confuse more often than they clarify. Hence, the literature is rife with inconsistent definitions of the value concept and blurred distinctions between value and related constructs like attitude, belief and norm (Wiener 1988). As a result, like a number of other grand abstractions – art, social justice, religion and democracy (e.g., Gallie 1955 1956), to name a few21 – values are “essentially contested concepts” (de Graaf 2003: 22). Moreover, the failure of many studies on values to provide any clear definition of values (Schreurs 2005) has led to the earlier mentioned “values literature confusion” (Agle and Caldwell 1999: 327). 2.2
Conceptual, ontological and epistemological aspects of values
How can a value exactly be defined, and to what extent is it possible to distinguish between different types of values, such as moral versus instrumental (amoral) or individual versus organizational? Values can, at the most basic level, be perceived as “anything good or bad” (Pepper 1959: 7). However, various disciplines within the social sciences have offered other classic descriptions such as ‘enduring, continuing or collective moral beliefs,’ ‘convictions,’ ‘standards,’ or ‘principles’ that influence individual and group choices among alternative courses of action (e.g., Brandsma 1977; Jacob, Flink and Schuchman 1962; Rokeach 1973). Values are also perceived as guidelines for action and decision making, as they “refer to people’s reasons for acting and judgments about such reasons” (Ozar 1997: 645) and as either means to achieve ends (Rokeach 1973) or “desirable end-states” themselves (Guth and Tagiuri 1965: 125). A somewhat more elaborate definition is presented by de Graaf (2003: 22): “Values are qualities that are appreciated for contributing to or constituting what is good, right, beautiful or worthy of praise and admiration.” Anthropologist Kluckhohn (1951: 395) conceptualizes a value as “an implicit or explicit conception of the desirable,” rather than what is good, right or beautiful, which clearly implies a cognitive, almost rational, as well as an affective element. Whereas the desirable may never be reached in its ultimate form, it is important that individuals and groups try to attain as much as is naturally possible of what is conceived or ought to be desirable. In daily organizational life, values not only address what ought to be but also what is; not only what is good or desirable, but also what is simply the right thing to do in a decision-making situation (in order to ultimately achieve what is good and desirable from an organizational perspective). Therefore, in this study values refer both to qualities that contribute to what is conceived as the organizational good, as well as to general standards of conduct, which, although broader and less direct then norms, act as guides in choices that have to be made. Thus, values constitute qualities as well as standards that have a certain weight when choices are being made. A subsequent question is whether there are different types of values. Rokeach (1973), as mentioned above, applies a well-known distinction between
20
a terminal (or intrinsic) value, a goal or end to be achieved (e.g., equality), and an instrumental (or extrinsic) value, a means to achieve a terminal value (e.g., equality as a means to greater freedom). Nevertheless, the author himself somewhat repudiates this distinction by stating that most values can be both terminal and instrumental. Even earlier, others, including Kluckhohn (1951), had disputed this distinction. Moreover, Schwartz (1992) finds no empirical evidence for the existence of either instrumental or terminal values. Another related distinction sometimes applied in the literature is that between moral values, values with a clear reference to right and wrong (e.g., honesty and equality), and instrumental values, more teleological values with no such reference (e.g., efficiency or effectiveness). However, this distinction also appears problematic in its usage. For example, in an organizational context, it can be argued that proclaiming efficiency to be a leading organizational principle is as much a moral choice as an instrumental one. Moreover, how would a value like accountability, which concerns duty and outcome, be classified? In reality, most values belong to multiple categories that overlap, making a strict distinction between moral and instrumental minimally useful. This is not to say, however, that all possible types of values have the same weight and are appropriate for inclusion in this study. On the contrary, values concerned with aesthetics – such as a world of beauty, cleanliness or inner harmony (Rokeach 1973) – are of no interest here. Rather, this study focuses on those values that impact organizational decision making and determine what is right and wrong, desired and undesired, in organizational conduct. The lack of agreement on a ‘proper’ usage of concepts such as value – like that on a ‘proper’ usage of concrete value statements such as impartiality – stems from what philosophers call the nominal nature of values. From this perspective, values are nothing more than linguistic agreements made operational into nouns like ‘reliability’ or ‘equity.’ Consequently, their meaning is derived not from the essence of the concept but from its usage (cf. Karssing 2003); a meaning that constantly changes and differs from context to context. A related ontological issue in the study of values is the danger of reification: the treatment of an abstraction as if it had concrete or material existence. Sometimes, such reification seems to be the problem with less well-rationalized codes of conduct, in which core values are presented as tools or instruments that can be easily retrieved from a drawer like a pencil, a calculator or a written law. However, despite the conceptual confusion and the different ontological attributions to the value construct, it is broadly agreed upon among scholars that values cannot actually be seen or heard and can only be observed in the ways they manifest (Beyer 1981; Kluckhohn 1951; Rokeach 1973), for instance through spoken, written or physical behavior and action (e.g., by filling in a questionnaire or writing a dissertation). As suggested eloquently by Schmidt and Posner (1986: 448), “[t]hey are so deep-seated that we never actually ‘see’ values themselves, what we ‘see’ are the ways in which people’s values manifest themselves: in opinions, attitudes, preferences, fears, and so on.” The best that can be said is that values never occur by themselves: they never appear
21
unaccompanied. Rather, they are attached to objects (“a gun is a bad thing”) or to people and their moods (“I feel good today”) and are manifested through behavior and action (“that is an effective decision”) (de Graaf 2003: 22). Although such a conception has epistemological and methodological implications for the way in which values can be studied (that is: obtaining valid knowledge on how values manifest themselves in real-life situations), it certainly does not exclude verbal or written manifestations of value preferences as a method of study. According to Kluckhohn (1951: 406), sometimes what people say about their values is, truer from a long-term viewpoint than are inferences drawn from their actions under special conditions: “As a matter of fact, people often lie by their acts and tell the truth with words. The whole conventional dichotomy between acting and telling is misleading because speech is a form of behavior itself.” The epistemological and methodological implications are dealt with more specifically in the next chapter, as they relate to the choice for a combination of methodologies. 2.3
Organizational values: A definition
Besides distinctions between different types of values that are neither specifically supported nor used here, many authors also apply different adjectives or levels to the concept of value, or rather, distinguish between different value systems. Such labels, as previously mentioned, include professional and democratic, as well as new and old, values (Kernaghan 2000, 2003); individual, organizational and legal values (Van Wart 1998); bureaucratic and social values (Gortner 1994); and cultural and stakeholder values (Seeger 1997). Others make even more elaborate distinctions, such as between different internal and external organizational levels and government stakeholders (Beck Jørgensen and Bozeman 2007: 359). Obviously, such distinctions between levels or systems are to some extent analytical and artificial. Nevertheless, previous studies argue convincingly that organizations have their own specific sets of values encoded in their cultures (e.g. Deal and Kennedy 1982, who state that values constitute the “bedrock of any corporate culture” (21), but also Peters and Waterman 2005, and Schein 2004) and part of the employee enculturation process involves abandoning individual morals and values as the basis of ethical judgment and replacing them with an organizationally based collective ethic (Jackall 1988). In addition, as pointed out previously, goals, values and knowledge that exist independent of organizational constituents, by determining in large part decisions within the institution, result in organizations having their own dynamics (e.g., de Graaf 2005; French 1984; March and Olsen 1989; Pruzan 2001). Although such a perspective implies that an organizational culture is not a static construct but is in part constructed and recoded by individuals entering the organization (e.g., Schein 2004), the focus of this present study remains those values that dominate the current decision-making practices of the organization. Therefore, in line with the definitions of central concepts presented
22
in chapter one, here, organizational values are defined as ‘qualities and standards that have a certain weight in organizational decision making.’ There is a slight difference with the concept of managerial values (cf. Posner and Schmidt 1986), although most empirical research on values in organizations concerns managers because they have a more informed view on important, strategic decision-making processes than, for example, the average street-level bureaucrat or call-center receptionist. In addition, an organization as such cannot be surveyed or interviewed. Therefore, even though the outcomes of the study will be based on individual responses (as discussed in more detail in chapter three) its object will be organizational, and not individual, values. Prior to any conceptualization of how values within the organizational value system can be related to public and private sector organizational decision making, attention must be paid to what constitutes the public and private in organizational life, and how these concepts are related to values in the literature. Nevertheless, because the public-private distinction constitutes an entire field of study in itself and is arguably one of the most classical issues of dispute within public administration,22 this discussion constitutes only an overview. 2.4
The significance of ‘publicness’
The public-private distinction is certainly not one that is applied exclusively to organizations. From a more philosophical viewpoint, the public and private domains are often characterized in terms of life spheres, as in Montefiore and Vines’ (1999) conceptualization of the private life sphere as that part of society over which the government has no authority. One well-known perspective on the public-private distinction is that of Hannah Arendt in her Human Condition (1958), which accords ‘public’ two different yet related meanings: on the one hand, “everything that appears in public can be seen and heard by everybody and has the widest possible publicity”; on the other, public entails “the world itself, in so far as it is common to all of us and distinguished from our privately owned place in it” (1958: 50–52). When discussing this grand dichotomy, Weintraub (1997: 14) makes the following observation: The public/private distinction is not unitary, but protean. It comprises not a single paired opposition, but a complex family of them, neither mutually reducible nor wholly unrelated. And these multiple (but also overlapping) discourses of public and private do not simply point to different phenomena; often they rest on different underlying images of the social world, are driven by different concerns, and raise very different issues. When we forget this, we not only talk past each other, but confuse ourselves as well.
Geuss (2000: 107) emphasizes that public and private cannot be determined a priori. Therefore we cannot ask for the implication of the distinction for a certain issue: “rather, first we must ask what this purported distinction is for, that is, why we want to make it at all.”
23
It is clear from the above that there is no one single public-private distinction, a conclusion also drawn by Rutgers (2003: 21), who states that traditions exist that share interpretations, but within those traditions strong diversity is apparent: “The predicate ‘public’ or ‘private’ is an ad hoc label, or it is derived from a specific normative perspective.” Moreover, the distinction cannot be validated empirically: “It is explicitly theoretical, ideological or at least normative in nature. That means that it is an analytical distinction that is helpful in order to observe and criticize phenomena (as being similar or different), or to prescribe a desirable reality” (15). Here, references to the private sector or domain allude to business organizations, while mention of the public sector or domain refers to government organizations. This current study will draw to a large extent on Coursey and Bozeman’s (1990) dimensional approach of ‘publicness’ to specify and make operational the analytical distinction and will analyze its predictability for value orientations in organizational decision making empirically. From a sectoral or organizational viewpoint, the distinction between public and private is arguably the most fundamental, yet at the same time most controversial, concept in public administration and organization science. Indeed, without such distinction, public administration would not exist as an academic discipline independent of business administration. Yet, as stated above, consensus has been reached neither on the shape and extent of this distinction nor on its salience and validity.23 Even the earliest administrative scientists trivialized as well as emphasized the intrinsic differences between public and private (now business) administration (Wilson 1887; Fayol 1929; Waldo 1980, 1984). It was Woodrow Wilson (1887: 209), arguably the first public administration scholar and later President of the United States, who stated that “the field of government is a field of business.” He thus saw the civil service reform attempts of his day – aimed at improving the politicized civil service of the time24 – as “opening the way for making it businesslike” (210). In describing the exchange between public and private sector techniques, the first Dutch professor of public administration, Van Poelje (1931), identified the ‘osmosis’ phenomenon. Such early observations seemingly imply that the more recent debate on value intermixing is perhaps not as recent as popularly thought, especially not in the United States where a ‘business-like government’ seems less of a general issue than in many European countries.25 On the other hand, with the same aim of depoliticizing organizational conduct in the public sector, Weber (1968) advocated efficiency and neutrality as central values in his ideal bureaucratic apparatus, and he was a European scholar. Therefore, the predispositions of many administrative scholars to simply contrast Weberian with NPM values is at best too simplistic, and perhaps even historically short sighted (see final section of this chapter). Nevertheless, the intensity and probability of modern-day NPM and CSR developments certainly differ from the externalities influencing administrations at the beginning of the twentieth century.
24
Again, it is clear that emphasizing and marginalizing public and private sector differences can go hand in hand; and in the debate on public and private sector differences, normative arguments are seemingly interchangeable with empirical observations, which Noordegraaf and Abma (2003: 857) comment on as follows: Normative arguments might, of course, under one condition, be more convincing than empirical arguments: They must match normative predispositions. The attractiveness of normative arguments depends on one’s ideas about democracy and the nature of ‘publicness.’ There is no firm basis for determining whether public organizations are unlike private sector organizations. There might be empirical ‘evidence,’ but this does not constitute a convincing reason for drawing the conclusion that they should not be like private organizations. The debate, then, becomes a debate in which the gap between ‘is’ (fact) and ‘ought’ (norm) plays a central role, and moreover, it becomes a debate with an ideological twist.
When only taking into account organizational structure, it seems fairly simple to determine the character or status of a particular organization based on legal status (i.e., ownership), whether funded by government or owned and financed by private investors. There is, however, more to the public-private distinction than just funding. For instance, Perry and Rainey (1988: 182) argue that the extent to which organizations are subject to external control by government or market is as least as important, as are the tasks that organizations fulfill: “The concepts are multidimensional, and this in turn complicates the definition” (183). One of the earliest theoretical statements on the relation between organizations and the political economy was offered by Dahl and Lindblom (1953), who observed that every society must choose between variants of two fundamental decision systems, which the authors refer to as ‘modes of social control.’ This latter dimension refers to the extent to which major components of an organization’s domain (e.g., rules, mission, goods and services) are subject to relatively greater external controls by a polyarchy or the markets (in: Perry and Rainey 1988: 193). Accordingly, Coursey and Bozeman (1990) examine two different concepts of ‘publicness,’ one based on the organizations’ legal status or ownership, the other a dimensional concept on which organizations can be more or less public depending on the degree of external political control of their resources and activities. However, whereas Bozeman’s (2004: xii) claim that “all organizations are public to one degree or another (making publicness a matter of degree)” clearly is endorsed here, a recurrent complication in assessing the implications of public-private distinctions relates to organizational functions or tasks, an aspect largely dismissed as a separate dimension in the debate on this issue. It is argued here that public and private versions of such institutions as hospitals, universities, accounting agencies and detention facilities might have much more in common with their functional counterparts in the other sector than with totally different functional types in the same sector (cf. Scott and Falcone 1998).
25
Moreover, organizational tasks, the way in which they should be executed and the question whether they are public or private tasks, are a key issue in the (Dutch) debate on privatization and agency creation (cf. Leeuw 1992). Therefore, this present study includes organizational tasks as a third dimension of the public-private distinction. In sum, this research applies the dimensional approach to characterize organizations and makes it operational as follows: −
− −
The extent to which organizations are constrained by political control (full ministerial responsibility, financial political control only, very limited direct political control, or no direct political control whatsoever); The way in which organizations are funded and financed (fully publicly financed, 99–50 percent publicly financed, 49–1 percent publicly financed, fully privately financed); The extent to which organizations perform public or private tasks in order to reach public or private goals (public = delivering goods and/or services to citizens with a nonprofit intention, meaning the goods and/or services are, in principle, accessible for everyone entitled to them; private = delivering goods and/or services to specific groups of consumers with a for-profit intention).
According to Boyne (2002: 112), publicness may well be a strong explanatory variable for managerial value differences between the public and private sector. Despite little empirical evidence, and that not always methodologically sound, “there appear to be fairly firm grounds for concluding that managerial values differ significantly between private firms and public agencies” (112). One interesting point Boyne raises is that it “is unclear whether the distinctive values of public managers precede or are a function of employment in the public sector” (113). Posner and Schmidt (1996: 287), two of the few other scholars who have done such comparative work, also ask whether the differences and similarities are related mostly to organizational or individual differences, and to what extent “people with certain values and perspectives join the public sector purposely (in order to actualize certain values) rather than the private sector.” The question then becomes whether sectoral traits determine managerial and organizational value preferences or whether managers prefer a career in a certain sector in the first place to achieve congruence between personal and organizational values (cf. Posner and Schmidt 1993). This a priori question on personal work values is the central locus of the debate on public service motivation or PSM (Houston 2000, 2006; Lewis and Frank 2004; Perry 1996, 1997, 2000; Perry et al. 2006).26 In this regard, a recent study by Lyons et al. (2006) points to differences and similarities between (para)public and private sector knowledge workers in terms of general values (very similar) and work values (significantly different). Even though such an individual focus goes beyond the realm of this present study, it is expected that
26
publicness will account for organizational value differences between public and private sector organizations just as it does for managerial values. 2.5
Origin and evolution of the organizational value system
If institutions have goals, values and knowledge that exist independent of organizational constituents, if there is such a thing as an organizational value system and if organizational culture is not a static construct, then where do these organizational values come from? Edgar Schein (2004), in his seminal work Organizational Culture and Leadership, argues that cultures begin with leaders who impose their own values and assumptions on the group. If that group is successful and these assumptions are taken for granted, then the culture will define what types of leadership are acceptable for later generations of members. However, as its environment changes to the point at which some of its assumptions are no longer valid, leadership comes into play again, now as the ability to step outside the culture and start evolutionary change processes that are more adaptive. As Schein (2004: 28-29) explains it, All group learning ultimately reflects someone’s original beliefs and values, someone’s sense of what ought to be, as distinct from what is. If a manager convinces the group to act on his/her belief and if the solution works and if the group has a shared perception of that success, then the perceived gradually starts a process of cognitive transformation. First, it will be transformed into a shared value or belief and, ultimately, into a shared assumption (if action based on it continues to be successful). The derived beliefs and moral/ethical rules remain conscious and are explicitly articulated because they serve the normative or moral function of guiding members of the group. A set of values that becomes embodied in an ideology or organizational philosophy can serve as a guide and as a way of dealing with the uncertainty of intrinsically uncontrollable or difficult events.
Pettigrew (1979) argues in this context that an entrepreneur who founds an organization tends to put a personal stamp on the organization’s cultural components by creating the initial culture and then finding followers who fit into that culture. Wiener’s (1988: 537) discussion of the source of shared values adds the element of tradition: “any given organizational core value can be derived primarily from organizational tradition or charismatic leadership”. Thus, he suggests, values rooted in tradition are by definition time tested and thus lend stability and predictability to organizational functioning. In contrast, values anchored in charismatic leadership are less stable and permanent “and their internalization by members is determined to a great extent by members’ personal identification with that leader” (537). Beyer (1981), while admitting that at least one possible origin of organizational ideologies and values is the organizational leadership, applies an external rather then an internal perspective. That is, values and ideologies do not originate in founding entrepreneurs and organizational leaders but rather in the
27
organization’s external environment: “Since all organizational resources, including members, were at some point derived from the environment, many of the ideologies and values held by members must be viewed as originating in the environment” (168). However, organizations are resource-transforming systems (cf. Jackall 1988). That is, in an organizational context, outcomes are more then simply a sum of individual decisions in that “organizations determine in large part the decisions and behavior of people” (de Graaf 2003: 35); thus, dynamics that transcend individuals partly replace individual values as a basis for decision making by an organizationally based collective ethic (Jackall 1988). Beyer (1981) therefore addresses how ideologies and values are imported into organizations from the environment, how they are transformed within organizations and how unique organizational ideologies and values are created. Since conflicts between ideologies and values, or between different ideologies and different values, can occur within an organization, between an organization and its environment or between environmental components, organizational decision-making processes are exposed to many possible sources of conflict. Both internal and external factors thus seemingly contribute to the origin and ‘existence’ of organizational values.27 2.6
Organizational values, culture and decision making
How can values be related to organizational decision making, and what is the difference (and the overlap) between organizational values and organizational culture and ideology as overarching concepts? An often-assumed relationship in sociology, psychology and various organizational and administrative sciences is the link between expressed values and related types of behavior,28 a relationship that has to some extent been validated empirically. For example, Bardi and Schwartz (2003) report statistical relations between particular value orientations and behavior that is supposedly in accordance with these orientations. Likewise, Corman, Perles and Yancini (1988) find that entrepreneurial behavior is strongly influenced by their values, which in turn permeate their business strategies. This finding is empirically validated by Gorgievski-Duijvesteijn and Ascalon’s (2005) discovery of relations between specific value types and criteria of entrepreneurial success. In addition, other studies show that the strategic decision-making behavior of family entrepreneurs is influenced by the importance attributed to specific values like honesty, credibility, lawfulness and innovativeness (Koiranen 2002: 183-185). Although the relation between expressed values and types of behavior has been shown statistically, it is questionable whether this relationship holds oneon-one. For example, Kluckhohn (1951: 403) argues that “the realm of values is that of ‘conduct’ (regularities of action-motivation which are related to or imply conceptions of desirable and undesirable behavior) and not that of ‘behavior’ at all.’” From this perspective, conduct might be perceived as a selective behavior like decision making, especially because many behavioral acts do not seemingly involve a conception of the desirable followed by a subsequent ‘preference
28
infused selection’ from available modes, means and ends of action. For example, acts guided by needs and impulses, such as eating an ice cream, do not seem subject to reflection and deliberation even on a subconscious level. Nevertheless, as Klamer (2000: 2) elaborates, this applies only to the act of eating the ice cream; as soon as a choice must be made whether or not to eat the ice cream, or what type of ice cream, values come into play: Even the purchase of an ice cream may involve a negotiation among many different little and big values. Because I like to follow my senses I might indulge yet because I also value being healthy I may refrain. Or I might have an agreement with my partner to reduce my weight and so have to factor the value of being trustworthy…And if I have decided to enjoy myself that very moment, will I care about whom to give the business? How much do I value the values of a socially minded and therefore high-cost company that I will pay extra for its products? Do I care about the esthetics of the place? Does it matter who the other customers are? No algorithm will do justice to the complicated process that constitutes the purchase of an ice cream.
Consequently, many values play a role in every decision, and any type of decision always involves values, value conflicts and value tradeoffs in ways so complex that we are, to a large extent, unaware of them, meaning that many remain unreflected. Nevertheless, according to Kluckhohn (1951: 402), values are operative when an individual or a group selects one line of thought or action rather than another, insofar as this choice is influenced by generalized codes. Yet it is questionable to what extent an individual is aware of this selection process. In some cases, a conscious choice is made between alternatives for action; at other times, an action appears inevitable so the actor is unaware that any selection has been made (402). Thus, values can be regarded as subjective, yet relatively persistent, factors that influence decision making (cf. Beyer 1981). Another such factor is ideology, which Beyer (1981: 166) interprets as “relatively coherent sets of beliefs that bind some people together and that explain their worlds in terms of cause-and-effect relations.” As a result of this subjectivity, ideologies and values may be hard to distinguish empirically. Specifically, whereas ideologies explain the hows and whys of events and have an effect on the predictions of the likelihood of outcomes, values make some courses of action or some outcomes more desirable than others and so also influence the choice of courses of action (167). Thus, ideologies involve predictions, whereas values involve preferences. Both, however, are strongly related to the concept of culture and affect the process of choosing between courses of action expected to produce different outcomes (i.e., decision making). According to Schein (2004), cultural analysis works best if culture is seen as manifesting itself at the level of acts and espoused values but having its essence in the set of underlying assumptions shared by a group (the ‘values-in-use’). Specifically, he identifies common words and concepts related to and associated
29
with culture – with things shared or held in common by groups – among the major categories of which are the following overt phenomena (12-13): − − − − − −
− −
Observed behavioral regularities when people act (conduct): the language they use, the customs and traditions they evolve and the rituals they employ; Group norms: the implicit values and standards that evolve in a group; Espoused values: the articulated, publicly announced principles and values that the group claims to be trying to achieve (codes of conduct, mission statements); Rules of the game: ‘the way we do things around here’; Climate: the feeling that is conveyed in a group by the physical layout and the way in which members of the organization interact with each other, with customers or with other outsiders; Habits of thinking, mental models and/or linguistic paradigms: the shared cognitive frames that guide group perceptions, thought and language and are taught to new members in the early socialization process; Shared meanings: the emergent understandings that are created by group members as they interact with each other; Root metaphors or integrating symbols: not necessarily consciously appreciated group ideas, feelings and images that become embodied in the group’s material artifacts (e.g., buildings, office layout, logos).
Why then is the word ‘culture’ necessary when so many other words exist? Culture, Schein (2004: 14) argues, adds two further critical elements to the concept of sharing: it implies some level of structural stability – that is, it is not only shared but deep and stable (less conscious, less tangible and less visible) – and it patterns or integrate the elements into a larger paradigm that lies at a deeper level. Thus, the essence of culture lies in shared basic assumptions. However, having emotionally invested in them, if organizational members are forced to discuss them, they tend not to examine but to defend them (cf. Jackall 1988). On the one hand, the differences between values and basic assumptions seem more or less clear, and the distinction valid; on the other, the distinction complicates the research framework. According to Schein (2004: 28-31), there are two types of value systems, a conscious (espoused), and a subconscious (inuse) system, each with different implications for organizational conduct. On which level, then, are the central organizational values of an organization situated? It is highly probable that values at the conscious level will predict much of the conduct observable on the artifactual level (Schein 2004: 29). However, if those values are not based on prior learning, they may also reflect what Argyris and Schön (1978) term espoused values, “which tend to predict what people will
30
say in various situations but which may be out of line with what they will actually do in situations where those values should, in fact, be operating” (Schein 2004: 30). In analyzing values, one must, according to Schein (2004: 30), “discriminate carefully between those that are congruent with underlying assumptions and those that are, in effect, either rationalizations or only aspirations for the future.” Often such lists of values, written down in codes of conduct and value statements, are not patterned on, and are sometimes even mutually contradictory to and often inconsistent with, observed behavior. Such behavior tends to be based on what Argyris (1976) has identified as ‘theories-inuse,’ being “the implicit [basic] assumptions that actually guide behavior, that tell group members how to perceive and think and feel about thing sand tend to be neither confronted nor debated; therefore, they are extremely difficult to change” (Schein 2004: 31). It is interesting to juxtapose this perspective with Kluckhohn’s (1951) statements on the detection of values. Sometimes what people say about their values is, according to Kluckhohn (1951: 406), truer from a long-term viewpoint than are inferences drawn from their actions under special conditions: “As a matter of fact, people often lie by their acts and tell the truth with words. The whole conventional dichotomy between acting and telling is misleading because speech is a form of behavior itself.”29 It is indeed questionable to what extent such a clear distinction between ‘espoused’ and ‘in-use’ can actually be made and what its empirical value is. Here, it is acknowledged that those values that are publicly espoused, for instance in codes and missions, are obviously more explicitly and consciously endorsed then less visible values that influence conduct on a more subconscious level. Yet the notion of organizational values as conceptualized in this chapter applies to both the values that are sometimes propagated explicitly and those that are less explicit but also appear to be important as decision-making preferences. In sum, the object of this study is organizational values, qualities and standards that have a certain weight in organizational decision making. However, in recognition of the fact that both explicit and conscious and implicit and unconscious values are important in representing what organizational decision making values most, the research will employ a combined methodological approach to detect the most prominent values in the Dutch public and private sector. 2.7
Lack of comparative empirical studies
Conceptualizing the organizational value system and its influence on organizational decision making is a first step in building the theoretical underpinnings of this study. Such a conceptualization, however, does not yet shed any light on which specific values government and business organizations actually consider to be the most important and most prominent. For example, is efficiency a public or a private sector value? Moreover, is it at all possible to so easily attribute specific values to specific sectors, as is done so often in the
31
literature and in codes of conduct, for instance by distinguishing between the classical public values lawfulness and impartiality, on the one hand, and business traits profitability and innovativeness, on the other? The task of attributing values to sectors is further complicated by the reality that, although a widely recognized and rapidly growing body of empirical research exists on organizational ethics and values for both public sector (e.g., Beck Jørgensen 2006; Beck Jørgensen and Bozeman 2007; Bowman and Williams 1997; Goss 2003; Kernaghan 1994, 2003; Schmidt and Posner 1986; Vrangbaek 2006) and private sector organizations (e.g., Agle and Caldwell 1999; Posner and Schmidt 1984, 1993; Hemingway and Maclagan 2004; Watson et al. 2004), there is a dearth of empirical comparative research that offers a reliable picture of value prominence in both the public and private sector. Admittedly, a few studies have comparatively and empirically investigated such issues as ethical climate (Solomon 1986; Wittmer and Coursey 1996) and ethical management (Berman and West 1994); however, with few exceptions, these do not specifically address values. One notable exception is Posner and Schmidt’s (1996) comparison of the organizational goals, stakeholders and personal traits (adapted from the England Personal Values Questionnaire (PVQ)), which in general addresses managerial values. However, even though these authors conclude that the differences between the public and private sector executives appear greater than the similarities, the differences are slight and sometimes even marginal. Moreover, once again, this study does not specifically address organizational values. Similarly, a number of recent empirical studies in the field of public service motivation (PSM), have focused on individual values and the predispositions of public and private sector employees rather than organizational values (e.g., Buelens and Van den Broeck 2007; Karl and Sutton 1998; Lyons et al. 2006; Stackman, Connor and Becker 2006). In addition, these studies make no use of value concepts and methodological instruments grounded in extant administrative theory but rather apply instruments from the fields of sociology and psychology like the Rokeach Value Survey (Rokeach 1973) or Schwartz’s general values model (Bardi and Schwartz 2003; Schwartz 1992). Therefore, despite the value of their findings, they make limited contributions to an accumulative and interdisciplinary body of knowledge on what is valued most in government and business organizations. 2.8
Which values belong to which sector?
Renewed attention to values has been fueled by the intense debate on the marketization of the public sector and – as a possible consequence – changing public sector values, on the one hand (Van Wart 1998), and unclear ethical rules for business (owing to CSR), on the other (Schultz 2004). As already mentioned, contradictory views exist among scholars on the differences, similarities and relationships between organizational values in public and private realms. Many authors emphasize differences and in most cases consider value intermixing to
32
be problematic and undesirable (e.g., Jacobs 1992; Lane 1994; Frederickson 1997, 2005; Schultz 2004). Others, however, emphasize similarities, and stress that the same moral criteria and values can (or even should) be applied to all types of organizations (e.g., Kaptein 1998; Caiden 1999). Yet the dominant sentiment in the literature seems to be that over the last 25 years, business-like values – some of which constitute the opposite of those traits just mentioned – have significantly altered the classical neutral Weberian bureaucratic organization (Bovens et al. (2001) and van den Heuvel et al. (2002) address this topic specifically for the Dutch case). A recurrent and contested issue in recent administrative discourse is the exact nature of the influence of NPM and its sister concepts on traditional public sector values like impartiality, fairness and neutrality (e.g., Eikenberry and kluver 2004; Frederickson 2005; Kernaghan 2000, 2003); particularly given that NPM is characterized by managerialism (Pollitt 1993) and embraces traditional business values like efficiency, effectiveness, innovation, profit, competence and quality (Lane 1994; Tait 1997). One problematic aspect of both the debate on public-private intermixing and the value research in public and private settings is that scholars tend to blithely attribute certain values to the realm of either government or business, with little disagreement over which specific values belong to which types of organizations and sectors. For example, within the NPM debate, values are classified as ‘old’ or ‘traditional,’ on the one hand, and ‘new’ or ‘emerging,’ on the other (e.g., Lane 1994; Tait 1997; Kernaghan 2000, 2003). Thus, van den Heuvel et al. (2002) conclude in their empirical study on public sector values in the Netherlands that ‘efficiency’ is an NPM value (as opposed to values characterized as Weberian), while Weber’s ideal bureaucracy specifically “stresses the importance of functional specialization for efficiency” (Rosenbloom 1983: 447). The Weberian ideal was in fact predominantly inspired by Taylor’s (1917) scientific management approach – reminiscent of Woodrow Wilson’s belief in public administration as “a field of business” (1887: 210) – which in the 1930s was at the pinnacle of its influence on public administration (Rosenbloom 1983: 446). Indeed, the well-known organizational scientists Gulick and Urwick (1937: 192) proclaimed efficiency to be the “axiom number one in the value scale of public administration” and advocated that “politics could not enter the structure of administration without producing inefficiency.” Of course, Weber [1921] (1972) also advocated administrative virtues such as impartiality, lawfulness and expertise, which today are still at the core of many prescriptive writings on the functioning of government organizations that aim to guide such functioning. It should also be noted that Weber did not mention innovativeness and entrepreneurship. Yet it is historically shortsighted to proclaim that an emphasis on the efficiency of government organizations is something from the last two decades. Even from a less functional and more moral viewpoint, such a claim seems hard to defend. For example, Simmons and
33
Dvorin (1977: 217, in: Rosenbloom 1983: 447) describe how the specific values of the managerial approach were captured in certain terms from the 1930s: The ‘goodness’ or ‘badness’ of a particular organizational pattern was a mathematical relationship of ‘inputs’ and ‘outputs.’ Where the latter was maximized and the former minimized, a moral ‘good’ resulted. Virtue or ‘goodness’ was therefore equated with the relationship of these two factors, that is, ‘efficiency,’ or ‘inefficiency.’ Mathematics was transformed into ethics.
Thus, apparently, an economic or, from an ethics perspective, utilitarian approach to government conduct is not so much new public management as some scholars imply when they consider efficiency to be at odds with the public sector value of fairness (e.g., Frederickson 2005: 178). Indeed, Frederickson’s statement, in the 2007 Donald C. Stone lecture,30 that good public management equals social equity raises the question of whether scholars are actually addressing different elements, as regards discussions on what should guide employee and management conduct in public organizations, and confuse social and democratic values with organizational and professional ones. A similar confusion surrounds the value ‘effectiveness.’ Specifically, it is unclear whether the measurement of output and outcome is a typical private sector technique (which is being increasingly applied within governmental organizations) or whether public sector organizations have traditionally focused on outcome as much, or even more, as on the aspect of cost-effectiveness or efficiency. In this context, the results of Schmidt and Posner’s (1986) work on values among public sector executives are particularly interesting in that the public managers surveyed attributed considerably higher scores to ‘effectiveness,’ ‘reputation’ and ‘efficiency’ than to ‘service to the public’ (1986: 448). More important, these data were collected in 1982 before the true expansion of NPM. Clearly, the theoretical and empirical studies mentioned are far from consonant and produce no clear picture of what is valued most in each sector. Does this observation mean, then, that the recent debate on the ‘managerial state’ (Clark and Newman 1997) is a fad or just a recurrent approach to government conduct that has been ‘hyped up’? Arguably, the answer is no, since a structural application of economic individualism to government organizations and its employees is something more than simple organizational efficiency and effectiveness (cf. Bozeman 2007). Moreover, although there may not be an objective ‘new state of affairs,’ the debate about managerialism in the public sector (as well as on CSR in the business sector) has expanded and proliferated at an astonishing rate during the last 25 years, leading to entire new discourses and opinions about what should be appropriate public and private sector conduct. Yet, again, the answer depends on the locus of discussion. Is the debate about values with a more general public and democratic character or about specific organizational values that some might automatically consider more instrumental and thus undesirable from a public governance perspective?
34
As a consequence, not only is the debate on the values that (should) guide public and private sector governance strongly ideological, it is also contaminated by a lack of clarity on what exactly is being debated within which context. Moreover, it is precisely this confusion that turns the highly relevant debate on ‘changing public sector values’ (Van Wart 1998) into a fragmented and sometimes rather unsatisfying clash of ideological stances. Accordingly, it seems arbitrary to attribute specific values to specific sectors and organizations solely on conceptual or theoretical grounds. Rather, given the purpose of this study, it seems more appropriate to deduct values from the literature through a systematic review rather than to make use of extant value research instruments like those of Rokeach (1973) and Schwartz (1992). Such a review enables the creation of an initial substantive sectoral hierarchy from which to derive a feasible set of key values for empirical research. These latter steps will be taken in chapter four through a literature review and a content analysis of the most prominent organizational values in the administrative and business ethics theory, codes of conduct, and empirical research. This analysis will result in a public-private value panorama and a set of values to be used as a survey research tool. In addition, it will identify a number of specific integrity violations and decision-making inefficiencies to be analyzed in relation to specific values, thereby allowing the Frederickson (2005), Friedman (1970), Jacobs (1992) and Schultz (2004) hypotheses on value intermixing to be tested. Not only is such intermixing of sector-specific values expected to raise vexing ethical questions, it might also lead to various moral and functional problems; in particular, an increase in corruption and unethical behavior in public sector organizations and inefficiency and waste in private sector organizations.
35
36
“You can’t see the dust if you don’t move the couch.” (Robert Abelson, 1928-2005)
3.
A Combined Methodological Approach
This chapter on the research methodology first juxtaposes the quantitative and qualitative research traditions and then offers three specific arguments in favor of combining the two approaches given the object of the study. Subsequently, it details the three phases of actual data collection and gives an overview of the different data analysis and coding techniques applied to the empirical data. 3.1
Researching values: Combined methodological perspectives
Values are not only hard to locate and define but also difficult to study and measure empirically. That is, values are often approached from an ideological rather then an empirical angle (cf. Schreurs 2005; Beck Jørgensen and Bozeman 2007) and thus, methodologically, they remain undefined and uncontextualized for a certain audience, such as public and private sector managers. In this study, however, values are attached to a specific context: they have been conceptually situated and defined in direct relation to organizational decision making. In addition, those values that are part of the empirical study were derived from relevant literature through systematic analysis and then defined precisely and uniformly (as detailed in chapter four). Then how to study what is valued most in public and private sector decision making? A first step in identifying the most important values for and the most significant value differences between government and business organizations in the Netherlands, is to acquire baseline quantitative data. Such comparative empirical data are currently almost nonexistent. Nevertheless, even though, given the process described above, the data will already be relatively structured and contextualized compared to many quantitative studies on values, general limitations apply to the use of quantitative methods for studying ‘essentially contested concepts.’ For example, surveys on sensitive subjects always generate a certain social desirability bias (i.e., respondents attributing very high absolute scores to values such as ‘honesty,’ even though it is clearly impossible to be honest at all times in every decision-making context). Moreover, even when the items in a questionnaire are worded carefully and specifically, it can be difficult for respondents to contextualize and grasp the conditions and circumstances under which something is of a certain importance; in other words, to distinguish how and when something is important in addition to establishing that something is important. The importance of the decision-making context is exactly why, to achieve a complete picture, quantitative results must be complemented by qualitative data from in-depth interviews, case studies or discourse analyses, which provide
37
more insight into the context in and conditions under which the value is of weight in the choice of action. Indeed, Boyne (2002: 117) specifically advocates such use of multiple methods in studying public and private sector differences: In addition to more sophisticated quantitative research, it would also be useful to pursue a variety of qualitative methods in order to explore the distinctions between private firms and public agencies. Such qualitative approaches may help to illuminate not only whether publicness matters but also why and how.
Within the framework of this study, in-depth interviews on the importance and actualization of values in organizational decision making seem particularly suitable for obtaining a more complete picture of public and corporate value orientations. Nevertheless, whereas making a plea for combining qualitative interviewing with quantitative survey research is simple, combining and integrating the two methodological techniques that are supported by epistemologies that are so often contrasted and juxtaposed as to make “packs of scholars take sides in longstanding philosophical disputes and clash over bedrock beliefs that drive their research” (Soss 2006: 131), is more problematic. It is difficult to conduct survey research on values without falling into the trap of one-dimensionality or running the danger that values deemed crucial in daily organizational decision making are instead merely espoused “truisms” (van Rekom, van Riel and Wierenga 2006: 175). At least as complex, it may be argued, but certainly more time consuming is the construction of an interview guide, the selection and contact of respondents willing to discuss a complex and sometimes sensitive subject, arrangement of a secure interview setting and investment in a fruitful interviewer-interviewee partnership. Thus, the remainder of the chapter details the way in which these complexities were dealt with within the research design and the empirical process itself. The clashes and contrasts between the quantitative and qualitative worldview constitute the first topic of discussion. 3.2
Quantitative versus qualitative worldview: Clashes and contrasts
Whereas the combination of quantitative and qualitative techniques like surveys and interviews is often termed ‘triangulation,’ methodologists apply a more subtle perspective because the triangulation concept itself can be ambiguous. For example, Hammersley (2004: 1) points out that the term triangulation used widely in social research is “often treated as if its meaning were clear, and as if what it referred to was unproblematic.” Yet, according to this same author, the term triangulation has been substantially criticized and variously interpreted.31 Thus, a more subtle distinction is relevant here for three reasons. First, many different and contrasting views exist on the relationship between the quantitative and qualitative approach (e.g., King, Keohane and Verba 1994), or somewhat more broadly, between the positivist and the hermeneutic tradition (Ragin 1994). Second, different purposes underlie the combination of both approaches
38
(Hammersley 2004). Finally, at least four different meanings attached to the triangulation concept are identifiable in the research, which itself is related to these different purposes.32 Hence, before addressing the specific purposes of the methodological combination in this study, a discussion of the different views on the (im)possibility and (un)desirability of a combined method is in order. Hammersley (2004) distinguishes five different views in the literature on the relation between quantitative and qualitative research: 1. Quantitative methods are superior, being essential to a scientific approach. 2. Qualitative methods are superior, being especially attuned to understanding human social life. 3. Quantitative and qualitative methods are complementary in that through combination, their contrasting strengths can be maintained and their weaknesses minimized. 4. Quantitative and qualitative approaches derive from incompatible philosophical traditions, each true to its own terms. Therefore, researchers must make a personal commitment to one approach or the other and not mix or combine the two approaches. 5. The distinction between quantitative and qualitative approaches is fundamentally erroneous: a more subtle set of distinctions is required that capture available options for dealing with various aspects of the research process, from formulation of research questions through case selection and choice of data collection methods to analysis and report writing. Given all the preceding discussion, it is clear that this present study endorses the third view. That is, predominantly for reasons of complementarity, it applies a combination of quantitative and qualitative methods that can be expected to produce a more relevant and more valid portrait of sectoral value orientations than would be the case if only one method had been used. However, the relevance of the fourth view – that the two approaches are incompatible and should not be mixed – cannot simply be dismissed, especially with respect to research on values. As said, it is heavily contested among scholars from different philosophical and methodological schools whether constructs like attitudes or values can be researched at all, and if they can how this should be done. Such a debate stems inherently from differing worldviews. From a quantitative perspective, scientific phenomena are relatively one-dimensional and a construct like value is often reified, objectified or materialized. From a qualitative perspective, scientific phenomena are attached to settings, context and language, and inferences are drawn with regard to the meaning and nature, the interpretation, of social phenomena.33 According to Soss (2006: 131), the literature on this subject “often seems to imply that an interpretive project is one that is carried out by a particular type of person (an interpretivist) whose
39
worldview is defined by a particular epistemological and ontological paradigm (interpretivism)” that is totally different from any other paradigm. Such disputes over worldviews sometimes seem to divert attention away from logical and pragmatic discussions on what should be the proper research method to answer specific academic questions (cf. King et al. 1994). Obviously, both methodological schools will continue to draw on distinct epistemologies, and quantitative and qualitative techniques will continue to differ because the objectives they serve also differ. However, such differences do not preclude the combination of both perspectives to produce a combination of different research techniques. Rather, researchers should commit to the best method for answering the given research question as adequately as possible. Accordingly, the advocacy for a strict personal commitment to one methodological school is here considered unnecessarily rigorous and, arguably, even methodologically counterproductive. Soss (2006: 131) summarizes this pragmatic and eclectic use of different conceptual and methodological traditions as follows: I find it most helpful to apply the label ‘interpretive’ to the logics of specific pieces of research rather than to researchers themselves or to any philosophical first-traditions one might attribute to researchers. The interpretive/positivist distinction, in this usage, is a matter of practice rather than identity or worldview.
Indeed, great methodological value can be derived from a useful combination of these different worldviews if triangulation is perceived as a form of “epistemological dialogue or juxtaposition” (Hammersley 2004: 1). Thus, successful application of such a combination is related to researcher rationale – the fundamental reasons for and purposes of such a combination. 3.3
Multiple purposes of the combined approach in this study
This study makes use of a combined methodological approach for a number of specific reasons. First, the research design that embodies different phases of the data collection uses one method to prepare the way for the other; that is, “retrospectively to investigate the operation of the other method so as to judge what errors might have to be involved. Thus, postal questionnaire data may be used to check conclusions reached on the basis of semi-structured or unstructured interviews, or vice versa; while interpretations of interview data might be checked through participant observation, or vice versa; and so on” (Hammersley 2004: 2). To illustrate, before the survey study in this research could commence, a number of expert interviews had to be conducted to solidify the object of research and test the survey methodology. Next, a content analysis of values in the literature and codes of conduct, as well as different types of integrity violations and decision-making inefficiencies, was necessary to engender a set of values and items for inclusion in the questionnaire. Once the questionnaires had been collected and initial analysis of the survey data had
40
taken place, it became clear which values were the most interesting for inclusion in the qualitative interview guide. It also became apparent that the significance of the publicness dimension would make it worthwhile to include organizations with different sectoral characteristics in the interview phase. Here, the purpose of the combination was facilitation or “navigation” (Hammersley 2004: 3). Second, the ways in which the concept of value and its weight in organizational decision making have been addressed and defined in chapter two imply that neither quantitative nor qualitative methodology alone would be sufficient to identify what is truly valued most in decision making, either explicitly or implicitly. Therefore, more or less the same question was addressed using a combination of two methods that carry “opposite threats to validity so as to try to discount the effects of those sources of error on the findings” (Hammersley 2004: 3). In other words, the skeleton of the quantitative data needed supplementation with the ‘flesh and blood’ derived through qualitative interviews.34 It is this goal that denotes the classical concept of triangulation. A third and final reason for using a combination of methods was that the qualitative interviews were designed to further substantiate the importance of specific values through questions such as to what extent and in which contexts certain values were important. Here, the combination was used to provide answers to different questions on the same topic and thus served the purpose of complementarity (5). 3.4
Three phases of data collection
The combined approach resulted in three different phases of data collection. The first – further research of the literature, expert interviews and the construction of a research instrument – took place from September 2003 until December 2004. The second, a postal questionnaire, was distributed between January 2005 and May 2005. These quantitative data were then collected and analyzed between May 2005 and July 2006. The third phase, a series of in-depth interviews, was conducted from November 2005 until July 2006. The resulting qualitative data were then analyzed between January 2006 and May 2007. 3.4.1
Phase I: Desk research, expert interviews and content analysis
The empirical research phase began with the construction of a conceptual and theoretical framework (see chapters one and two). Such construction was based first on additional research into the literature on administrative ethics, business ethics, the public-private distinction in public administration and organization science, comparative research on public and private sector organizations and survey research methods. Additional information on public-private comparisons, public-private management and organizational values and organizational culture was then derived from expert interviews on a range of issues with Dutch scholars from different disciplinary and methodological backgrounds.35 This combined input resulted in a state-of-the-art overview on public and private
41
sector values research that formed the basis for a value panorama constructed through a content analysis, which culminated in a value set suitable for use as a survey research tool (see chapter four). This content analysis, “a research technique for making replicable and valid inferences from data to their context” (Krippendorff 1981: 21), used a clustering system and a search protocol of nine observation questions that determined the importance, relevance and characteristics of the values deduced from the literature (see Table 3.1).36 Table 3.1: Observation questions that guided the content analysis Observation question
Answer categories
a. General data
Document code
b. Value
(Name value spelled out)
c. Definition present?
Yes No
d. Definition
(Definition spelled out + source)
e. Type of value
Moral (original label from author)
f. Justification
Empirical Conceptual Unknown
g. Relative importance
Low High Unknown
h. Type of organization
Society Public sector Government organization Other type of public organization Private sector Unknown
i. Organizational level
Group Larger group (unit) Organization as a whole Unknown
The clustering of the hundreds of value statements derived from the literature was conducted as follows. First, only those values that appeared more than once in the literature were chosen for inclusion in the analysis. The most important remaining values were then divided among so-called mother values with a strong resemblance in meaning (61 for the public sector; 45 for the private sector). Further clustering, in which values that overlapped were either filtered out or integrated, resulted in 30 values for each sector. The observation
42
questions or differentiating variables were then used to create an empirical hierarchy and select a top 13 for each sector. A specific value’s rank in the hierarchy was assigned based on its score on each of four crucial determinants: the extent to which the value was (i) explicitly described as a public or private sector value; (ii) considered organizational; (iii) empirically validated and (iv) described as (highly) important in the particular piece of literature. For instance, values that were explicitly characterized as organizational and highly relevant to the specific sector were determined to be more important than values mentioned in the text without an organizational or sectoral specification. For the public sector, the review included seven relevant books on administrative ethics and values – in particular, Cooper (1998, 2001), Heidenheimer, Johnston and Levine (1989), Lawton (1998), Sampford and Preston (1998), Van Wart (1998), and Williams and Doig (2000) – as well as 46 issues of two prominent journals, Public Integrity (1999–2003) and Public Administration Review (1999–2002). Also included were four government reports on ethics and values2 and five public sector codes of conduct.3 An extensive review of the business ethics literature on prominent private sector values was also conducted that included the following 11 books: Bird (1996), Boatright (2000), Bok (1978), Bowie (1992), Donaldson and Dunfee (1999), Gauthier (1986), Kimman (1991), Montefiore and Vines (1999), Nash (1990), Quinn and Davies (1999), and Wempe and Nelis (1991). Also included were 25 issues of three journals: the Journal of Business Ethics (1999–2001), Business Ethics Quarterly (2000–2001) and Business and Society (2001), as well as research on codes of conduct by Kaptein (2004). The results of this content analysis are reported extensively in chapter four. 3.4.2
Phase II: Self-administered mail survey
Based upon the value set derived through the literature review and content analysis, and assisted by the data gleaned from the expert interviews, the tutorials and earlier survey research on values by research group colleagues van den Heuvel at al. (2002) and Lasthuizen and Kolthoff (conducted in 2003; see Kolthoff 2007), a seven-page self-administered questionnaire (Tourangeau, Rips, and Rasinski 2000) was constructed in November and December 2004. This questionnaire not only measured organizational values but made a first attempt to establish a relation between specific values and specific categories of integrity and efficiency problems (see Appendix A for the survey instrument). All participants responded to 17 statements on the extent to which they had encountered different types of integrity problems during the last year, including cronyism or fraud and decision making problems like unnecessary delay or excessive participation procedures (see Table 4.8). A five-point Likert-type scale (from ‘often’ to ‘never’) was used for attitude measurement. The questionnaire explicitly asked respondents to rate and rank those values considered “most important when decisions are being made within the unit or
43
organization that you supervise,” emphasizing the values that guide organizational decision making over the respondent’s individual moral opinions. Participants were thus asked to report for their entire organization (or subunit) rather than displaying their personal managerial preferences. To reduce the effect of individual respondent interpretation, the 20 organizational values were accompanied by clear definitions. To characterize the extent to which the participating organizations could be characterized as public or private, the survey also included questions on three traditionally distinctive features of public and private sector organizations: organizational funding, political authority and control and organizational tasks (cf. Dahl and Lindblom 1953; Perry and Rainey 1988; Rainey and Bozeman 2000; Scott and Falcone 1998). Based on these responses, fewer than 19 percent of the participating organizations could to some extent be characterized as hybrid; the majority of the sample consisted of core public and core private organizations (Coursey and Bozeman 1990). However, to assess the influence of the participating organizations’ publicness, and thus specify their sectoral characteristics, a new variable ‘publicness’ was computed (Bozeman and Bretschneider 1994; Coursey and Bozeman 1990; Rainey and Bozeman 2000). To characterize the management level and work experience of the respondents, they were also asked about the number of employees they supervised; the number of years they had been working for the present organization; and their age, gender and work experience in another sector (yes or no). These control variables were included to monitor what effect(s) other variables besides publicness might have on the perception of organizational values. Boyne (2002: 104-105) rightly underscores the importance of this aspect: Few empirical tests take account of any other influences on the organizational variables [i.e., values]. This simple methodology takes the view that “publicness” matters to extremes: it is tantamount to the assumption that only publicness matters. This assumption is highly implausible. For example, organizational environments may differ between industries and geographical locations, yet tests of publicness rarely hold such variables constant. Similarly, although structural characteristics of organizations are likely to vary with their size, and the values of managers may vary with their age, sex and prior professional experience, such variables are seldom taken into account.
When research involves concepts such as values, with which individuals inherently have normative associations, the issue of social desirability becomes important (cf. Fowler 2003). For example, when respondents are asked to rate each value in a list of twenty from 1 to 10, they might feel inclined to declare selfless and altruistic values like ‘social justice’ and ‘honesty,’ together with a number of other values, to be very important in decision-making situations. However, in real-life situations, it might be very difficult, if not impossible, to actualize these values in each and every decision. Thus, as Posner and Schmidt (1986: 448) point out, “responses to a questionnaire may not correspond exactly
44
with how people behave; questionnaire responses are likely to be more positive and idealistic than behavioral responses which occur when managers feel under pressure, confronted with conflicting information and competing loyalties.” Consequently, respondents’ selection of a specific number of listed values and their ranking of values in order of importance may provide some insight into what is truly valued most, but the picture may be incomplete in that a large number of values may be seldom – or never – mentioned as being among the most important. Clearly, both methods have advantages and disadvantages (Agle and Caldwell 1999: 367–368). For instance, advocates of rating state that, in actual decision making, agents attribute equal importance to several different values at once without being aware of possible conflicts between them (Hitlin and Pavilian 2004; Schwartz 1999). Thus, making such conflicts transparent (with the assistance of additional statistical analyses) is an interesting aspect of the rating method. In general, rating is also easier to analyze statistically than ranking. One disadvantage, however, is that rating each value produces a more general constructed hierarchy because differences between values, as well as between organizations and sectors, become marginalized and respondents are not obliged to choose what is really valued most in the case of conflictual situations (Rokeach 1973). It is precisely because of these pros and cons that this research applies a within-method triangulation (Denzin 1970) using a questionnaire that contains both rating and ranking questions. In addition, respondents were asked to rank not only the five most important ‘actual’ values, but also the five values that ‘should be’ most important (cf. Lawton 1998). Thus, the divide between fact and norm was deliberately enlarged so that more insight could be gained into possible social desirability biases. Between March and May 2005, following a pretest of 16 public and private sector managers from a comparable population, the questionnaires were distributed among 778 (effectively, 766) managers of government organizations (response rate: 30.16 percent) and 500 (effectively, 497) managers of business organizations (response rate: 30.44 percent).37 Being comparable to the response rates of earlier mail surveys of top managers and considerably higher than those for most business surveys (e.g., Posner and Schmidt 1984, 1993, 1996), these response rates, although somewhat lower than those for most public sector surveys (e.g., Bowman and Williams 1997; Goss 2003; Kim 2001; van den Heuvel et al. 2002), were considered respectable given the type of respondent. Initially, the response from the private sector respondents, at about 15 percent, was disappointing. However, after an initial analysis of non-response among the private sector respondents through telephone inventory (cf. Fowler 2003), a second group of questionnaires together with reminders was sent to a specific portion of the population that had agreed to participate but had lost the questionnaire or simply forgotten about it owing to professional obligations. Within a number of weeks, the response then doubled to 30.44 percent. For the public sector, the response was immediately higher and, based on the private sector experience, public executives were given an extra two weeks to respond.
45
Table 3.2: Survey respondents’ most important characteristics Public sector (n = 231)
Private sector (n = 151)
Age: 26–35 36–45 46–55 56 and older
0% [1%] 20% [19%] 55% [51%] 25% [29%]
1% [3%] 17% [18%] 41% [37%] 41% [31%]
Gender M: F:
85% [85%] 15% [15%]
97% [94%] 3% [6%]
Number of employees supervised: 500
56% 27% 17%
36% 27% 37%
Working at present organization: 10 years
6% 31% 9% 54%
4% 24% 17% 55%
n/a
4259
33%
29%
Average number of employees in entire organization Worked in other sector
The questionnaires were distributed in cooperation with two professional associations: the Senior Public Service (ABD in Dutch) and the Dutch Centre of Executive and Non-executive Directors (NCD in Dutch). ABD, the professional association of the top management group of the Dutch federal government, maintains a database of almost 800 heads of directorates, departments and agencies that automatically become members upon reaching a certain hierarchical and salary level. NCD is a professional association of 4,500 executives and non-executive board members of small, medium and large companies in various fields (predominantly finance, consultancy, industry, law and infrastructure), with voluntary membership. The ABD sample included all members, but the NCD sample comprised 500 managers chosen randomly, 400 managers of companies with at least 50 (but fewer than 1,000) employees and
46
100 managers of companies with at least 1,000 employees. This particular NCD sampling was designed to achieve the best comparability with the ABD members, who ranged from bureau chiefs supervising a few dozen employees to department heads supervising up to 30,000 employees. Top managers were explicitly chosen as a research population to avoid a confounded relation with occupational type. That is, to achieve a certain level of homogeneity that excludes a number of external factors that might account for too much variance. Moreover, top managers can be expected to have more influence on the workings of public and private organizations, and a more informed view on decision-making processes than, for example, the average street-level bureaucrat or call-center receptionist. Therefore, this group is of particular interest to the central research question. Admittedly, the distribution and variance of ABD respondents among the different departments was not perfect, but the ABD sample appeared sufficiently valid and representative of Dutch federal public sector executives (all differences: less than 5 percent). With regard to gender and age, the sample closely resembled the population, as shown in Table 3.2, which outlines the most important respondent characteristics and gives population data – only available for gender and age – between brackets. The final sample consisted of 382 fully completed and usable questionnaires. 3.4.3
Phase III: Face-to-face in-depth structured interviews
The first step in the qualitative research phase was the construction of an interview guide, “a listing of areas to be covered in the interview along with, for each area, a listing of topics or questions that together will suggest lines of inquiry” (Weiss 1994: 48). Following an introductory talk in which the interviewer laid out the purpose of the research project and interview, each respondent was asked 17 central questions. Special attention was paid to the order of the topics and the questions (see Appendix B for the full interview guide). The qualitative research method was a structured or standardized interview, which “consists of a set of questions carefully worded and arranged for the purpose of taking each respondent through the same sequence, and asking each respondent the same questions with essentially the same words” . Thus, each respondent was asked the same questions during an in-depth conversation that lasted between 45 and 65 minutes, depending on time availability and the progress being made during the interview. Each interview began with a few short questions on career history, present function and organizational characteristics. The respondent was then asked to describe in detail a recent decision he or she had been part of, the deliberations and considerations involved and the resultant choice of a certain direction. This technique was designed to establish an indirect link between decision-making conduct and value prioritization (cf. van Rekom et al. 2006) and initially probe the significance of the sectoral decision-making context. The interview core addressed the role and importance of several specific values or pairs of values in
47
this and other important decisions in general, and the way in which these values were either in conflict or strongly related in certain decision-making contexts. Respondents were also asked whether the same set of values applied to every decision or whether every decision implied a different hierarchy and a different value mix, what they thought of differences and similarities to other sectors, and to what extent a strong, shared value system (a shared view on ‘what to be and where to go’) existed within their organizations. Although many respondents initially considered the role of values in organizational decision making abstract, most were able to make transparent – for themselves and for the interviewer – what was really valued most during important decisions. They also identified the conflicting, contradictory and sometimes problematic tradeoffs in the prioritization and actualization of certain values. The interviewer-interviewee partnership, also known as the interviewing relationship, is of great importance to the success of in-depth interviews. For the purpose of this study, the following research perspective (Weiss 1994: 65-66) was applied to this interaction: Interviewer and respondent work together to produce information useful to the research project; the respondent will provide observations, external and internal, accepting the interviewer’s guidance regarding topics and the kind of report that is needed; the interviewer will not question the respondent’s appraisals, choices, motives, right to observations, or personal worth, and will ensure that the respondent will not be damaged or disadvantaged because of the respondent’s participation in the interview, in particular the interviewer will treat the respondent’s participation and communications as confidential information.
Nevertheless, on occasion, the role of “respectful student awaiting instruction” (Weiss 1994: 66) was applied, because in some cases acknowledging the professional status of the respondents helped to disarm them so that they would disclose relevant information and account truthfully for reallife events. Rather than selecting respondents randomly on the basis of probability parameters, as is the case in a quantitative approach, the selection of respondents or cases in a qualitative context aims at maximizing range and depth (cf. Weiss 1994: 23). It seemed clear that, if the results of both research phases were to be combined and integrated, the type of respondent and organization that was to participate in the interview phase had to resemble those that responded to the survey. Therefore, not only did the sample have to include a wide range of companies as well as a number of federal government organizations, respondents again had to be top-level managers. In addition, a deliberate choice was made to include a number of so-called parapublic organizations (Lyons et al. 2006) or quangos (Van Thiel 2000), such as hospitals, schools, universities and contract agencies (Agentschappen), which operate somewhat autonomously from the public core and within a market-like financial budgeting regime. This inclusion was based on the expectation that such organizations more often encounter confrontational and conflicting public and private sector demands,
48
interests, techniques, norms and values within their decision-making processes. The participating organizations together with the (sub)sectors to which they belong are shown in Table 3.3 (see Appendix C for a full list of participating organizations and managers). Table 3.3: Organizations participating in the interview phase (n = 38) Public sector
Government agencies
Parapublic organizations
Private sector
Provincial Government of Noord-Holland (n = 5)
National Agency of Correctional Institutions (n = 5)38
VU University (n = 4)
Organon (n = 5)
Ministry of Health, Welfare and Sport (n = 5)
Agency for Sale of Used Government Goods (n = 1)39
Zaans Medical Centre (n = 1)
Rabobank (n = 1)
Ministry of Finance (n = 4)
Royal NIVRA (n = 1)
Ministry of the Interior (n = 1) 15
KPMG (n = 1) Unilever (n = 1) DLA Piper (n = 1) Shell (n = 1) Van Doorne (n = 1)
6
6
11
Besides attempting to maximize range, the sampling was also infused by a certain degree of pragmatism: the research group and the researcher’s network were used to recruit interesting and relevant respondents. Although convenience sampling may not be the ideal base for generalization (Weiss 1994: 26), good reasons existed for using this technique to a certain extent: (i) the respondents’ own assessment of generalizability, which is strongly related to snowball sampling; (ii) the interviewer’s own identification of others worth recruiting, and (iii) the depth of the studied phenomenon itself (the idea that an identity in structure and functioning, a certain amount of universalism with regard to the phenomenon studied, exists among a certain group of respondents. For these reasons, some organizations were represented by several manager participants but others by only one (in some cases, the highest ranking general manager or CEO). Just as was the case for the survey respondents, the majority of interview participants were male (87 percent), averagely aged above 45 (M = 50 years old) and holding a senior management position, supervising from a few dozen up to thousands of employees. Although organizational, rather than managerial, values were the subject of study, the overall objective was to paint a broad picture of values in
49
organizational decision making rather than doing case studies of particular organizations. Almost all managers participating in the interviews also filled out the questionnaire, so that a match could be made between the value orientations of both populations. In the end, a total of 38 in-depth interviews were conducted. 3.5
Data analysis and reporting
The data analysis aimed to draw inferences on the general – organizational and sectoral – rather than the individual level. Single respondents and cases were, therefore, less important than the overall issue: the dependent variables that represented the most important values in organizational decision making. Thus, even though out of necessity individual participants were surveyed and interviewed in this study, making them the subjects of research, the objects of analysis were the aggregate values that play a role in organizational decision making. The empirical results are presented in the next four chapters. The remainder of this chapter first provides a very concise overview of the analysis and reporting of the quantitative data, followed by a somewhat more elaborate discussion of the analysis and reporting of the interview data. The analysis of the survey data, done using SPSS 15.0, employed various, although mostly standard multivariate, techniques to investigate what was valued most. First, it compared the means (M) for the value ratings and created a sum of scores ( ) for the value rankings (taking into account both frequencies and ranks), after which a factor analysis investigated whether certain clusters or dimensions could be identified within the more general set of 20 values. An independent sample t-test and a multivariate test of differences using partial eta square ( 2) then assessed the significance of differences between the organizations. Additional correlational analyses tested whether specific values were (strongly) positively or negatively related to each other. Following Boyne’s (2002: 118) suggestion that “publicness variables should be tested in multivariate statistical models that control for other variables, such as the external and internal features of organizations and the personal characteristics of respondents,” multiple regression and correlational analyses were employed to measure the predictability of publicness. These techniques also assessed the control variables for value preferences in participating organizations and the relative effect of these variables on one another. The same techniques were used to observe possible relations between specific values and specific integrity and decision-making problems. The primary objective of this study is to portray public and private value patterns on an organizational and sectoral rather than individual level. Therefore, the data analysis was issue focused rather than case focused, as is the case in a multiple case-study design aimed at formulating research propositions, and took place at the “level of the generalized” rather than the “level of the concrete” (Weiss 1994: 152). Consequently, the participating organizations were not studied as distinct cases, but rather the statements on the values in organizational
50
decision making constituted the locus of analysis.40 The aim of issue-focused analysis is “to describe what has been learned from all respondents about people in their situation” (153); in other words, to paint a general but at the same time contextual picture. Likewise, analysis of qualitative data involves analytic generalization rather then statistical generalization (Yin 2003). Thus, instead of generalizing the results to a larger population and testing the theoretical propositions, the aim here was to inform existing theory with new insights. Although clearly not a case-study analysis as such, this research shares many similarities with a multiple-case study analysis in terms of initial data coding and sorting. The logical choice for data analysis was a “retrospective comparison of cases”, an in-depth analysis of a large set of aspects (organizational decisions, organizational values and a number of related issues) in a number of cases (the 38 respondents). According to Eisenhardt (1989), the advantage of such a design is that it allows the researcher to recognize general patterns in different settings. However, its disadvantage is that every case with its own context and contingencies must be reduced to a more abstract level to enable between-case comparisons . Despite Weiss’s (1994: 154) contention that the coding starts as soon as the first interview report is written, the initial coding of the interview material for this study began after completion of the first series of 10 interviews because of a short hiatus between this series and the 28 interviews that followed. Because 5 of these initial 10 interviews were conducted in a public sector organization and 5 in a private sector organization, the initial coding was not disproportionately biased by either of the two sectors. As Strauss (1987), Weiss (1994) and Miles and Huberman (1994) rightly argue, data analysis is not simply a question of retrospective comparison of cases. Rather, data analysis begins as soon as there is data collection. Indeed, as Miles and Huberman (1994: 49) observe, “the more investigators have developed understandings during data collection, the surer they can be of the adequacy of the data collection and the less daunting will be the task of fully analyzing the data.” The consequence of such a research strategy for the present study was that typing out every interview as seven to eight pages of text resulted in immense quantities of data (over 250 pages of literal transcriptions) that needed to be systematically analyzed. Following the suggestion of the above researchers, coding of these literal transcriptions began with a monster-grid – a data matrix created in Excel with the respondents on one axis and the seventeen interview questions on the other – that can be perceived as a more elaborate version of what Weiss (1994: 157) calls “creating excerpt files.” Thus, the grid cells were filled not with numbers but with various “verbal comments and citations” from the interviews. Consistent with explorative research, the option of insights and novel findings based on other variables emerging from the data was left open. Following Bijlsma-Frankema and Drooglever Fortuijn (1997), the analysis included either telling citations from respondent answers or written summaries of the answers on a particular theme that adhered as closely as possible to the respondents’ own words. The next step involved reading all the responses to a
51
particular theme to derive first impressions of overall patterns that were then juxtaposed with the empirical data. This inductive process, described by Weiss (1994: 158) as “local integration,” is clearly not just a matter of counting. After all, besides the fact that respondents had not been randomly selected and that 38 interviews and 17 organizations are, for quantitative purposes, too small a number, the goal of this explorative phase was to consider the nuances and context of every case. Thus, it not only mattered that a respondent considered a certain value important and its usage different from that in the other sector, it was equally important to observe what and how strong that importance was and how the respondent worded the differences. As a result, the inductive analytical process was repeated many times before the final analysis was written. Following this initial interpretation using the monster grid, the 38 interviews were converted to text documents and imported as separate ‘hermeneutic units’ into Atlas.ti 5.0, a widely used software tool for coding qualitative data in a structured manner. Each core statement or quotation of about 10–20 lines on the importance of specific values received a label or an initial ‘open code,’ one structured sentence that summarized and characterized the statement’s core. All structures were coded except for the few that had no specific relevance for answering the research question (cf. Klostermann 2003: 43). Particular attention was paid to how, when and to what extent the value was important in organizational decision making. During the process of coding and sorting and going back and forth between data and codes, more definitive codes were gradually established as new codes were created or old ones adapted (43). Appendix D provides an example of the way in which the importance of the value ‘responsiveness’ was iteratively coded for the first five respondents. Not only does the chronological order of an interview (i.e., the interview guide) structure the coding process, it also influences the way in which the story is told in the final report both in terms of chronology and content (Weiss 1994: 153). Thus, to grasp the importance of certain values on a subconscious level and the significance of the sectoral decision-making context, the qualitative analysis focused particularly on the decisions, deliberations and considerations accompanying decision making. Next, based on the most relevant quotations (or parts of them) from participant responses, it characterized each value or pair of values explicitly addressed in the interview in terms of the way it seemed to be important in the decision-making process. Finally, all such observations were combined as a cogently written report; that is, “as a coherent story, so that the material presented early in the report prepares the reader for material that will appear later and later material draws on the earlier, and the reader in the end can grasp the report entire” (153).
52
Part II: Empirical Results
53
54
4.
A Public-Private Value Panorama
This chapter discusses the results of a literature review and content analysis of the most prominent organizational values in the administrative and business ethics literature, public and private sector codes of conduct, and empirical research on values. The analysis culminates in the construction of a publicprivate value panorama and a value set for use as a tool in the empirical research. In addition, based upon the literature and empirical research, it identifies a number of integrity violations and decision-making inefficiencies that are then made operational. 4.1
Organizational values in administrative ethics
For the public sector, the review included seven relevant books on administrative ethics and on values, in particular: Cooper (1998, 2001), Heidenheimer, Johnston and Levine (1989), Lawton (1998), Sampford and Preston (1998), Van Wart (1998), and Williams and Doig (2000), as well as 46 issues of Public Integrity (1999-2003) and the Public Administration Review (1999-2002).41 Of 544 values belonging to 294 different types, clustering of the extensive number of value statements – using 15 different options of combinations of variables that have been described in 3.4.1 – resulted in 30 that were the most prominent (see Table 4.1 below). Table 4.1: Cluster results for the administrative ethics literature Cluster 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15.
Honesty Humaneness Social justice Impartiality Transparency Integrity Obedience Reliability Responsibility Expertise Accountability Efficiency Courage Prudence Serviceability
Total 434 422 402 380 379 365 357 329 327 314 294 276 254 220 215
16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30.
55
Cooperativeness Responsiveness Dedication Effectiveness Innovativeness Lawfulness Loyalty Consistency Autonomy Stability Representativeness Competitiveness Profitability Collegiality Self-fulfillment
191 184 183 181 179 152 146 111 99 99 88 77 59 48 16
Ultimately, 13 of these 30, shown in bold in the same Table, were determined to be the most relevant for public sector organizations. Not only were these values mentioned most often in the literature, they were also in the highest regions of the different clustering combinations, meaning that they were most often empirically justified and explicitly mentioned as ‘public,’ ‘organizational’ and ‘(very) important’ in the studies that were part of the analysis. As a single value, ‘impartiality’ was, for instance, mentioned more often than ‘honesty’ and ‘social justice,’ but because it appeared in less combinations of clustering variables, it ended up lower in the final hierarchy. This set of values shows some remarkable characteristics; for example, ‘courage,’ arguably a traditional business value, was among the 13 most prominent values, whereas ‘lawfulness,’ an often emphasized public sector value, was absent (although, as Table 4.1 shows, it was present in the top 30 of the most important public sector values). Thus, it might be asked whether the fairly recent literature has already been ‘infected’ by NPM developments. For example, Beck Jørgensen and Bozeman (2007: 357), who encounter similar findings in their study on public values, suggest that a large proportion of the literature may be very much of its time: “In particular, much of the literature praises recent reforms such as New Public Management (NPM) and Reinventing Government. However, there is an emerging literature that, as a reaction, praises the old virtues of classic administration or, alternatively, launches new progressive models such as ‘new public governance’ or ‘new public service’.” Hence, the set extrapolated for this study seems to be more a valid representation of prominent values in recent administrative ethics literature than a valid representation of relevant public sector organizational values. More specifically, it relates to the overrepresentation in this literature of articles and essays with a more normative and prescriptive character than empirically informed writings. Perhaps an investigation of more practitioner-oriented sets of public sector values and specific empirical research on codes of conduct and civil servants’ value orientations can paint a different and complementary picture. Among these, one well-known moral framework for public officials was developed in the United Kingdom by the Committee on Standards in Public Life chaired by Lord Nolan, which sketched the following Seven Principles of Public Life (1995). First, holders of public office should make decisions based on the public interest: private interests or obligations to outside individuals and organizations should have no influence (‘selflessness’ and ‘integrity’). Similarly, they should make choices on the basis of merit (‘objectivity’), be accountable for their decisions and actions (‘accountability’) and be as open as possible (‘openness’). Holders of public office also have a duty to declare any private interests and resolve possible conflicts of interest (‘honesty’). Finally, they should promote and support these principles through leadership and example (‘leadership’). Many of these values have also been used in other contexts; for instance, by the Committee of Independent Experts (1999), which reported on fraud, mismanagement and nepotism in the European Commission; in the Queensland
56
Public Sector Ethics Act; and in codes of conduct such as the United Nations’ International Code of Conduct for Public Officials (1996: A/RES/51/59). The UN code contains a set of basic standards of integrity and performance expected from public officials. The term ‘public officials’ is deemed to include all persons vested with the power and authority to make, implement, enforce, amend or revoke government decisions and to render services to the public, with or without remuneration. The code includes the following general principles: public officials shall act in the public interests; function efficiently and effectively, in accordance with the law and with integrity; and shall be attentive, fair and impartial. The code also contains rules pertaining to “conflicts of interest and disqualification, disclosure of assets, acceptance of gifts and other favors, confidential information and political activity” (Pieth and Eigen 1999: 665–666). The values mentioned in the research on the 59 codes of conduct for Dutch government organizations (Ethicon 2003 (see Table 4.2)) show some similarities to the Nolan Committee principles (moreover, 3 of the 7 principles appear in the list of prominent values, while the other 4 principles resemble some of the values mentioned). Somewhat striking is the absence of ‘accountability’ and ‘lawfulness’ in the Dutch public sector codes, although ‘responsibility’ and ‘serviceability’ are to some extent related in meaning. Table 4.2: Core values in the Dutch public sector codes of conduct (Ethicon 2003) Core values (translated from the Dutch) 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13.
Percent of codes that mention the value (n = 59)
Integrity Transparency/openness Responsibility Trust/trustworthiness Carefulness Independence Reliability Professionalism Restraint Functionality Credibility Honesty Serviceability
98 88 78 76 76 75 68 44 39 36 31 27 24
When the academic literature value set of the 13 most prominent values for the public sector (bolded in Table 4.1) is compared with the Nolan principles and the core values in the Dutch public sector codes of conduct, a number of similarities and differences become obvious.42 Whereas six values are exactly represented in both the academic literature value set and in the Nolan principles and Dutch conduct codes (‘honesty,’ ‘transparency,’ ‘integrity,’ ‘reliability,’
57
‘responsibility’ and ‘accountability’), other values not identical but to some extent related in meaning (‘impartiality’ ‘independence’ and ‘objectivity,’ ‘expertise’ ‘professionalism’) are present in both sets. However, it should be noted that the codes do not even mention ‘efficiency’ and ‘effectiveness.’ Another somewhat surprising similarity is the absence from both sets of ‘lawfulness,’ arguably a crucial public sector value. Perhaps this value is considered so self-evident to public sector organizations and employees that making it explicit is unnecessary. One interesting difference involves the presence of ‘humaneness’ and ‘social justice,’ two values with a strong social character, in the list derived from the public administration literature but the absence of both in the public sector codes. One explanation might be that these values are considered crucial and important within an academic context but do not apply to day-to-day organizational reality. Equally interesting is the presence of ‘obedience’ and ‘efficiency’ in the academic set and their absence in the codes. However, as Lawton (1998) has already shown, obedience is an important organizational value in practice but, because of its somewhat archaic and hierarchical connotation, is not considered something to strive for and thus not a value commonly mentioned in codes of conduct.43 In 1999 and 2000, van den Heuvel et al. (2002) conducted a broad and extensive survey on the values and perceptions of integrity of Dutch politicians and civil servants (n = 678) working in assorted positions in various policy fields within the different administrative areas of the Dutch government (local, regional, federal). Many of the civil servants considered ‘expertise’ very important for proper fulfillment of professional obligations, followed at some distance by ‘honesty,’ ‘efficiency,’ ‘lawfulness,’ ‘serviceability’ and ‘dedication’ (see Table 4.3, in which those six values that received the highest scores from civil servants are bolded). The review for this present study of the international literature and the Dutch research on codes of conduct and civil servants’ value orientations paints a rather uniform and sometimes rather ideal image of public sector value preferences. After a similar review for the business sector, the different sources of public and private sector value preferences are juxtaposed to construct the most relevant and feasible set of values for the empirical research.
58
Table 4.3: Most important values for Dutch politicians and civil servants (van den Heuvel et al. 2002) Politicians Civil servants (value among (value among three most three most important for important for own behavior own behavior in %) in %)
Value
Acceptability (acting in accordance with public wishes ) Collegiality (acting loyally and in solidarity with colleagues) Expertise (Acting with competence, skill and knowledge Serviceability (being respectful and helpful towards the public) Efficiency (attaining goals at the least possible costs) Honesty (being truthful and keeping promises) Obedience (complying with policies and superiors’ instructions) Impartiality (acting in a way unbiased by specific group interests) Integrity (acting in a way unbiased by self-interest) Openness (acting openly, transparently and without secrecy) Profitability (acting to achieve political or bureaucratic gain) Lawfulness (acting in accordance with laws, rules and regulations) Dedication (performing tasks with diligence and perseverance)
20
4
5
19
26
66
17
29
14 60 0
31 35 1
32
18
39 39
20 21
1
1
22
31
24
26
4.2 Organizational values in business ethics An extensive review of the business ethics literature directed at prominent private sector values included the following 11 books: Bird (1996), Boatright (2000), Bok (1978), Bowie (1992), Donaldson and Dunfee (1999), Gauthier (1986), Kimman (1991), Montefiore and Vines (1999), Nash (1990), Quinn and Davies (1999), and Wempe and Nelis (1991). Also included were 25 issues of three journals: the Journal of Business Ethics (1999–2001), Business Ethics Quarterly (2000–2001) and Business and Society (2001), as well as Kaptein’s (2004) research on codes of conduct (as previously mentioned, the analysis did not include government documents). These sources yielded 210 values belonging to 142 different value types, from which the most relevant 13 were
59
ultimately selected. Table 4.4 lists the most prominent organizational values for the private sector. Table 4.4: Most prominent private sector organizational values Private sector organizational values 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13.
Honesty Social responsibility Customer orientation Innovativeness Accountability Self-fulfillment Expertise Effectiveness Reliability Profitability Collegiality Sustainability Entrepreneurship
Yet again, this set of values exhibits some surprising characteristics; specifically, the business bottom line of ‘profitability,’ still touted by neoclassical economists and corporations like Shell as the most important business value (see www.shell.com), is in 10th position, while the fashionable ‘social responsibility’ ranks second. This outcome might have been expected given that the literature studied was from the field of business ethics and not from more standard economic or business literature. Further examination of the top 13 values shows that – as in the public sector – ‘honesty’ is ranked as the most prominent value. Also present in both the public and private sector value set are ‘accountability,’ ‘expertise,’ ‘effectiveness’ and ‘reliability,’ which are apparently important in both government and business sector organizations, at least according to the literature studied. Most surprising is the absence of ‘integrity.’ Kaptein (2004: 22), after analyzing the business codes of the 200 largest multinationals worldwide (of whom 52.5 percent had a code), provided a summary of the most frequently mentioned values (see Table 4.5 for the top 13), whose somewhat fragmentized character he attributes to cultural differences: The core values cited in the codes diverge strongly. American codes make comparatively less mention of values than do European codes. Open communication can be found in 35% of the European codes, while it amounts to 25% in American codes. European codes mention teamwork 1.5 times more often than do Japanese and American codes. ‘Humility,’ ‘harmony,’ ‘dedication,’ ‘innovation,’ ‘creativity’ and ‘team spirit’ are largely found in
60
business codes of Asian origin. ‘Innovation’ and ‘creativity,’ for example, are respectively mentioned 73% and 46% more in Asian codes than in American and European codes. Noteworthy is that the value [of] ‘effectiveness’ is seldom mentioned explicitly. Table 4.5: Core values in the business codes of the 200 largest multinationals (Kaptein 2004) Percent of codes mentioning the value (n = 105)
Core values
1. 2. 3. 5. 6. 7. 8. 11. 12. 13.
Teamwork/mutual support/cooperation/ team spirit Responsibility/conscientiousness (Open) Communication Innovation, creativity Customer orientation Flexibility Efficiency Professionalism Entrepreneurship Pride/dignity Loyalty Motivation/enthusiasm/energy/spirit/encouragement Participation
43 33 29 29 19 17 16 14 14 14 13 12 11
When the academic value set is compared with the core values of multinational corporations, a number of differences and similarities emerge. For example, ‘integrity’ – mentioned so often in public sector codes, the administrative ethics literature and Dutch-based multinationals’ codes of conduct (see e.g., www.shell.com and www.abnamro.com) – is once again absent. Any explanation for this absence would be merely speculative. However, three values do appear in both the codes and the value set (‘customer orientation,’ ‘innovativeness,’ and ‘entrepreneurship’). Perhaps the most interesting differences involve the academic research value set’s emphasis on the customer and individual versus the business codes’ emphasis on ‘teamwork’ and ‘cooperation.’ The absence of ‘responsibility’ and ‘responsiveness’ in the academic research set is surprising, since responsibility is often mentioned as an important principle or value in organizational literature. Moreover, although the term ‘profitability’ is often prominent in business principles and mission statements, sometimes phrased in terms of financial continuity or stability, it is absent from the business codes list of most frequently mentioned values. Perhaps it is simply more appealing to include values that seem to have a strong moral connotation than to include the most prominent, but also self-evident, business value. It must be noted, however, that this business value is in tenth position in Table 4.4, and thus is not given prominence in the literature either. Not surprisingly, obedience as a value seems
61
minimally associated with modern-day business discourse, whereas loyalty is among the oft-mentioned values in international business codes. 4.3
Constructing a public-private sector value panorama
Before a comparison can be made between the most prominent business and government organizational values, it should again be noted that identical values may be important in different ways, and with regard to different aspects, in a public and private context. For instance, ‘social responsibility’ probably denotes something different in the private sector than ‘responsibility’ in the public sector. Similarly, ‘accountability’ is a hydra-headed concept with multiple pathologies (e.g., Koppel 2005), including the distinction between ‘process accountability’ for the public sector and ‘output accountability’ for the private sector (cf. Erkillä 2007). It may be that the qualitative research will shed more light on this phenomenon. The identical organizational values distinguishable in Tables 4.1 and 4.4 – ‘honesty,’ ‘accountability,’ ‘expertise,’ ‘reliability’ and ‘(social) responsibility’ – might be viewed as common core organizational values. The other 16 values can be considered more or less public or more or less private. A closer look at their position in the broader framework of values, while taking into account the 30 most relevant values distinguished for both sectors (see Table 4.1 for the public sector), can determine whether they possess a more private or more public character. For instance, the prominent private sector values ‘profitability,’ ‘collegiality,’ ‘innovativeness,’ ‘self-fulfillment’ and ‘effectiveness’ were absent from the list of the 13 most prominent public sector values but present in the list of 30. Moreover, the order in which they were ranked (e.g., number 30 is closer to the private pole than number 28) suggests that they might be considered common core values somewhat more to the private end of the spectrum. Because values derived from empirical research are considered at least as relevant as values derived from recent literature, the present analysis also takes into account the empirical research. According to the latter, values absent from those most frequently mentioned in the literature and codes – such as ‘lawfulness,’ ‘dedication’ and ‘responsiveness’ – were considered important by politicians and public managers.44 Therefore, to construct a feasible value set for this current research, the two sets of values were weighted, integrated and supplemented when necessary, although, of course, such an exercise is always somewhat arbitrary. The construction process included the following steps. First, ‘integrity’ – not a separate value in the terminology as described in chapter one, but a sign that individuals and organizations are acting in accordance with all the relevant sets of values (and norms and rules) – was omitted. However, its content as a specific value is partly characterized by the remaining values of ‘impartiality’ and ‘honesty.’ One other value, ‘incorruptibility,’ was added to characterize the individual dismissal of selfinterest.
62
Second, a number of values that are strongly related in meaning were combined. For example, the more general social values of ‘humaneness,’ ‘social justice’ and ‘social responsibility,’ which all refer to the commitment to a just and humane society, were combined as ‘social justice.’ Conversely, for the private sector, ‘entrepreneurship,’ which is closely related to being able and willing to take risks, innovate (‘innovativeness’) and seek opportunities for profit (‘profitability’), was weeded out. Third, serious reflection was given to whether ‘courage’ is a value connected with individual character or whether it can be characterized as an organizational value. On the one hand, courage concerns the willingness in difficult circumstances to act with determination in accordance with values. Thus, it contrasts with opportunistic behavior but comes close to ‘reliability.’ On the other hand, courage involves the extent to which employees and organizations are willing to take and face risks. Perceived as such, it closely resembles ‘innovativeness.’ Thus, ‘courage’ was also omitted. Finally, it was asked whether the review of the literature might not have led to the omission of certain traditional business or public values that are worth incorporating into future research. For example, ‘lawfulness’ although strikingly absent from both the public sector codes of conduct and the most relevant public sector values in the literature, was valued quite highly by politicians and civil servants (Lawton 1998; van den Heuvel et al. 2002). Nor is lawfulness the only such problematic omission. For instance, ‘dedication’ can be considered an important organizational value for both the public and private sector, while ‘serviceability,’ and to a certain extent ‘responsiveness,’ can be considered the public counterparts of the private sector value of ‘customer orientation.’ Thus, to present respondents with a more neutral value set, ‘customer orientation,’ with its strong private sector connotation, was replaced by ‘serviceability.’ Not only did this neutrality invigorate the decision to also omit ‘entrepreneurship,’ it was expected to produce more interesting and more nuanced empirical results.45 As a result, the values ‘lawfulness,’ ‘dedication,’ ‘responsiveness’ and ‘serviceability’ were added to the set. This integration and supplementation enabled the construction of a ‘publicprivate value panorama’ (Figure 4.1), a continuum that attributes organizational values to both sectors with common core organizational values as the fulcrum.
63
Figure 4.1 Panorama of organizational values with a public and a private sector pole
PUBLIC Social Justice Impartiality Transparency Obedience Efficiency Lawfulness Dedication Serviceability Incorruptibility Responsiveness
Honesty Accountability Expertise Reliability Effectiveness Self-fulfillment Innovativeness Collegiality Profitability Sustainability
PRIVATE
64
4.4
The value set as a survey research tool
The set of 20 organizational values (presented in Table 4.6), intended as a feasible value survey research tool, consists of a balanced mix between characteristic public and private sector values and values that might be considered part of a common core of organizational qualities and standards. Definitions were drawn from the literature if present (46 percent of such values were clearly defined) and, because each value presupposes a certain mode of organizational conduct, set up in accordance with the active definition of the main construct, ‘having a certain weight in the choice of action.’ Table 4.6: Mixed set of public, private and common core organizational values Set of organizational values Accountability: Acting willingly to justify/explain actions to relevant stakeholders Collegiality: Acting loyally and showing solidarity towards colleagues Dedication: Acting with diligence, enthusiasm and perseverance Effectiveness: Acting to achieve the desired results Efficiency: Acting to achieve results with minimal means Expertise: Acting with competence, skill and knowledge Honesty: Acting truthfully and comply with promises Impartiality: Acting without prejudice or bias toward specific group interests Incorruptibility: Acting without prejudice and bias toward private interests Innovativeness: Acting with initiative and creativity (to invent or introduce new policies or products) Lawfulness: Acting in accordance with existing laws and rules Obedience: Acting in compliance with the instructions and policies (of superiors and the organization) Profitability: Acting to achieve gain (financial or other) Reliability: Acting in a trustworthy and consistent way toward relevant stakeholders Responsiveness: Acting in accordance with the preferences of citizens and customers Self-fulfillment: Acting to stimulate the (professional) development and well-being of employees Serviceability: Acting helpfully and offer quality and service towards citizens and customers Social justice: Acting out of commitment to a just society Sustainability: Acting out of commitment to nature and the environment Transparency: Acting openly, visibly and controllably
As argued in the discussion of central research concepts, such clear delineation is of particular importance in research on values; for instance, de Vries (2002) shows convincingly that civil servants within different administrative cultures interpret a seemingly universal concept like ‘honesty’ very differently. The current task of empirical research is to show whether distinct public and private value patterns are distinguishable or whether the most
65
important organizational values are shared among government and business organizations, with similarities exceeding differences. To this end and based upon the value set created, the following hypotheses are formulated: H1
In public sector organizations, dedication, efficiency, impartiality, incorruptibility, lawfulness, obedience, responsiveness, serviceability, social justice, and transparency are considered more important than in private sector organizations.
H2
In private sector organizations, collegiality, effectiveness, innovativeness, profitability, self-fulfillment and sustainability are considered more important than in public sector organizations.
H3
In both public and private sector organizations, accountability, expertise, honesty and reliability are valued highly and can be considered common core organizational values.
These hypotheses presuppose what might be termed ‘value solidity.’ Specifically, they assume that the main values in the public and the business sector did remain relatively stable (in accordance with the public-private panorama) and that less fluidity occurs between sectors than NPM and CSR might suggest. That is, despite these latter developments’ influence on the administrative and business ethics literature, their tenets may represent what ‘ought to be’ rather than what ‘is.’ 4.5
Value intermixing, and integrity and decision making problems
The current investigation will also test the assumptions put forward by Frederickson (2005), Friedman (1970), Jacobs (1992) and Schultz (2004), which in combination imply that the intermixing of sector specific values may not only raise vexing ethical questions but may also lead to moral and functional problems (in particular, an increase in corruption and unethical behavior in public sector organizations, and inefficiency and waste in private sector organizations). To do so, it will identify a number of specific integrity violations and decision-making inefficiencies and then assess whether a relation exists between certain values and the perception of certain problems and whether specific values or sectoral characteristics – the degree of ‘publicness’ – predict the perception of their occurrence.46 To identify specific moral problems that might be related to the presence of sector strange values in organizations, the analysis will use Huberts, van den Heuvel and Verberk’s (1999) categorical typology of integrity violations (see Table 4.7).
66
Table 4.7: Types of integrity violations (Huberts et al. 1999) Types of integrity violations 1. Corruption: bribing Misuse of public power because of private gain; asking for, offering, accepting bribes 2. Corruption: nepotism, cronyism, patronage Misuse of public authority to favor friends, family, party 3. Fraud and theft Improper private gain acquired from the organization (with no involvement of external actors) 4. Conflict of (private and public) interest Personal interest (through assets, jobs, gifts) interferes (or might interfere) with public interest 5. Improper use of authority (for noble causes) Use of illegal/improper methods to achieve organizational goals 6. Misuse and manipulation of information Lying, cheating, manipulating information, breaching confidentiality of information 7. Discrimination and sexual harassment Misbehavior towards colleagues or citizens and customers 8. Waste and abuse of resources Failure to comply with organizational standards, improper performance, incorrect or dysfunctional internal behavior 9. Private time misconduct Conduct in private time that harms the public’s trust in administration/government
This typology, an outcome of their analysis of the literature on police integrity and corruption,47 has since been used in a number of empirical studies (e.g., Huberts, Lasthuizen and Peeters 2006; Kolthoff 2007; Lasthuizen, Huberts and Kaptein 2002).48 In line with a previously developed questionnaire on integrity violations, values and leadership used by Kolthoff (2007) and Lasthuizen (forthcoming), the first eight categories were made operational in the survey instrument as 11 integrity violations. Category nine was omitted because it refers specifically to police officers whose professional codes also include their private time conduct. Also integrated into the questionnaire were six examples of decision-making inefficiency (delay, unnecessary delay, little attention to implementability, excessive participation procedures, too many costs, and decision-making processes that do not result in a real decision; cf. Russo and Schoemaker 2002). In Table 4.8 the resulting 17 items are listed for inclusion in the questionnaire and – to some extent – the interviews.
67
Table 4.8: Survey items on integrity and decision-making problems Integrity violations 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11.
Leaking information to the press The favoring of friends and family members outside the organization Cronyism in job appointment procedures Unjust treatment of internal and external clients and relations Providing information to third parties in exchange for payment Acts of fraud Side jobs or other employee activities that can harm organizational interests Bypassing norms and regulations to achieve certain objectives The use of organizational resources for private purpose Intimidations and insults among employees The use of office hours for private purposes Decision-making inefficiencies
12. Unnecessary delay in decision-making situations 13. Decision making that pays too little attention to the implementability of the decision 14. Delay in decision-making situations 15. Decision making situations that involve too much cost 16. Decision-making situations that do not result in a real decision 17. Excessive participation procedures in decision-making situations
Based upon the general predispositions in the literature and the values and items distinguished in this chapter, the following hypotheses can be formulated: H4 In public sector organizations, perceived integrity violations are positively related to the values collegiality, effectiveness, innovativeness, profitability, self-fulfillment and sustainability H5 In private sector organizations, perceived decision-making inefficiencies are positively related to the values dedication, efficiency, impartiality, incorruptibility, lawfulness, obedience, responsiveness, serviceability, social justice, and transparency These hypotheses will be empirically tested in the next two chapters.
68
5.
Crucial Values in Public Sector Organizations
This chapter shows what is valued most in government organizations, by empirically extending the previous discussion of public sector values that figure prominently in the literature. It does so by addressing the public managers’ responses to the survey questionnaire and in-depth interviews.49 After presenting the ratings and rankings of organizational values, the discussion examines the importance of specific values within the context of public sector decisionmaking and analyzes how, when, and under which conditions and circumstances these values are important. It also examines the extent to which values coincide with one another (complementarity) or hinder each other’s actualization (conflict). The chapter concludes with an analysis of the possible relation between values, integrity violations and decision-making inefficiencies.
5.1
What is valued more and less in government
The central research question of which organizational values are most important in decision making and how certain values are related to certain problems will be answered in this chapter for the public sector. It does so through a systematic presentation, analysis and preliminary discussion of the research results, beginning with the quantitative ratings and rankings of values given by the public managers surveyed. The survey respondents rated all 20 values in the value set (see Table 4.6) independently on a scale from 1 (‘not at all important’) to 10 (‘very important’), depending on their relative importance in organizational decision making. Such a scale implies that all values will receive relatively high scores because respondents need not select a limited number of values as truly important. Arguably, many values in the set are important for government organizations despite seeming to conflict or even be contradictory in daily organizational life. Not surprisingly, the variance in scores was considerable (5.67), with the public managers attributing explicit scores from 3.27 for ‘profitability’ to 8.94 for ‘incorruptibility.’ However, when profitability was excluded, the variance decreased to 3.06. The mean (M) of all public sector values with ‘profitability’ excluded was 7.45. Therefore, those values with a higher score are considered the most important and are shown in bold in Table 5.1. Overall, the survey ratings, which identify ‘incorruptibility,’ ‘accountability,’ ‘honesty,’ ‘lawfulness,’ ‘reliability,’ ‘transparency,’ ‘impartiality,’ ‘expertise,’ ‘effectiveness’ and ‘dedication’ as the most important public sector values, reflect a fairly traditional and consistent value pattern that is consistent with the literature. Specifically, these values are frequently mentioned as the most crucial in the administrative ethics and business ethics literature (e.g., Kaptein and Wempe 2002: 237–46; Kernaghan 2003: 712), in Dutch public sector codes of conduct (Ethicon 2003) and in earlier research on civil servants (van den Heuvel et al. 2002), although the variable ‘integrity’ is often used instead of the closely related ‘incorruptibility.’ 69
Table 5.1: Mean of value ratings by public sector managers (n = 231)
Value
Mean SD
1. Incorruptibility Acting without prejudice and bias toward private interests 2. Accountability Acting willingly to justify and explain actions to the relevant stakeholders 3. Honesty Acting truthfully and complying with promises 4. Lawfulness Acting in accordance with existing laws and rules 5. Reliability Acting in a trustworthy and consistent way toward relevant stakeholders 6. Transparency Acting openly, visibly and controllably 7. Impartiality Acting without prejudice or bias toward specific group interests 8. Expertise Acting with competence, skill and knowledge 9. Effectiveness Acting to achieve the desired results 10. Dedication Acting with diligence, enthusiasm and perseverance 11. Serviceability Acting helpfully and offering quality and service to citizens and customers 12. Efficiency Acting to achieve results with minimal means 13. Collegiality Acting loyally and showing solidarity with colleagues 14. Responsiveness Acting in accordance with the preferences of citizens and customers 15. Innovativeness Acting with initiative and creativity (to invent or introduce new policies or products) 16. Social justice Acting out of commitment to a just society 17. Obedience Acting in compliance with instructions and policies (of superiors and the organization) 18. Self-fulfillment Acting to stimulate employees’ (professional) development and wellbeing 19. Sustainability Acting out of commitment to nature and the environment 20. Profitability Acting to achieve gain (financial or other)
70
8.94
Std. Error 1.18 .08
8.42
1.33
.09
8.26
1.36
.09
8.08
1.62
.11
8.07
1.30
.09
8.07
1.34
.09
7.99
1.60
.11
7.94
1.26
.08
7.79
1.33
.09
7.56
1.25
.08
7.26
1.60
.11
7.02
1.51
.01
6.98
1.39
.09
6.74
1.58
.10
6.73
1.53
.10
6.61
1.92
.13
6.33
1.55
.10
6.26
1.49
.10
5.88
2.35
.15
3.27
2.02
.13
A comparison of Tables 4.1 (literature) and Table 5.1 (survey), however, reveals a number of interesting differences in the value hierarchy. Whereas ‘dedication,’ ‘effectiveness’ and ‘lawfulness’ rated 18th, 19th, and 21st, respectively, in the top 30 values in the administrative ethics literature, here, in alignment with earlier empirical research, all three were among the ten most important public sector values. In addition, ‘accountability,’ rated number 11 in the literature, rated second most important here. In contrast, ‘obedience,’ (rated 7th and 17th, respectively) and in particular social justice (3rd and 16th) are apparently valued much lower in real life than in the literature. Moreover, new or emerging values (Kernaghan 1994, 2000, 2003) traditionally associated with the private sector, like ‘innovation,’ ‘self-fulfillment’ and ‘profitability,’ are not among the manager’s rating of the most important public sector organizational values, although ‘effectiveness,’ whose ‘newness’ is disputed (see earlier discussion) is among the 10 most important public sector values. Thus, the rating results do not support the claim that classical public service values are devaluated or degraded by the emergence of classical business sector values. The results become even more surprising when the focus shifts to what is valued least. For instance, ‘collegiality’ and ‘responsiveness’ rate relatively low compared to the importance attributed to them in the literature and in other empirical studies on public sector values (Ethicon 2003; van den Heuvel et al. 2002; Vrangbaek 2006). On the other hand, these two values are absent from the 13 most prominent public sector values in Table 4.1. The same is true to a lesser extent for ‘serviceability’ (rated at number 11), a value associated with customer-oriented government that has been at the core of many government reforms and campaigns in the Netherlands and abroad and has often been promoted by public sector executives and politicians (see e.g., www.andereoverheid.nl). The classical Weberian ‘obedience’ received a low rating, arguably because it has a hierarchical and Tayloristic connotation that is out of alignment with modern-day thinking on (public) management. As already pointed out, the lower rankings of the typically business values of ‘innovativeness,’ ‘self-fulfillment,’ ‘sustainability’ and ‘profitability’ were less surprising, although it must be noted that governmental campaigns in the Netherlands to attract young highly educated potential employees from the tight labor market (see e.g., www.werkenbijhetrijk.nl) promote sustainability and emphasize self-fulfillment and self-development as important assets for government employment. ‘Profitability,’ perceived as the least important and rated by far the lowest, is clearly a value of little importance in public sector decision making. One question often raised in research of this type is what effect confounding variables – for example, age, gender, management experience or previous employment in the other sector – might have on managerial perceptions of important values (cf. Boyne 2002; de Graaf and van der Wal 2008). Here, however, given the questionnaire’s organizational focus, these variables were not expected to have much influence; that is, it should not matter whether a male or a female manager characterized the organizational value system. This
71
organizational focus was largely corroborated by the multivariate analyses, but a few interesting differences between subgroups in the sample are worth mentioning. First, in an independent samples t-test between three pairs of subgroups based on gender, age group and other-sector experience, fewer senior managers (aged below 46) attributed higher scores to ‘responsiveness’ and ‘social justice’ than did their older counterparts; however, remarkably, they attributed lower scores to ‘dedication,’ ‘expertise,’ ‘honesty,’ ‘sustainability’ and ‘impartiality.’ Nevertheless, absolute differences were small, except those for ‘sustainability’ (0.78) and ‘impartiality’ (0.65), which were both statistically significant. Second, a regression analysis showed no significant effects of the control variables on the values, with the exception of ‘past employment in the other sector’ for ‘transparency’ (.034, ß -.179, p
