How performance measurement systems in the healthcare sector influence the motivation, capacity and opportunity of employees to perform goal congruent behaviour: A qualitative approach

(1)

How performance measurement systems in the healthcare sector influence

the motivation, capacity and opportunity of employees to perform goal

congruent behaviour: A qualitative approach

MSc Organizational & Management Control University of Groningen

Faculty of Economics and Business

June 2015

Lodewijk van Dooren van Ostadestraat 185-III

(2)

Preface

This thesis is the last step in finishing my master’s degree Business Administration: Organizational & Management Control at the University of Groningen. During my studies I have gained a lot of valuable theoretical and practical knowledge about a broad spectrum of organizational theories and life in general. I hope to continue this steep learning curve at the same pace in the coming future. My interest in the management of people within organizations made me choose for this particular subject. Performance measurement is highly relevant these days and I believe it might be even more relevant in the coming few decades. I hope that reading this thesis will inspire others as much as it did me.

Finally, I would like to thank several persons who helped me finish this thesis. At first, I would like to thank drs. M.M. Bergervoet who constantly provided me with constructive and valuable feedback during the whole research process. Furthermore, I would like to thank my parents for their constant support and confidence in me during my studies. Lastly, I would like to thank all other persons that had a positive influence on this thesis.

(3)

Abstract

Performance measurement (PM) in the healthcare sector is increasing and there is a growing recognition that the ability of healthcare organizations to evaluate and report on performance is a critical building block for further improvement of healthcare delivery. This study combines management control (MC) literature and the Triad-model (Poiesz, 1999) of behaviour in order to answer the main research question in this research of how performance measurement systems (PMS) in the healthcare sector influence the motivation, capacity and opportunity of employees to perform goal congruent behaviour. In order to answer this question, a qualitative case study was conducted at a large Dutch non-academic hospital. This study found that there are several ways in which PM’s influence the dimensions of motivation, capacity and opportunity. This was however to a great extent determined by the specific characteristics of the PMS and the dimensions. Furthermore, the study explains how PM causes individuals to better achieve the main organizational goals. This study addresses several gaps in the current literature by being the first study that combines the PM literature and the Triad-model of behaviour in order explain the influence of PM on behaviour by taking the individual perspective as a starting point and by providing valuable insights of PM in the healthcare sector. From a more practical perspective, several important insights are discussed that managers can use when reasoning about PM in their organization. Future research in the area should take the form of multiple case studies and more longitudinal research is needed in order to fully understand the influence of PM on the behaviour of individuals in the healthcare sector.

(4)

Chapter 1. Introduction

Nowadays, everyone is measuring performance (Behn, 2003). Especially in the healthcare sector there is growing recognition that the ability to evaluate and report on performance is a critical building block for further improvement of healthcare delivery (Miller & Leatherman, 1999). Healthcare organizations follow the tendency of private companies to measure the performance of their employees in attempt to react to the increased pressure for competition in this sector. Most managers would agree that employees make a critical difference when it comes to organizational performance (Bakker & Schaufeli, 2008) and that it is the behaviour of employees, which ultimately leads to the achievement of the organizational goals (Hoogervorst et al., 2004). Therefore, understanding why behaviour takes place in general and how performance measurement (PM) influences this, is highly important.

In the last decade, researchers identified some serious gaps in the PM literature. Firstly, while much has been written in theory about the use of PM’s to influence employee behaviour, these theories explain little about their operation in practice, especially in the public sector (Propper & Wilson, 2003). The public sector is quite different from the private sector and therefore the impact of PM’s is somewhat contested (Teelken, 2008). The healthcare sector is a very specific sector that recently went trough some turbulent changes. Nowadays, healthcare organizations can only survive if they are able to deliver a high level of quality against reasonable costs. Secondly, PM is mostly studied from a top-management perspective where the focus lies on how PM allows them to monitor whether certain objectives have been achieved or not and how they use it to motivate employees, formulate strategies, communicate expectations and indicate possibilities for improvement (Behn, 2003; Kaplan & Norton, 1992; Simons, 1994; Wouter & Wilderom, 2008). Surprisingly, there are only a few studies that take the individual as a starting point. This study argues that the way in which individuals experience the PMS it vital to understand as this partly determines the effect on their behaviour. Thirdly, there is very little evidence that PM in the healthcare sector leads to better attainment of the organizational goals (Propper & Wilson, 2003). In theory, the performance measurement system (PMS) helps the organization to achieve its goals. If this is indeed the case remains largely unknown.

(7)

PMS (interactive/diagnostic). Starting from these typologies, this study provides a description of how PMS’s in the healthcare sector influence the behaviour of employees. Furthermore, this research makes use of a behavioural model from the economic psychology: the Triad-model developed by Poiesz (1999). This model is based on the assumption that behaviour only takes place when three conditions are met: the person must be willing to display a certain type of behaviour, the person must be able to display the type of behaviour and the person should be given the opportunity to display this type behaviour. In other words, it is needed there is a sufficient level of respectively motivation, capacity and opportunity to display a certain type of behaviour. This research will also investigate if the PMS stimulates the behaviour in the direction of the main organizational goals: quality and efficiency. Together, this will lead to an increased understanding of the influence of PM’s on the behaviour of the healthcare employee and enables us to understand if this is leading to goal congruence.

This research contributes to the existing literature in several ways. Firstly, by combining the MC literature with literature from the economic psychology, this study was able to gain a better understanding of the effect of PM on the behaviour of employee. This is the first study to combine these streams of literature in order to get a better understanding the impact of PM in the behaviour of healthcare employees. Secondly, we will add to the PM literature by looking from the perspective of the individual instead of the top-management perspective. Thirdly, we will add to the understanding of the use of PM in the public sector whereby we simultaneously provide more insight on if these measures actually help to achieve the organizational goals.

From a more practical perspective this study also has multiple benefits. To begin, at the most general level, it forces managers to think in a structured and explicit way about behaviour of individual employees and its underlying causes (Poiesz, 1999). According to Dubin (1976) practitioners often have the tendency to derive a problem on the basis of observed symptoms. This study forces to think about the ‘why’ behind the seen behaviour structured around 3 dimensions, which cause the behaviour. Managers who understand this, will be able to use these systems in such ways that it leads to goal congruent behaviour. In addition, since many MC policies, are in fact behavioural policies (Poiesz, 1999), managers could use this study as a model to think about the effects of the policies they use in their organizations.

From the introduction above, is should be clear that the central research question addressed in this study is:

(8)

In order to be able to give a complete answer to this question we will address a set of sub questions that are presented below:

1. What are the characteristics of the healthcare context? a. What are the main goals of healthcare organizations? 2. What are management controls?

3. What are PMS’s?

a. How does the management use them? b. How do employees perceive them? 4. Why does behaviour take place?

a. What are the forces that drive this behaviour? 5. How do PMS’s influence the forces that drive behaviour?

a. Does this lead to goal congruence?

In Chapter 2, this study concentrates on the sub-questions proposed above and concludes by presenting the conceptual model. In Chapter 3 the research methods, the research case, the data collection and the methods of data analysis are discussed. Chapter 4 presents the findings from the data and discusses some unexpected findings. Lastly, in Chapter 5 the conclusions of the research are discussed and limitations and recommendations for future research are given.

Chapter 2. Literature

In this chapter the theories and variables used in this research are introduced and operationalized. The first section introduces the healthcare sector and provides detail about the context of this research. The second section explains the concept of MC, discusses a typology of MC’s and indicates how MC’s are used in organizations. The third section focuses one particular type of MC, namely, PM. The fourth section provides some theories that explain why behaviour takes place in general and what drives the behaviour in organizations. In the fifth section, this research combines the different theories and literature in order to give a preliminary answer to our research question. The last section presents the variables and relations that are discussed in a conceptual model.

2.1 The Healthcare Context

(9)

limits the ability to measure the output of healthcare institutions and employees. Moreover, the way of delivering the outputs, the so-called transformation process, is often ambiguous and hard to predict. It is not clear which actions exactly have to be taken to reach certain goals. Every patient is different and thus, requires a different ‘transformation process’. Additionally, healthcare is mostly non-marketed to its patients and information regarding price and quality is often difficult to obtain and compare (Sharpe et al, 2007). Healthcare insurers negotiate prices and terms with the hospitals and the end-users often do not receive a single bill for services delivered. Also, third-party and government interventions are frequent and multiple laws and supervisory institutions determine for a great deal how healthcare institutions operate. To complicate even more, in comparison to commercial organizations, errors and service failure may result in disastrous consequences for the end-users as patients may die even from the smallest errors (Ndubisi, 2012). Furthermore, the core-employees within a healthcare institution are often highly trained professionals with a high level of autonomy who work in differentiated domains and therefore, often with different goals. This not a complete list of characteristics of the healthcare sector, nor are they unique to healthcare alone but the interplay and combination of these characteristics is.

Producing health and wellbeing is the overall main objective of healthcare organizations (Grossman, 1972). This is realized when healthcare is of considerable quality and when it is affordable to the main public. Treatments thus have to be performed in a high quality and, in order to keep available for the public; they must provide this care at reasonable costs (Mills & Spencer, 2005). Quality of care and efficiency are thus vital for healthcare organizations and employees should perform behaviour in line with these goals. However, this is simpler said than done. Logically, there is much possibility of conflict between these two main goals, as there is generally a positive relationship between the quality of care and its costs. Moreover, it is likely that some groups within the organization give more attention to one of the two values. Arguably, medical specialists are generally less concerned with the costs of care, and focus working towards achieving the highest quality possible when diagnosing and treating patients. They do care less about the time and effort they spend in caring for a certain patient as long as the outcome of the care is positive. In contrast, the unit-managers and top-management are probably more concerned with cost-efficient care. However, in order to function appropriately and to keep functioning in the future, healthcare organizations must pursue both values simultaneously (Mills & Spencer, 2005). The extent to which healthcare organizations are able to balance these two values seems crucial for their success.

(10)

early 1990s made considerable progress in changing the Dutch healthcare system from a supply side government regulated market, to a system of managed competition (Schut & Van de Ven, 2005; Van de Ven & Schut, 2008). The idea of this system of managed competition is that healthcare insurers and providers compete with each other on price and quality, while the government establishes certain rules to insure that overall public objectives are met (Van Kleef, 2012). In 2006 a major step towards this objective was taken by introducing the Health Insurance Act. This made it possible for every Dutch citizen to buy a basic health insurance package from private health insurers, which stimulated competition on price and quality. As a result, insurers stimulated healthcare providers to increase the quality and efficiency of their organizations and to provide insight on their performance (Van de Ven & Schut, 2009).

2.1.1 Goals Congruence

Managers should make sure employees act in accordance with the organizational goals. They do this by making use of different kinds of MC’s. This is clearly reflected by Ouchi (1979) who argues that MC’s provide the means for cooperation among units or individuals, even in case of partially congruent or even contradictory objectives. Before we will explain how these MC’s lead to goal congruence, it might be helpful to have a better understanding of the meaning of goal congruence in healthcare institutions.

Goal congruence is the alignment of different perspectives from groups within the organization to the main organizational goals (Bouillon, 2006). A shared interest of the multiple groups within the organization to reach this goal congruence may be a powerful facilitator of cooperation between groups (O’Toole, 2003). As stated before, the goal of healthcare institutions is making the world as healthy as possible by providing high quality care against reasonable costs. This means that, in the end, quality of care and efficiency are the main drivers of performance for healthcare organizations. Subsequently, healthcare employees behave goal congruent when they act in accordance with these two goals. The PMS within an organization should help managers to channel the efforts of the employees towards these organizational goals.

In terms of the quality objective, patient safety, patient centeredness, effectively of treatments, broad

scale of possible treatment, complication rate and survival rates could be indicators that determine

(11)

2.2 Management Controls

Since PM is a specific form of MC, it makes sense to better understand the concept of MC before continuing. This section shortly discusses the concept of MC and presents some background information. In addition, a typology of MC is presented and to gain a better understanding of MC in organizations.

2.2.1 Background and Definitions

Not surprisingly, managers see MC’s as vital instruments because they are means to influence employee behaviour and thus, the organizational performance (Hameed & Waheed, 2011). MC was first defined by Anthony (1965) as ‘the process by which managers ensure that resources are obtained and used effectively and efficiently in the accomplishment of the organization’s objectives’. In this definition he clearly refers to the fact that MC’s are used to steer resources, such as employees, towards the organizational goals. Merchant & Van der Stede (2007) explicitly refer to the steering of the behaviour of individuals by defining MC literally as dealing with employees’ behaviour. They state, ‘it is people in the organisation who make things happen. MC’s are necessary to guard against the possibilities that people will do something the organisation does not want them to do or fail to do something they should do’. Examples of MC’s are the PM’s, culture, rules, procedures, organisational structures and budgets within organizations. This is not a complete list; there are many different sorts of MC’s active within organisations.

(12)

This is in line with definitions of Abernethy and Chua (1996), Flamholtz et al. (1985) and Ouchi (1979), which are based on the assumption that someone (senior management, managers) is seeking to control the behaviour of others (middle management, employees).

2.2.2 Classifications and Typologies

In order to group different types of MC’s and to classify them according to certain characteristics researchers came up with various typologies MC’s. In their study, Langfield and Smith (1997) already described various typologies such as formal and informal controls (Anthony et al., 1989) output and behaviour controls (Ouchi, 1977) market, bureaucracy and clan controls (Ouchi, 1979) administrative and social controls (Hopwood, 1976) and results, action and personnel controls (Merchant, 1985). These typologies ought to classify MC’s in terms of what they are made of how they function. For example, formal controls include rules, operating procedures and budgeting systems, while informal controls include the unwritten policies of the organization (Langfield and Smith, 1997).

More recent research in typologies implies that it might be more relevant to look at how these MC’s are used instead of what a MCS consists of. One of the influential works in this matter is the work of Simons (1994). It describes the role of PM in organizations. His framework is based on four levers of control: beliefs systems, boundary systems, diagnostic control systems and interactive control systems. He defines beliefs systems as ‘the explicit set of values that senior managers communicate formally and reinforce systematically to provide purpose and direction for the organization’. Beliefs systems thus communicate core values and motivate employees to take appropriate actions and search for new opportunities (Widener, 2007; Tuomela, 2005). Boundary systems ‘delineate the acceptable domain of strategic activity for organizational participants’ (Simons, 2000). They are thus quite similar to beliefs systems in terms that they motivate to search for new opportunities, but a boundary system defines the limits of this search (Widener, 2007). Diagnostic control systems are the key performance indicators (KPI) that are communicated and monitored (Tuomela, 2005). As is the case with the boundary systems, diagnostic systems act as a constraint on the employee behaviour (Simons, 2000) because they focus on behaviour that employee have to perform. Lastly, interactive

control systems are use for forward-looking dialogue to discuss uncertainties and to develop and learn

(13)

in different roles in the organization, that is, different levers of control (Widener, 2007). This depends on how they are used by the organizational managers.

2.3 Performance Measurement Systems

In this section we will provide more information on the use of PMS’s in organizations. Although a PMS is only one aspect of the total MCS, this study devotes a separate section explaining the use of PMS’s within organizations because this is an essential part of the study. To understand the effects of PMS’s in organizations is it important to know how PMS is defined, how employees perceive the PMS and how managers can use it in the organization.

Tuomela (2005) defines PMS as ‘collections of financial and/or non-financial performance indicators that managers use to evaluate their own, their unit’s or their subordinates performance’. There is a wide range of different financial and non-financial indicators to evaluate performance. Most measures of financial performance fall into the categories of investor returns or accounting returns such as shareholder values, price/earnings ratios and product costs (Cochran & Wood, 1984). Non-financial measures try to evaluate performance according to indicators such as customer satisfaction, employee satisfaction and quality of delivered products or services (Abdel-Maksoud et al., 2005). It is recognized that management can use these measures to focus manager’s attention, reveal priorities for improvement and rewarding employees (Behn 2003; Jordan & Messner, 2012; Wouters & Wilderom, 2008). By selecting the right performance measures and incentives the management should be able to achieve goal congruence (Bouillon, 2006).

Traditionally, PM’s predominantly made use of financial indicators. However, nowadays it is recognized that exclusive reliance on financial measures in a PMS is insufficient to achieve long-term performance (Kaplan & Norton, 2001). One of the main reasons for this is that financial performance indicators have only limited capability of indicating future performance; they focus more on past performance. Furthermore, financial indicators have often been criticized for not emphasizing elements that lead to good or poor results (Norreklit, 2000), encouraging short-term decision-making (Banks & Wheelwright, 1979) and driving excessive risk-aversion and gamesmanship (Merchant, 1985). Non-financial indicators overcome many of these problems and since they are capable of including future performance perspectives of organizations, the last two decades researches have made an attempt to develop PM frameworks that balance between financial and non-financial performance measures (Bourne et al., 2000). By balancing financial and non-financial information, these frameworks have the ability to indicate past performance and give indications of the future performance.

(14)

& Norton (1992). They describe their framework as ‘a set of measures that gives top-management a fast but comprehensive view of business. The balanced scorecard includes financial measures that reveal the results of actions already taken. It complements the operational indicators with measures on customer satisfaction, internal processes and the organization innovation and improvement activities’. They state that in contrast with the more traditional view of PMS, which really try to control the behaviour of employees, PM in the balanced scorecard is used to help employees perform better their tasks by creating a clear vision and communicating direction. Interestingly, this indicates that these frameworks assume that PM is something used to improve and guide employees instead of solely controlling them. This relates to the notions of Adler and Borys (1996) of enabling form of control, which is discussed below.

2.3.1 Enabling and Coercive

PMS’s are mostly studied from top-management perspective whereby researchers describe how it allows them to monitor if organizational objectives have been achieved (Wouters & Wilderom, 2008). However, since this study is focussed on the perspective of the individual, it would be helpful to use a MC typology, which likewise takes this into account. One way of characterizing the PMS from the employee’s perspective is by using the framework of enabling and coercive formalisation suggested by Adler & Borys (1996). It characterizes a PMS as more enabling or coercive depending on how it is perceived by the employees subject to it. Important to note is that this typology only makes a distinction in how employees perceive the PMS, it does not make for example suppositions about which of the two views it better.

(15)

achieved when management does not solely rely on the PM information, but uses it in a complementary manner to estimate performance.

Adler & Borys (1996) indicate that employees perceive a PMS as enabling when it facilitates their actions instead of disproportionally constraining them and when it is used to support them by providing feedback, identifying problems, learning and revealing opportunities. In addition, Jordan & Messner (2012) state that a PMS is perceived as more enabling when the development process of such system is organized with user involvement, rather than being made exclusively by outside experts. In addition, they explain that a PMS can even be perceived as enabling when it does not fully reflect the performance of employees, that is, when it is incomplete. Even incomplete PMS’s can be perceived as enabling when the four features described above are satisfied to a certain extent.

In contrast, Jordan & Messner (2012) state that when these features are not satisfied and when the PMS only serves higher management needs and is used as mean to control employees instead of serving employees to better master tasks, it will be perceived as more coercive. They add to this that especially during the introduction of a new control system, feelings of coercion are easily created. In their research it also became clear that when employees perceive the PMS as coercive and at the same time incomplete, it likely that this will trigger tensions and general dissatisfaction about the PMS (Jordan & Messner, 2012).

2.3.2 Diagnostic and Interactive

Another way to characterize PMS’s is to distinguish in how de management uses the PMS. Simons (1999) makes this distinction and discusses diagnostic and interactive use of a PMS. It is used diagnostically when it sets goals and measures outputs, and computes variances in order to estimate performance (Simons, 2000). When the PMS is used diagnostically, it functions routine based and focuses on general KPI’s. In this case, the PMS looks for exceptions, compares the results and provides feedback on performance (Widener, 2007). Diagnostic PMS’s provide managers with the information on results that are not meeting expectations and are typically involved with single loop learning (Argyris, 1977). It can be compared with a dashboard in a car. Management uses a PMS in a diagnostic way to gain control of the situation.

(16)

in performing an equivalent task in the future. Simons (2000) argues in line with this and argues that interactive use of controls is enhances the abilities and skills of employees. It is thus not just used to provide information about whether things are going well or not, but it really tries to discover the underlying logic behind the outcomes achieved. While the interactive use of a PMS uses significantly more effort from the management, the effect on the long run is higher compared to the diagnostic use of the PMS.

While in the past the discussion among researchers has been whether PMS are used diagnostically or interactively (Simons, 1999), nowadays it is recognized that PMS’s can be used diagnostically, interactively or in both ways. We argue, in line with Henri (2006) and Widener (2007), that diagnostic and interactive use of the PMS is complementary. It is for example possible to use some of the PM’s diagnostically on weekly basis and interactively on a monthly basis. Important to remember is that not the PMS itself, but the way how it is used determines whether it is classified as diagnostic or interactive.

2.4 Behavioural Theories

This section continues with the presentation of two behavioural theories to better understand the concept of behaviour, why behaviour takes place and what forces drive this behaviour. Again, the focus lies on behaviour of individuals.

2.4.1 Theory of Planned Behaviour

One of the most influential behavioural theories is the theory of planned behaviour (TPB) described by Ajzen (1985, 1991), which is an extension of his theory of reasoned action (TRA). The TRA is based on the proposition that an individual’s behaviour is determined by the intention to perform that behaviour (Fishbein and Ajzen, 1975; Ajzen and Fishbein, 1980), which is basically the combination of two factors: an individual’s general feeling of favourableness for displaying a certain behaviour (attitude) along with the perception that important people surrounding the individual think that the behaviour should (not) be performed (subjective norm).

(17)

significant barriers or constraints to perform the behaviour (Armitage, 2001). A schematic representation of Ajzen’s (1991) theory can be found in appendix A.

While the theory of planned behaviour was found to be quite useful in explaining health related behaviours (Godin & Kok, 1996), this study uses the Triad-model of behaviour to understand the influence of a PMS on behaviour. The main reason for this was the simplicity of the Triad-model that allows for a simple but clear analysis. Furthermore, since to our knowledge there are no other studies that used the Triad-model to explain healthcare related behaviour, the use of this model could lead to significant new insights.

2.4.2 The Triad-model of Behaviour

Poiesz (1999) developed the Triad-model to explain how behaviour is determined at individual level. The Triad-model is an elaboration of two leading models used in the social psychology and marketing research literature. The first one is the elaboration likelihood model of Petty & Cacioppo (1986) that describes how attitudes form and change over time (Angst & Agarwal, 2009). The second model on which Poiesz (1999) model builds is the MOA-model that describes how individuals process marketing information (Andrews, 1988; MacInnis & Jaworski, 1989). This model uses the factors of motivation, opportunity and ability (MOA) to determine the probability that individuals react to certain marketing advertisements. The Triad-model can be theoretically positioned as an elaboration of the Elaboration Likelihood model and the MOA-model (Poiesz, 1999) and can be used in a wide range of settings to determine the probability of behaviour. An overview of these preceding models of the Triad-model together with their constructs can be found in appendix B.

In the Triad-model, behaviour is determined as a result of the interplay of three dimensions: the

motivation of an individual to perform certain behaviour, the capacity of an individual to perform

certain behaviour and the opportunity of an individual to perform certain behaviour. The model uses these three dimensions to summarize all causes of the occurrence of certain behaviour. By multiplying the scores of the three dimensions, the model predicts the probability that certain behaviour will take place. However, this study is not very much interested in the exact probabilities of behaviour since our goal is to understand how the behaviour is influenced by the PMS and if this leads to goal congruence. Before elaborating further on the model, the three dimensions of the Triad-model will be defined and characterized in more detail below:

(18)

individuals to perform a certain tasks or behaviour. Ryan & Deci (2000) tend to agree with this by stating that if someone is motivated for something, this means this individual is ‘to be moved’ to do something.

Motivation has lots of different sources or drivers. To explain this, imagine for example a common student. It is possible that this student is motivated to do his homework because of interest and curiosity for the subject. On the other hand, it can likewise be possible that this student is motivated to do his homework because he will get a present from his parents when he finishes it. In these cases, the amount of motivation does not necessarily vary but the nature or source of the motivation does (Ryan & Deci, 2000). People can be motivated to do something because they value an activity or, alternatively, because there is a strong external coercion to do something (Ryan & Deci, 2000). This reflects the much used distinction of intrinsic and extrinsic motivation. These two types of motivation reflect the two poles of a motivation internalization continuum (Ryan & Connell, 1989). The more the motivation is coming from the person itself, that is, the more internalized, the more we can call the motivation intrinsic.

Literature refers to intrinsic motivation when someone is moved to do something because it is interesting and enjoyable. Individuals who are intrinsically motivated thus simply motivated regardless of rewards or compensations (Deci and Ryan, 1985). This type of motivation clearly comes from within the individual. Comparisons between people whose motivation is authentic (intrinsic) or more externally controlled (extrinsic) reveal that intrinsically motivated individuals typically have more interest, excitement and confidence, which is highly associated with enhanced performance, persistence and creativity (Ryan & Deci, 2000). Autonomy is one of the drivers of intrinsic motivation and research indicates that tangible rewards, deadlines, directives, pressured evaluations and imposed goals can diminish intrinsic motivation because they lead to a reduced perceived autonomy (Ryan & Deci, 2000).

In contrast, extrinsic motivation refers to doing something for instrumental reasons, which is the case when there is a reward or compensation coupled to the performance of certain behaviour (Gagne et al., 2010). Ryan & Deci (2000) also state that ‘extrinsic motivation refers to the performance of an activity in order to attain some separable outcome and, thus, contrasts with intrinsic motivation, which refers to doing an activity for the inherent satisfaction of the activity itself’.

(19)

Physical capacity relates to the characteristics such as length, strength or health of individuals. Financial capacity relates to the extent to which an individual has the economic means to perform

certain behaviour. Mental capacity refers to the general mental ability of individuals to perform certain tasks. According to Winterton et al. (2005), mental capacity relates to the fuzzy en weakly defined concept of an individuals competence, which is best defined as the ‘underlying characteristics of individuals that are causally related to effective or superior performance, which is generalizable across situations, and enduring for a reasonably long period of time’ (Boyatzis, 1982). Weinert (2001) lists a range of factors that are found to positively relate to an individual’s degree of competence: ability, knowledge, understanding, skill, action and experience. Other research likewise demonstrated that higher scores on general mental ability are positively associated with the main predictors of job performance (Schmid et al., 1992). In essence, this is in line with this study because this study likewise proposes that the capacity dimension positively influences the degree to which goal congruence can be achieved and thus, performance. Finally, Personal instruments relate to the devices, which we perceive to belong to us and that we can handle ourselves or perceive as part of our functioning (e.g. tools, dictionary, glasses).

The capacity dimension is quite similar to the concept of perceived behavioural control in the theory of planned behaviour (Ajzen, 1991). However, this concept relates to the combination of capacity and opportunity of an individual to estimate the perception of easy of performing behaviour. In contrast, in the Triad-model, opportunities to perform behaviour are not seen as characteristics of capacity but as a separate dimension. In this study, it is the combination of the different forms of capacity discussed above that compass the capacity dimension of an individual.

Like is the case with the other dimensions, the capacity dimension should be seen in relation with the behaviour to be performed. An individual does not have a high or a low score on capacity, but does have a high or low score on capacity to perform certain behaviour. To illustrate this, a cardiac surgeon would most probably have a low capacity to set up a hedge fund (financial institution), while a financial genius would have a low capacity of performing heart surgery. However, the capacity of a cardiac surgeon to perform hearth surgery would be quite high, as well as the capacity of the financial genius to set up a hedge fund. Furthermore, it important to understand that capacity and motivation relate to the individual at issue while opportunity relates to the circumstances where the individual is in.

(20)

guide the behaviour. Moreover, the availability of data that can be analysed can be an example of a circumstance that influences the behaviour. Also, the extent to which important others approve or disapprove certain behaviour would fall into the opportunity dimension. Again there is a clear difference with the theory of planned behaviour. While in the last, important others would fall into the factors that define intention, the Triad-model describes this as an opportunity to perform behaviour because this pressure stems from outside the individual. This list is far from complete since there are numerous examples to illustrate the opportunity dimension.

Other researchers also mention the opportunities to perform behaviour. Siemsen at al. (2008) mention the concept of opportunity in their research and define it as the environmental or contextual mechanisms that enable action. In addition, Sarver (1983) describes the importance of opportunity as the ‘context of opportunity’ by which he meant to which extent the situation where someone is in provides an opportunity for a person to act in a manner consistent with his attitude, beliefs and intention towards a specific behaviour.

To clarify the definition of opportunity, imagine a person who is both motivated and sufficiently capable to make a fire. When he stands in the garden of his house, he clearly has the opportunity to make the fire and therefore, he will probably do it. However, when this person is in a school building, he is inhibited to actually make the fire. There is little opportunity to do it because there are rules against making fires inside schools and because the lack of a proper smoke evacuation system, the excessive smoke formation will not make it a pleasant situation.

Opportunity sometimes looks like a side term in the light of motivation but it does have a clear own character because it relates to the circumstances where the person is in, instead of relating to a person at issue (Poiesz, 1991). This separation between external and internal pressures from the viewpoint of the employees is highly important to distinguish between the dimensions. However, this does not mean that opportunity to perform does not relate to motivation. As we will see in the next paragraph, there could be reciprocal effects in the model.

2.4.3 Reciprocal effects in the model

(21)

occurs when for example the high-potentials of organizations have the status of trainees. Ford et al. (1992) explain in their research that trainees have more opportunity to learn because they receive above average supervisory support and training. Of course, this will have an effect on the capacity of the trainees as well.

There are many other examples possible that illustrate the reciprocal relationships within the Triad-model. However, this study will not further investigate the reciprocal effects within the model since the goals of this study is to understand the effect of the Triad dimensions on goal congruent behaviour. It is questionable if an elaborate discussion about the reciprocal effects within the model does significantly add to this goal.

2.5 Healthcare, PM and the implications on Behaviour

Now that all concepts made known in the introduction are defined and explained, this section continues with an elaboration of how these concepts relate to each other. Firstly, the relation between the healthcare sector and the PMS is discussed. After this, the relations between the PMS and the Triad dimensions are discussed whereby is indicated whether or not this will contribute to goal congruence.

2.5.1 Healthcare and PMS

While there is much written about PM in the healthcare sector in general, not much is published on whether the PMS’s are perceived as enabling or coercive in this sector. Therefore, it is quite hard to predict on how the PMS is perceived by the employees. However, by linking various characteristics of the healthcare sector to the PM literature this study was able to make some implications.

(22)

With regard to a more interactive or diagnostic use of the PMS in the healthcare sector the current literature is not very conclusive either. Research does indicate that performance indicators in healthcare are generally based on patient satisfaction surveys, costs of care, number of treatments and duration of treatments (Marshall et al., 2006). There is little empirical evidence that PM in the healthcare sector actually leads to improved service provision or better quality of care (Teelken, 2008). This indicates that PM’s in the healthcare sector are not used in such a way that they stimulate learning, which would be the case if the PMS were used interactively. Furthermore, we have explained that the interactive use of the PMS takes a lot time from the general management. Therefore this research expects that foremost of the PM’s within the PMS are used diagnostically in the organization since it would not be doable for any management to use all the PM’s interactively. However, most probably there are also some PM’s that will be used more interactively such that they are thoroughly discussed on unit level in order to really improve the way of doing things.

2.5.2 PM and Motivation

Literature has proven that PM is a very useful mean to increase the motivation of employees to reach certain goals (Behn, 2003). Especially the attachment of rewards to PM’s is very effective for increasing employee motivation. Malina and Selto (2001) agree with this but add that performance indicators and targets should also be controllable and challenging but attainable. When these conditions are not met, PM’s have the potential to decease the level of motivation. Likewise, Decoene and Bruggeman (2006) conclude in their research that there is a clear relation between PM’s and the level of motivation of employees. They indicate that PM’s do have the potential to increase the motivation of employees when they have a high level of strategic alignment, are controllable and are technical valid.

(23)

this case intrinsic motivation is believed to increase (Frey and Oberholzer-Gee, 1997). It remains unclear what the effect of the shift between the two types of motivation is on the total level of motivation. However, an important insight from this could be that at least when employees perceive the PM as supportive and confirming their performance, this may lead to increased intrinsic motivation. This new insight leads to the expectation that when management uses the PMS more interactively instead of diagnostically, it is likely that employees will be more motivated to perform goal congruent behaviour. Recall that diagnostic use of the PMS if mostly used for control purposes while an interactive use emphasizes learning and support, which is helpful for the specialists.

It is also reasonable to believe that the distinction between enabling or coercive is also important when considering the relation between the PMS and the motivation. As is explained, when employees perceive the PMS as coercive, this could give rise to dissatisfaction, a concept closely related to motivation. Especially for medical specialists, who are highly trained professionals, it could be very frustrating to have the feeling to be constantly controlled and to be forced to work in line with certain PM’s in which they do not see the relevance or which are a incomplete representation of their performance. Therefore, it is expected that when the employees perceive the PMS as enabling, this will have a more positive influence on their motivation to perform goal congruent behaviour. In contrast, when the PMS is perceived as coercive, we expect that the PMS can have a negative influence on the level of motivation.

2.5.3 PM and Capacity

Current literature also shines light on the relation between PM and the capacity of employees to achieve organizational goals. Kaplan and Norton (1992) suggest that organizations need to make their PM’s in a way that employees can learn from them and can use them to increase their knowledge on how to improve performance. Tuomela (2005) states that when appropriate feedback mechanisms are in place, the PMS can help individuals to identify problem areas in the organization that once solved, will increase the capacity to handle future tasks. Davenport (2006) even takes this belief a step further and states that the ultimate goal of a PMS should be learning, rather than control. These authors clearly share the view that individuals should learn from the PMS, which increases the capacity to perform goal congruent behaviour and to achieve the organizational goals.

(24)

diagnostic use of a PMS facilitates single loop learning (Argyris, 1997). Following this logic, both manners of use of the PMS lead to more capacity to perform goal congruent behaviour. However, interactive use is expected to have more effect on the capacity than the diagnostic, especially in the long run because it enables an individual to learn and to become better in future tasks. Diagnostic use serves mainly as a sort of instrument that helps to keep control like a dashboard in a car.

Taking into account the notions of enabling and coercive PMS is likewise important when investigating the effect of the PMS on the capacity of employees. Adler & Borys (1996) state that enabling procedures help employees to do their jobs faster and more effectively. Therefore, this study expects that the more the PMS is perceived as enabling, the more positive the influence on the capacity dimension. In contrast, PMS’s are perceived as coercive when they constrain the specialists in doing their tasks and when they only serve higher management needs. This could lead to either no effect or even a negative effect on the capacity dimension. A negative effect could be the case when the coercive PMS takes an excessive amount of time from the specialists, which means less time to spend on caring for their patients. This would for example decrease the capacity to deliver high quality of care. In line with the arguments above, this research argues that the more enabling the PMS’s have a more pronounced positive influence on the capacity of the employees.

2.5.4 PM and Opportunity:

(25)

are easily translated in valuable information in a timely manner, PMS’s can increase the opportunity dimension of the Triad-model.

Another way in which the PMS influences the opportunity dimension is that they communicate the direction of behaviour desired by the management. It is widely known that when things are being measured, most employees will behave in a certain extent towards these measurements. It thus serves as a tool to communicate to the employees what is important to the organization and therefore works stimulating on the behaviour that is intended with these measures. However, when performance indicators do not exactly measure the behaviour that is intended, this could stimulate ineffective behaviour. When healthcare professionals are for example only measured on basis of the number of patients they treat each day, this could obviously stimulate the behaviour that specialists spend as less time as possible with the patients. This refers to the work of Smith (1995) who argues that wrongly constructed PM’s can have some unintended consequences on behaviour. Among others, he defines the problem of tunnel vision ‘an emphasis on phenomena that are quantified in the PM scheme at the expense of un-quantified aspects of performance’. The problem is here that in an extreme case, employees do nothing else than what is being measured. Another problem he defines is sub optimization, which is ‘the use of narrow local objectives by managers, at the expense of the objectives of the organization as a whole’. However, this research expects that because of their high intrinsic motivation, healthcare professionals generally attach more value to their patients than to the measurements. This makes it less likely this problem will become as extreme as in the example. We argue in line with Behn (2003) who states that the PMS can influence the behaviour of employees in both desired and undesired ways, that is, to goal congruence or not.

2.6 Theoretical Framework

(26)

depending on the combination between characteristics of the PMS and the concerned dimension. All together, this resulted in the theoretical framework displayed in the figure 1.

Figure 1: Conceptual model of the research

Next to the reciprocal effects within Triad-model that were discussed, this study realizes that there are also other reciprocal effects within this conceptual model. For example, it is logical to assume that when PMS are used interactively, they will be perceived as enabling because they can support employees in doing their work better in the future. In addition, it is reasonable to argue that when employees increase the quality of care, the efficiency of care also increases because fewer patients have to come back to the hospital for extra treatment. However, like is the case with the reciprocal effects within the Triad-model, we will not discuss these in the rest of this paper. Taking all these effects into account would be extremely difficult and time consuming and we do not believe that this will add significantly to the results of this study.

Chapter 3. Methodology

This chapter begins with an elaboration on the type of research and the chosen research method for this study. In the second section, the method of data collection will be described and the research case will be presented. The third section provides more detail about the techniques used to optimally interpret the collected data.

3.1 Type of Research and Research Method

This study is based on the theory development process as described by Van der Bij (2014). This type of research was most appropriate because the subjects of this study were not yet totally addressed in the literature. Since the aim of this study is to get a better understanding of how PMS’s influence the motivation, capacity and opportunity of healthcare employees to perform goal congruent behaviour and because only specific parts of this question are answered in the current literature, a theory development process was an excellent starting point.

(27)

the understanding of the dynamics of certain phenomena within specific settings (Eisenhardt, 1989). To be more specific, an explanatory case study was conducted. Explanatory case studies seek to use theory in combination with generated data in order to explain the presumed causal relations in real-life settings with the aim of understanding and explaining the specifics, rather than to produce generalisations (Baxter & Jack, 2008; Scapens, 2004). This fitted very well with the aim of this study since it this study used the combination of various MC and PM theories in order to understand the effect of PMS’s on employee behaviour in the healthcare sector.

A deductive reasoning approach was used, which is a research process that begins with general theories or generalizations to investigate if these are also applicable to specific circumstances, such as the healthcare environment (Eisenhardt & Graebner, 2007; Hyde, 2000). As is explained in more detail in the next section, the basic design type of this study was a single-case design with a single unit of analysis. The research is cross-sectional and as a result describes the situation at a single point of time. While preferable, due to the time constraints, the researcher was not able to perform a longitudinal research.

3.2 Data Collection

For our research we selected the Dutch hospital ‘Jeroen Bosch Ziekenhuis’ (JBZ) in the city of ‘s-Hertogenbosch. In 2011 three hospitals in the region joined forces to form the JBZ and as a result this hospital is now the largest employer of the whole region ‘s-Hertogenbosch with more than 4000 employees and 240 medical specialists. Almost all different specialties were represented within the hospital and it provides more than 500,000 outpatient visits and 60,000 hospital admissions on a yearly basis. The service area of the JBZ consists of around the approximately 635,000 people and last year, the hospitals annual turnover was over 330 million euro’s. These numbers make that this hospital is regarded as one of the largest non-academic hospitals in the Netherlands. The organizational chart of the JBZ is included in appendix C.

(28)

As Yin (1984) explains, case studies can be based on quantitative data only, qualitative data only or both, depending on the circumstances. However, qualitative data are useful to provide rich descriptions and understanding of certain phenomena within a certain context (Sofaer, 1999). In line with this reasoning, qualitative data were the appropriate data type for this study. Typically, semi-structured interviews are most common when doing explanatory case study research (Saunders et al., 2009). Likewise, this research made use this kind of interview in order to be able to collect information on predefined general areas from each interview (McNamara, 2009). The advantages of the increased flexibility when using semi-structured interviews greatly outweighed the forego possibility of performing statistical analyses, which is the case with more structured interviews (Horton et al., 2004). To ensure reliability of the interview data, an interview protocol was used during the data collection process. The interview protocol can be found in appendix D. Next to the interviews, data were collected by means of informal conversations, observations, internal documents and different online sources. To enhance the validity of the research, multiple sources of data were used during the data collection process, also known as, triangulation (Yin, 2003). In addition, one of the key informants of the research reviewed the draft version of the results in order to aim for even more validity.

Separate interviews were conducted with four medical specialists of the department. In addition, one interview was conducted with the unit manager of the dermatology department and one with the chief of the ‘results-accountable unit’ (in Dutch: Resultaat Verantwoordelijke Eenheid, RVE) together with the quality manager of this unit. All six interviews were recorded in order to allow for transcription and to enable the researcher to concentrate on the things being said and on non-verbal ques. Interview data were gathered during a time period of two weeks. The interviews typically had a time frame of 45 minutes and were held in the Dutch language. Information regarding the interviews can be found in the table 1. With regard to the other sources of data, we received those predominantly in digital form. It gave us valuable insights in the way in which the PM’s and their resulting data was used throughout the department. We used this information in advance to prepare for the interviews and to validate our findings and results.

Table 1: List of persons interviewed at the JBZ

Date:& Funtion: Code: Duration:

05#May#15 Dermatologist MS1 454m4

06#May#15 Unit4manager MA1 554m4

06#May#15 RVE4manager4+4Quality4manager MA24+4MA3 454m

11#May#15 Dermatologist4(also4medical4manager) MS2 304m

11#May#15 Dermatologist MS3 404m

(29)

3.4 Data Analysis

After the recordings from the interviews were transcribed, written notes from the interviews were added to the transcriptions. The methods described in the paper of Lacy & Luff (2001) were used to analyse the data that was collected. Firstly, the data was organized by giving comprehensive codes to the interviews and interviewees (see table 1). In order to further guarantee the anonymity of the interviewees, original recordings were deleted and safely kept on a USB stick. The second step included the familiarisation with the data. All transcripts and other data collected was re-read and the first cautious insights were written down. After this, preliminary codes were given to the transcripts. These codes are consistent phrases, expressions, or ideas that were common among research participants (Kvale, 2008). This was done in the qualitative analysis program ‘NVivo’. During the coding process, more specific and detailed codes began to emerge and themes and categories were identified. These themes and categories were used to identify the most emerging subjects in the research and were the main basis for the formation of the results in this study.

Chapter 4. Results

In this chapter the results of the interviews, observations and informal conversations are presented. Findings from the literature review are (dis)confirmed with support of quotes of the interviewees. The first section presents a description of the general variables of the conceptual model. The second section continues with a presentation of the findings regarding the relations between these different variables. The third section presents the conclusion of the findings resulting from the research. At last, the fourth section focuses on some unexpected results that were found and discusses these shortly.

4.1 Description of the variables from the research question

This section presents the information found regarding the variables of the conceptual model. The first part of this section discusses the findings on the healthcare sector while the second part focuses on the PM’s that were in the JBZ. The third part continues with the findings regarding the main organizational goals of the JBZ. The last three parts discusses the Triad dimensions in more detail. 4.1.1 Healthcare Context

As indicated in the literature review, the transformation process in the healthcare sector is often not clear and performance hard to measure. This was confirmed during the interviews and three out of four specialists mentioned that performance, especially in terms of quality, was extremely hard to measure. They found efficiency to be slightly more measurable since is easier to be expressed in numbers. In the case of performance of quality, one specialist mentioned.

(30)

In addition, both managers and specialists stressed that these days, there is an enormous increased pressure on the costs within the hospital. Especially the specialists expressed their concern in the increased power of the healthcare insurers. They have to do more, with less financial resources and this was not always easy for them.

‘’I really want to try to help as many people as possible even when nowadays we have limited resources. Sometimes it can be a problem that these days everything has to be cheaper and cheaper.’’ – MS1

4.1.2 PMS

From the interviews it became clear that the JBZ and the organizations evaluating the performance of the JBZ use multiple different PM’s to evaluate the performance of individual specialists, medical units or even the whole hospital. These PM’s all have a different scope and are used by different means within the hospital. Three main features are used to distinguish the various PM’s from each other: the evaluation level, by who the PM is deployed, and last, the source of information. This is explained in more detail below.

The PM’s in the JBZ were used to evaluate the performance on four different levels: individual, unit, process and hospital. Individual level PM’s aimed to evaluate the performance of individual specialists. Furthermore, the unit level PM’s focussed on evaluating the performance of a whole unit, in our case the dermatology unit of the JBZ. The process level PM’s evaluated the process by which the patients were treated. This could thus be for example the process of how dermatologists treat a patient with a melanoma. Lastly, hospital level PM’s evaluated the performance of the hospital as a whole by allowing comparing between hospitals.

The second way to distinguish between the different PM’s was by focussing on who deployed the PM. There were four possibilities for this: the hospital itself, national institutions and the dermatology association. Hospital PM’s were deployed by the hospital itself and compared only results within the hospital. Nationwide institutions PMS compared the evaluation on one of the above-described scope levels trough the whole nation. This could be for example trough comparison made by the press, the government or hospital associations. Furthermore, PMS deployed by the Dutch dermatology association compared the evaluations of dermatology departments between different hospitals.

(31)

most prominent. Patient sourced PM’s used the information given by patients regarding their experiences in the hospital. The second type of PM’s used general KPI’s to evaluate the performance. This could be the general financials but also other data, such as numbers of patients treated and incidence of complications. The third type, the colleague based PM’s, made use of the opinion of peer colleagues.

More information regarding the different PM’s that were used in the JBZ is given in table 2. It is important to realise that there was a clear division between the more general and less detailed PM’s on the one hand and the more specific and detailed PM’s on the other. The first four PM’s indicated in table 2 made use of specific information about individual specialists or treatments. These measurements generated a rich source of data that was used for improvement activities and evaluation. In contrast, the last five PM’s were based on more general aggregated indicators of performance. They indicated for example the general satisfaction of visits or the amount of costs made for certain treatments. Often they were based on standard KPI’s resulting in much broader and less detailed information. As is discussed below, this had several implications for how they are used in the hospital.

One of the first things that emerged was the difference in attitude towards the (in)completeness of PMS within the hospital. The managers indicated that, in contrast to the past, the real performance of medical specialists was nowadays almost perfectly captured in the PMS. In the eyes of the managers the PMS was very much complete and thus accurately reflects the real performance of specialists.

‘’I think they do reflect the performance quite good. Of course not totally, that’s utopia. But I think that in these days we definitely reach the 90%. When I compare this with the past, we even didn’t meet a 50%. A real improvement as you will understand.’’ – MA1

(32)

Table 2: Different PM’s in the dermatology department of the JBZ

‘’When you would have a doctor who functions extremely badly, this will come forward in several measurements. However, that is when we talk about extreme situations. The difference between a 7 or a 9 [on a scale of 10] is not easily seen in the PMS.’’ – MS3

This was especially problematic in the broader en and less specific PM’s deployed by parties outside the hospitals. These were based on general KPI’s in order to allow for comparison and as a result, they were often based on vague indicators that were easy to manipulate. There was even sometimes discussion between the specialists, management and the initiators of these PM’s (for example the Dutch newspaper Algemeen Dagblad) about the relation between the measured indicators and the goals of hospitals. It seemed thus evident that these PM’s did not completely measure the performance of specialists, departments or hospitals.

PM:

Medical Incidents

Appraisal & Assessment 360

Patient Satisfaction Survey

Mirror Talks

ZorgkaartNederland.nl

Self Evaluation list Hospital KPI's PKKK Press PM's

PM Evaluation level: Deployed by: Source of info: Times p/y:

Medical Incidents Individual Hospital Patient When it occurs

Appraisal & Assesment 360 Individual Hospital Patient + Colleagues 0.5 (once in two years)

Patient Satisfaction Survey Indivudual + Unit Hospital Patient 2

Mirror Talks Process Hospital Patient 3

ZorgkaartNederland.nl Individual + Hospital National institutions Patient + KPI's N/A

Self Evaluation list Individual Dermatology Association Specialist 1

Hospital Indicators Unit + Hospital Hospital KPI's 4

PKKK Unit Hospital KPI's 3

Press PM's Hospital National institutions (press) KPI's 1

The (non)financial indicators about for example costs, quality and efficiency. The more general KPI's that the management of the hospital uses to asses performance.

Meeting with the board to discuss the production, costs, quality of care and the quality of labour.

Initiatives from the press (mostly newspapers and medical magazines) such as the 'AD top 100'. They use general KPI's to compare them amongs eachother.

Description:

This is a mandatory registration in which you have to register when a patient has had a complication. This could be a small wound dehiscence but also life threatening complications. The bigger things will be investigated by the hospital.

A kind of year review for medical specialists that is initiated from the hospital. Around 10 people who worked close with the specialist are being asked to give 3 good and 3 improvement points and some other questions. The specialist gets feedback on the results with focus points for the next two years.

A 2 times per year survey which is filled in by the patients of the department. The survey contains questions about how the patients experienced the encounter in the hospital and their specific treatment.

An initiative from the hospital that discusses a certain treatment, for example, a treatment of a melanoma. Around 10 patients are invited to discuss what went well and what improvement need to be made.

An online platform on which patients can evaluate their medical specialist on multiple indicators and give comments. This way, specialist performance can be compared across the country.

How performance measurement systems in the healthcare sector influence the motivation, capacity and opportunity of employees to perform goal congruent behaviour: A qualitative approach