Management in the South African public health sector: An x-inefficiency perspective

Abstract

Inefficiency in the South African public health sector contributes significantly to the country's relatively poor health outcomes, yet it is poorly understood and overshadowed by health care financing and payment issues. This paper explores this situation from the perspective that the public health sector forms part of a complex adaptive system from which inefficiencies emerge endogenously. Leibenstein's seminal work provides the basis for a conceptual framework that explores and describes management in the public health sector from an x-inefficiency perspective. Further research into x-inefficiency and its underlying causes may be useful to guide the design of appropriate health policies in the South African context. Unless reforms targeted at improving x-efficiency are prioritised, increased revenue allocation to the public health sector will have a sub-optimal impact on health outcomes.

Keywords:

• complex adaptive systems
• human behaviour
• rules and institutions

1. Introduction

Despite the allocation of an increasing share of general government expenditure to health as a priority area since 1994 (Black et al., 2008:91), South Africa has failed to improve health outcomes significantly. In fact, the most critical health indicators are worse than those of comparable middle- and low-income countries that spend substantially less than 8% of their GDP on health care. The South African health system's failure to adequately address the challenges of universal coverage, quality and cost results in poor outcomes relative to total expenditure. This implies that a degree of inefficiency is at play within the system. Part of the challenge has been to redistribute scarce resources more equitably among a growing population. A further challenge has been the increasing quadruple burden of disease (i.e. emerging chronic diseases, poverty-related diseases, the HIV/AIDS pandemic, and injuries), exacerbated by the HIV component, and sub-optimal access to social co-determinants of health, such as water provision and waste management. The crisis in the public health sector is multi-faceted in its origin, and therefore the solution should be too.

The health sector forms part of a complex adaptive health system in which many interconnected participating agents continuously interact and adapt (Plsek & Greenhalgh, 2001:625). A complex adaptive system (CAS), by definition, is unpredictable and non-linear (Kernick, 2004), rendering precise quantification impossible and relative quantification challenging. Still, mainstream economic thinking prevails, resulting in most efficiency studies being quantitative and measurement focused. Studies measuring inefficiency, while useful in establishing the relative degree of inefficiency, provide limited contributions to x-inefficiency theory since they cannot capture underlying dynamic real-life phenomena.

In contrast, Leibenstein's x-efficiency theory offers qualitative insights mainly aimed at a deeper understanding of how participating human agents make decisions. Realistic and relevant aspects of human behaviour and institutions are incorporated into this approach, signifying a departure from the assumptions attached to neoclassical theories. This paper uses Leibenstein's approach to inefficiency, finding it appropriate and useful for analysing a CAS.

The South African Government has placed great emphasis on redressing the financing and payment mechanisms of the current system in order to achieve universal access to quality health care. While important, this specific focus overshadows critical issues pertaining to public health sector inefficiency. This paper therefore explores the x-inefficiency paradigm arising from public sector management weaknesses and the way it affects the behaviour of participants. Management problems are explored from a perspective that views the public health sector as part of a CAS from which inefficiencies endogenously emerge due to the interaction between participating agents operating within the constraints of rules and institutions. It highlights the benefit of viewing inefficiency in the public health sector from this perspective, and in so doing intends to stimulate debate and further research into the neglected area of x-inefficiency.

Failure to understand the causes of x-inefficiency and its contribution to the health crisis will result in inappropriate policy design. As long as management inefficiencies are not thoroughly addressed, increased revenue redistributed to the public health system will have a less than optimal impact on health outcomes.

2. Methodology

This paper takes a qualitative approach. We begin by distinguishing between x-efficiency1 and other types of efficiency. This is necessary since x-efficiency is often misunderstood and used in the wrong context. We describe the characteristics of a CAS and concepts linked to it, contextualise the viewpoint that the health sector forms part of a CAS, and highlight the link between Leibenstein's theory and modern-day complexity theory.

The x-(in)efficiency theory implicitly captures aspects of behavioural and public choice theories, and explicitly focuses on the individual agent and the degree of effort the agent exerts when faced with information and incentive challenges within the constraints of rules and institutions. We review Leibenstein's (1966) seminal work on x-efficiency and locate it within the broader context of complexity and complementary theories. In so doing, we attempt to construct a holistic understanding of x-inefficiency as an undesirable phenomenon that emerges from complex adaptive systems, such as health systems. This in turn provides the conceptual framework that we use to explore and interpret management inefficiencies in the public health sector in the South African context.

It should be noted that the available literature on x-inefficiency is minimal compared with literature related to other types of efficiency. This is largely because the mainstream economic rationale favours modelling and quantification of phenomena. Furthermore, most of the research cited in this paper was conducted in the international private sector. The paper is thus mainly theoretical in nature, containing very few examples in the local context. This highlights the need for further research on x-inefficiency in the public sector, particularly the South African public health sector, given the extent of the planned financial expenditure in the medium term.

The scope and qualitative nature of this paper does not permit a very detailed exploration of all the factors that contribute to x-inefficiency in the public health sector. The focus has therefore been limited to exploring inefficiencies arising from management in the public health sector.

3. Conceptual framework

3.1 Allocative and technical efficiency and x-efficiency

Economic efficiency is typically defined in terms of allocative and technical efficiency. Allocative efficiency describes a situation where scarce resources of a country are distributed optimally according to society's preferences, while technical efficiency describes a situation in which available resources are employed in the most cost-efficient manner (Black, 2008:15–16, 20). Research into these types of efficiency is important and they cannot be ignored. However, this paper focuses on x-efficiency, which is closely linked to technical efficiency since it too describes the effectiveness with which a set of inputs can produce outputs (Leibenstein, 1966). Consequently these terms have many related features even though there are important differences between the theories underlying technical and x-efficiency (Button & Weyman-Jones, 1992:439).

Leibenstein (1977:312) highlights fundamental differences between technical efficiency and x-efficiency, with the most critical difference stemming from the way the problem of inefficiency is viewed. Technical (in)efficiency suggests that the problem is technical in nature, related to input techniques, while x-(in)efficiency views the problem as being intrinsic to the nature of human behaviour. This key distinction is evident in the objectives of work done by Farrell (1957) on technical efficiency and Leibenstein (1966) on x-efficiency. Farrell focused primarily on the measurement and practical computation of productive efficiency. Furthermore, his work was conducted within a neoclassical economic modelling framework and therefore excluded real-world characteristics such as human behaviour and institutions (Beinhocker, 2006:116). In contrast, Leibenstein's x-efficiency theory offers qualitative insights into inefficiency that are mainly aimed at gaining a deeper understanding of how participating human agents make decisions.

3.2 A complex adaptive system

A health system, by definition, is inherently a complex adaptive system (Smith & Hanson, 2012:13). The public health sector forms part of the health system, so a complexity approach can be employed to explore dynamics within the South African public health sector. This approach necessitates taking a less positivist attitude towards what is generally considered useful knowledge (Lessard, 2007:1754). While health system frameworks have advanced in embracing the relational nature of health systems and their dynamic complexity, most evidently in the frameworks designed by Frenk (1994) and Brinkerhoff & Bossert (2008), little guidance is provided in terms of incorporating complexity into efforts to strengthen health systems (Gilson, 2012:37).

A CAS is made up of many relatively independent agents that are interlinked and interact and can change their actions on the basis of events occurring in the process of interaction (Kochugovindan & Vriend, 1998:55). The adaptive ability of a CAS can be likened to what evolutionary theorists term ‘niche construction’, which refers to the way organisms evolve and in so doing affect their environment, which in turn affects their evolution (Leland et al., 2000; Bowles & Gintis, 2002:15). Based on this understanding of complex adaptive systems, complexity approaches view macro patterns as emergent phenomena that arise endogenously from the interactions of economic agents and their environment within a system (Beinhocker, 2006:167). The subsequent hypothesis is that emergent phenomena are likely to arise from three origins: behaviour of participants, institutional structures of the systems and exogenous inputs into the system (Beinhocker, 2006:185).

Inefficiency in the South African public health sector can thus be viewed as a problem generated by the system itself. The behaviour of all the participating agents operating within an institutional structure creates and influences x-inefficiency. Much of Leibenstein's (1966) seminal work on x-efficiency implicitly captured these aspects of complexity as he sought to understand the root causes of inefficiency.

The next section looks at the most important and relevant assumptions of x-inefficiency theory (Leibenstein, 1979) and how it ties in with complex adaptive systems.

3.3 The emergence of x-inefficiency

Leibenstein proposed that the individual, and not the organisation, is the basic decision unit. This includes individuals making decisions on behalf of organisations, such as facility managers or hospital board members. This assumption is in keeping with complexity theory, where human behaviour plays a central role in influencing patterns that emerge from a system. Complexity theory ascribes inductive rationality to human behaviour, rather than perfect rationality (Beinhocker, 2006:117), in other words it views the human decision-making process as being influenced by a variety of internal and external forces, rather than automated responses made by the perfectly rational neoclassical ‘robot’. Simon's theory of bounded rationality (1992) states that humans operate in a world of information asymmetry using their large but constrained intellectual powers. He further postulated that human beings ‘satisfice’,2 an action that involves making the best possible decisions with the limited information available. Studies by early behavioural economists confirm his theory that people do not behave in the ways assumed by traditional economists (Beinhocker, 2006:119).

Parallel reasoning led Leibenstein to assume that individuals are non-maximisers and that their behaviour is dependent on a balance between internal and external pressures. Internal pressures are subject to a balance of opposing motivating forces: ‘constraint concern’, such as conscience and the desire for approval, versus ‘animal spirits’, such as the urge to avoid responsibility. Frequently, the equilibrium point of internal and external pressures is evident in choices made according to habit, social norms, rules, institutions or ethics. This implies that the behaviour of each individual agent, or employee, will depend on the particular combination of all these factors.

In addition, Leibenstein assumed that all individuals have some discretion with respect to the nature and amount of effort delivered, implying that incomplete employment contracts are at play. Organisations normally purchase labour in units of time, not in units of output or effort, and therefore employment contracts are considered incomplete. Individuals also face incomplete information in terms of their job expectations, necessitating self-interpretation of their jobs when determining their effort. The individual's satisfaction derived from effort is the sum of the satisfaction derived from the income earned from the job and effort put into it.

Leibenstein's model assumes that the individual prefers some effort to no effort, but a point is reached where additional effort becomes less satisfying and results in a decrease in total satisfaction (Tomer, 1981:353). The selected effort is therefore unlikely to be in the best interests of either the employee or the employer. Furthermore, Leibenstein hypothesised that inert areas exist within specified parameters. This implies that a selected effort position linked to a certain level of satisfaction will not be changed easily even if a different position is associated with a higher level of satisfaction.

Leibenstein assumed that if external pressure (for example from peers, authority, or institutional pressure and incentives) is strong enough, it can steer behaviour towards approximating optimisation. He acknowledged that even though certain personality types achieve optimising behaviour most of the time, the model assumes that this is not generally the case. Therefore the lower the degree of constraint concern, the greater the degree to which the effort position will reflect self-interest and the less it will reflect loyalty to the organisation's interests (Leibenstein, 1978:30). This behaviour by employees, acting as agents on behalf of the organisation's principals, is referred to as the principal-agent problem. Both principals and agents are opportunistic: agents seek to minimise the production and transaction costs and maximise the benefits, while principals seek to maximise their benefit to the detriment of the agents. Unless they are operating in an environment where external pressure is applied, agents may engage in behaviour such as rent seeking, cheating, breach of contract and incomplete disclosure. The degree of such opportunism differs very widely depending on the work environment and cultural setting (Preker & Harding, 2000:10). Complexity studies conducted by Stanley et al. (1994) have demonstrated how the emergence of undesirable opportunistic behaviour is altered depending on the ‘rules of the game’. This highlights the importance of rules and institutions in complex environments where participating agents are influenced by incentives and available information. Rules and institutions should thus be carefully constructed, and relevant to the context and level of worth ethic, in order to influence human behaviour in a favourable way. No ‘one-size-fits-all’ approach can be adopted when constructing the appropriate rules and institutions.

Leibenstein's model can be used to predict how employees will behave in an environment where the rules are more explicit. Such environments could take a multitude of forms, such as combinations of structures that are able to enforce punishments and rewards, in the form of either social approval or monetary incentives. Employees operating in such environments are expected to adopt a higher effort position, thereby increasing productivity and subsequent output. Efforts to create such environments are thus symptomatic of management choosing to put pressure on employees to perform better. Leibenstein concludes that the application of external pressure is the solution to worker productivity. This implies a focus on rule and institution changes to influence worker behaviour.

Leibenstein noted that even though individuals respond to pressure by increasing their constraint concern, they find this extra effort uncomfortable. Eventually individual workers or groups induce others, through work slowdowns, resistance, or union activities, to return to the previous level of constraint concern. Alternatively stated, x-efficiency theory does not maintain that management can shift long-run employee effort levels at will. The larger the group, the larger the inert areas, and the more difficult it will be for management to induce higher levels of productivity. Peer pressure plays a significant role in influencing an individual's behaviour, and thus the effectiveness of the organisation is dependent on peer interaction and the relationship between employees and management at all levels. This is characteristic of complex adaptive systems.

The more an organisation is sheltered from competition, such as government departments, the more likely it is that x-inefficient effort positions will be chosen by employees and managerial staff (Leibenstein, 1979:18). Standard neoclassical theory assumes that organisations operate at the minimum cost level, because it also assumes a perfectly competitive market, requiring no additional external force from management to maintain the minimum cost level. However, this assumption is not valid in x-efficiency theory, which views costs as a coiled spring: unless sufficient external pressure is applied to change employees' effort positions, the coiled spring will uncoil, in other words costs will shoot up. Managers of sheltered organisations have to exert greater effort when applying external motivating incentives (whether rewards or punishment) and be fully competent at performing this function. Since public sector managers act as agents of government, this critical management function may not be executed adequately, as implied by the principal-agent theory.

Tomer (1981:354) expands on Leibenstein's theory by adding a work environment variable and a future growth variable, and expanding the personality variable. The work environment variable comprises four dimensions: the match between the individual and the characteristics of the job and the organisation; the existing structure and supervision of the job; the existence of clear, meaningful goals of the job and organisation; and the nature and enforcement of implicit contracts (Tomer, 1981:354). The future growth variable represents the worker's perception that the job will lead to growth opportunities. The expanded personality variable includes drive, maturity and psychic health, i.e. Maslow's (1943) concept of self-actualisation. The size of the personality variable will determine the extent to which an individual may be motivated.

These variables are supported by motivation theory, in particular, Herzberg's (1968:55) need-satisfaction theory which hypothesises that a motivated worker is an energised and goal directed individual who wants to work and does not require external pressure to do so. Although motivation is an internal force, research suggests that motivation to do a job well is not automatic, but comes from a correspondence between basic needs and the job situation. Herzberg lists achievement, recognition of achievement, the work itself, growth, advancement and responsibility as motivating factors intrinsic to the job.

Complexity theorists turn to evolutionary psychology to gain insights into what motivates human behaviour. From an evolutionary perspective, humans exist to fulfil the sole purpose of gene replication (Beinhocker, 2006:314). This school of thought concludes that present-day human behaviour is a combination of nature and nurture; in other words, inherent survival and reproductive instincts passed down from ancestors interact with ever-changing local environments to which humans adapt. Factors motivating human behaviour are thus not only intrinsic but also depend on the context in which humans operate.

In summary, Leibenstein's theory provides a useful basis from which to analyse the fundamental causes of x-inefficiency. When combined with complexity and worker motivation theory, a more comprehensive understanding of x-inefficiency is constructed.

Working from the premise that the public health sector forms part of a CAS, x-inefficiency can be viewed as an undesirable side effect that emerges endogenously because of the interaction between participating agents who are trying to ‘satisfice’ within the confines of rules and regulations. Participating agents are individuals who are motivated by internal and external factors, with non-maximisation being the norm rather than the exception. The implication for governments, who cannot change human behaviour directly, is that rules and institutions are vital for indirectly making agent behaviour more efficient.

4. Exploring management inefficiencies in the public health sector

In management theory, human capital is viewed as the most important of all assets because of its pivotal role in the manipulation of all other inputs to make an organisation productive (Akpan, 2011:159). This is evident in the organisation's strategic decisions to attract the desired labour (Saltman & Figueras, 1997). Inefficient management in any organisation thus demands urgent attention since ‘managers determine not only their own productivity but also the productivity of all cooperating units in the organization’ (Leibenstein, 1966:397). It thus follows that management inefficiencies should be identified and corrected before addressing worker inefficiencies since the reverse would be ineffective and counter-productive.

In the following sections we narrow the focus to issues specific to management inefficiency in the South African public health sector (while noting that management is not the only factor contributing to inefficiency in the sector). We explore management in the public health sector from an x-inefficiency perspective in the context of a CAS framework, as described in the previous section. It becomes difficult to separate the agent behaviour, i.e. the human element, from the rule and institutional elements since they are interlinked, with causality running in both directions simultaneously. However, for the sake of structure and cohesion, we look at management x-inefficiency from the perspective first of human behaviour and then of rules and institutions.

4.1 The human manager

Managers are people, not robots. While people possess strong adaptive inductive abilities, they are nevertheless vulnerable to human error, such as biased judgements and interpretations, generalising, intuitive reasoning and poor risk-assessment ability (Beinhocker, 2006:122). Mental strengths and weaknesses, as well as an array of emotional characteristics influenced by internal and external factors, all combine to cause the behaviour displayed. Since managers are people, they are capable of displaying any behavioural pattern typical of participating agents operating in a CAS, including x-inefficiency.

International evidence suggests that one of the crucial factors in the state's ability to achieve its goals is the people who work in the public sector; public servants, especially managers, must therefore have the skills and motivation to do their jobs well (NPC, 2011:21). Problems arise when managers, lacking skills or experience, cannot manage their staff. Research conducted at the primary health care (PHC)) facility level found that the lack of training for appointed staff, inadequate systems (e.g. for staff recruitment) and the relatively low rank of facility managers make managerial positions at the facility level unattractive to qualified managers (Leon et al., 2001:211). Similarly, research conducted at a district hospital level shows that management teams have not been sufficiently prepared and supported in the execution of health sector reforms (Penn-Kekana et al., 2004). This problem can be overcome with education, training and the selection of suitably qualified managers. However, the problem becomes more difficult to deal with when managers are skilled and competent but still choose not to optimise productivity.

The x-efficiency theory assumes that individuals are non-maximisers and will choose a sub-optimal level of effort. This assumption is compounded by the principal-agent problem, since managers play dual roles: they are principals of their employees yet agents of their employers. Managers are therefore highly vulnerable to engaging in inefficient behaviour. And since managers operate not in isolation but within a complex adaptive system where the behaviour of one agent affects the behaviour of others, this type of inefficiency may have negative spillover effects on colleagues.

Should a point be reached where the majority of management has adopted a lower effort level, replacing the entire management team is a theoretically viable option to take the organisation to higher effort and production levels. Empirical studies conducted in the private sector suggest that replacing the entire management team improves performance more effectively than replacing only the chief executive (Dailey & Schwenk, 1996). This corroborates Leibenstein's (1979) theory that larger groups have larger inert areas, making achievement of higher productivity levels more challenging. The effects of peer pressure in an inefficient management culture should therefore not be taken lightly.

4.2 The importance of leadership

Although management and leadership are closely linked, a distinction should be made between these two concepts. Leaders can inspire and motivate others, whereas managers may be able merely to maintain the status quo within a work environment without providing any form of leadership. Since worker motivation plays such an integral role in worker productivity, it becomes essential for policymakers to fully comprehend and address leadership issues when developing policies for increasing organisational efficiency.

The wealth of recent literature on the theoretical consequences of leadership in public organisations is mainly qualitative, with only a few empirical tests of the performance effects of leadership styles and skills (Boyne, 2004:112). Empirical evidence shows that leadership makes a difference to public service performance. In addition, the impact of leadership change is dependent on previous levels of performance, and is likely to be particularly noticeable in organisations that have been under-performing for a long time (Miller, 1993). Two areas that have yet to be explored are the difference between internal and external appointments and the effects of leadership change on performance (Boyne, 2004:112).

In South Africa, hospital managers do not function independently but rely on leadership and support from provincial and national management structures, i.e. institutional support (Harrison, 2009:2; Engelbrecht & Crisp, 2010:204). The current PHC work environment often lacks such support, and where it is present, decision-making processes are sluggish (NPC, 2011:22). This makes for demotivated and less-than-efficient staff (Gilson et al., 2004:63) – a finding that is unsurprising, given the link between motivation and efficiency as described by Leibenstein.

While it is hard to measure this objectively, it appears that leadership is weak in the public health sector in many provinces and municipalities. Conclusive evidence of unethical behaviour, i.e. corruption, fraud, nepotism and cronyism, may be considered a reasonable proxy indicator of poor leadership. This finding can be explained using Leibenstein's theory that employees, acting as non-maximising agents, will exploit systemic weaknesses in the work environment to the detriment of the principal, the employer. Thus public employees will exploit poor leadership, and as long as poor leadership persists it will be naïve to expect significant improvements in their productivity. Leadership change, where needed, is thus of critical importance.

Lack of effective leadership has a major negative impact on x-efficiency. Since weak leadership is harder to identify than weak management, solutions to this problem are more complex to develop. Weak leadership is exacerbated by political interference and serious thought needs to be given to transforming public system operations into a technocrat-led system.

4.3 Pushing boundaries: Rules and institutions

A comprehensive body of literature on the performance of the private sector suggests that institutions influence behaviour and that different environments require different structures to work best (Boyne, 2004:104). The reasoning is that rules and institutions need to be designed according to the context in which participating agents, such as managers, operate. The health system framework designed by Roberts et al. (2008) embraces this approach by identifying regulation and behaviour as two of the key ‘control knobs’ that influence efficiency.3

Leibenstein prescribes the addition of appropriate incentives to work environments, whether punitive or remunerative, to achieve higher levels of efficiency. This can be achieved by fine-tuning rules and institutions. Such measures would limit managers' ability to exploit loopholes in the system.

Incomplete contracts allow individuals to use their discretion when choosing the kind of work they do, the pace at which they work and the quality of their work (Leibenstein, 1979). To mitigate the adverse effects of incomplete contracts, performance expectations should be established at the start of a contractual relationship (Preker & Harding, 2000:16). However, this is problematic in the South African public health sector since two independent authorities, the Department of Public Service and Administration and the National Department of Health, manage different aspects of the same contract, the former determining the post level and conditions of service for PHC workers and the latter setting the performance standards for the quantity and quality of their work (Chabikuli et al., 2005:110).

Employment contracts should be more explicit in order to limit the discretion used by managers when selecting their effort level. This would curtail the self-interpretation of job descriptions which leads to lower effort choices. Possible improvements to contractual agreements include shortening the service term and detailing outcome measures. Such interventions would hold workers accountable to quantifiable performance targets. Care should be taken to ensure that these targets are realistic and achievable, since increased work pressure which fails to improve outcomes may demoralise staff and decrease productivity in the long run.

Empirical studies come to the general conclusion that a performance culture, achieved by instituting greater competitive measures, is associated with better service outcomes in the public sector (Boyne, 2004:111). This is in keeping with Leibenstein's (1979) theories on x-efficiency in sheltered organisations. In 2001, the routine, years-of-service based promotion of civil servants was replaced by a system where supervisors motivate for the employee to be promoted or awarded a bonus. However, the absence of contractual performance indicators and targets makes this method of decision making subjective and vulnerable to nepotism, cronyism and favouritism (Chabikuli et al., 2005:111).

In recent years the public sector has attempted to institute best-practice management techniques from the private sector to replicate an incentive environment (Osborne & Gaebler, 1993). A study conducted in Ghana by Akazili et al. (2008:149) concluded that incentives were a key motivating factor for the promotion of efficiency in Ghana's health care system. Clinical staff receiving incentives were nine times more likely to be efficient than their counterparts who received no incentives (Akazili et al., 2008:154). Leibenstein's (1979) suggested remedy for x-inefficiency included the use of social or monetary incentives, which would motivate workers to reach higher levels of productivity. Incentives form part of the external measures Leibenstein (1979) prescribed to persuade individuals to perform optimally.

5. Conclusion

This paper views the public health sector as part of a complex adaptive health system, and thus views the interaction between participating agents, operating within a context of rules and institutions, as the source of emerging factors such as inefficiency. The inefficiency of public health care managers in South Africa corroborates Leibenstein's theory that individuals do not maximise productivity, but instead exercise discretion when selecting the amount of effort to put into work. Weak management, including a lack of leadership, are thus major contributors to x-inefficiency in the South African public health sector. This has a negative impact on public health service delivery and, consequently, health outcomes. Increasing the x-efficiency of public health managers should therefore be a major policy focus and take priority over correcting the x-inefficiency of public health workers.

While governments cannot change agent behaviour directly, they can influence participant behaviour indirectly via rules and institutions, i.e. by changing the work environment. Employees require a functioning framework of rules and institutions in which accountability measures and incentives foster productivity and thus optimal output. An increase in public management efficiency is of paramount importance if public health services are to be delivered as efficiently as possible. Failure to address management issues adequately leaves little hope for improvement in the efficiency levels of health care workers.

With a largely x-inefficient public health sector, South Africa has significant room for improvement in outcomes should x-inefficiency diminish. More qualitative research is therefore required on the emergence of x-inefficiency in the public health sector to gain a deeper understanding of the nature, extent and cause of this problem. Such research would be useful to guide policy design, which should be structured to induce participating agents to evolve away from x-inefficient behaviour.

One of the main roles of government is to allocate scarce resources efficiently. Yet inefficiency in the public health sector, a form of government failure, further amplifies generic government failures such as rent-seeking, and political and bureaucratic failure. Reforms targeted at improving x-efficiency must take priority over other reforms that can only succeed in an x-efficient environment. Failure to reduce x-inefficiency dramatically before implementing reforms that require a significant increase in health budgets will simply perpetuate x-inefficiency in the public health system, at the expense of the taxpayer and society. More than a mere increase in inputs, the public health sector requires an improvement in the level of efficiency with which such inputs are converted into health outcomes. Government focus must therefore be on solidifying the public health sector base, by improving its efficiency, before it is built upon and expanded.

Acknowledgements

The authors wish to thank Elizabeth Stoltz, of the University of the Western Cape and Stellenbosch University, and Servaas Van Der Berg, of Stellenbosch University, for their helpful comments, advice and moral support.

Notes

¹A term coined by Leibenstein, x-efficiency is an open-ended concept which describes the effectiveness with which a set of inputs can produce outputs. The ‘x’ captures all instances of non-allocative inefficiencies which arise from activities within the organisation (Leibenstein, 1979:14). Some studies refer to x-(in)efficiency but actually capture technical (in)efficiency.

²A portmanteau word combining ‘satisfy’ with ‘suffice’. See www.realtech.co.za/realwiki.php?title=Satisfice Accessed 29 July 2012.

³Roberts et al. (2008) identify five ‘control knobs’: financing, payment, regulation, organisation and behaviour.