Developing administrative capacity: an agenda for research and practice

Abstract Administrative capacity is necessary for achieving policy success and preventing policy fiascoes. However, a fragmented literature, sparse empirics, and a focus on developed countries have led to a gap in examining how such capacity should be measured and built; particularly in developing contexts. This paper assesses administrative capacity frameworks, indices, and reforms with a specific focus on the organizational-operational dimension that is crucial to effective implementation, in order to provide recommendations for advancing the theory and practice of capacity building. Our observations suggest that both scholars and practitioners need to be more cognizant of the fact that capacity building is a continuous process, requiring specific assessment and measurement of the components of the organizational-operational dimension. Such components include enhancing administrative autonomy and competence, and arrangements incentivizing coordination and collaboration, while safeguarding adequate control and oversight.


Introduction
Capacity building has become one of the most prevalent concepts in the literature about public policy and development. Different concepts are used somewhat intertwined, and with different dependencies, such as state capacity, policy capacity, managerial capacity, and administrative capacity (Howitt 1978;Ingraham and Donahue 2000;Painter and Pierre 2005;Wu, Ramesh, and Howlett 2015). Wu, Ramesh, and Howlett (2015) distinguish three necessary competency types; analytical, operational, and political capacity, applying to three levels; the individual, organizational, and systematic (see Table 1). According to them, capacity results from the combinations of skills and resources at each level, while the interplay between different competencies can explain cases of policy failure or success.
Following their terminology, this paper zooms in on organizational-operational capacity, required of government agencies as well as non-government and private organizations engaged in the operations of implementation and delivery. Effective and rigorous implementation and the capacity to actually deliver on policy promises (e.g. Barber, Moffit, and Kihn 2010;Barber 2015) is one of the key elements distinguishing high performing countries from those that show mediocre performance (Everest-Phillips 2018). Organizational-operational capacity is closely related to administrative capacity as defined by Painter and Pierre (2005, 2): the ability of the government to manage its human and physical resources to deliver on its objectives. It focuses specifically on how policy implementation and delivery are organized, and how agencies and networks and their managers, responsible for delivering on policy promises, operate and perform.
The aim of this paper is to synthesize literature about these dimensions of capacity, in order to generate theoretical and practical knowledge for building more capable public sectors, particularly in developing countries. This is paramount since the current state of the literature, according to Wu, Ramesh, and Howlett (2015), lacks consistent gathering of empirical research that can help measure, build, and bridge capacity gaps in different sectors and contexts.

Why capacity matters
Capacity to effectively implement policy plans and programs is crucial to the success of nations. Indeed, having a capable public sector that is able to optimally align resources with actions and actually implement designed policies, is widely considered to be a crucial factor in any state's quality of government (Farazmand 2009;Fukuyama 2013;Rothstein 2015;Wu, Ramesh, and Howlett 2015), and the success of its development efforts (Caiden 1973). In fact, public policies would be inconsequential if governments did not have the capacity to implement them, no matter how representative or democratically formulated such policies would be (Jreisat 2012), or how intricately they would be designed (Howlett 2019). Successful implementation of government initiatives requires capable organizations (Pritchett, Woolcock, and Andrews 2013) that can define problems, set priorities, allocate resources, and carry out requisite implementation of policies to deliver public goods and services efficiently and effectively (Caiden 1973).
None of today's great landmarks could have been achieved without such capability, including the deployment of physical and financial resources, the use of long-term planning, and the monitoring of performance (Caiden 1973). Even more so, today's complex and rapidly changing policy environment necessitates greater efforts at  (Wu et al. 2015

The development challenge
However, for developing countries, operating effectively in today's policy environment is an even bigger challenge. They often suffer from weak implementation capacity (Caiden 1973), as evidenced by the inability of state organizations to undertake even basic administrative functions, such as policing, social service delivery, tax collection, and others. Inadequate capacity has an adverse impact on the outputs and outcomes of government activities in the provision of vital public services such as education, health, and justice (Pritchett, Woolcock, and Andrews 2013). Such poor outcomes can often be attributed to regulatory failures in policy implementation due to lack of capacity (Wu, Ramesh, and Howlett 2015). Enhancing administrative capacity, therefore, is an essential task. Short term efforts costing billions of dollars aimed at building capacity in developing countries, often externally enforced by "global governance players" such as the World Bank, IMF, UNDP or OECD, have so far been unable to demonstrate actual success in achieving their goals (Jones, Guthrie, and Steane 2001;Jones and Kettl 2003;Pritchett, Woolcock, and Andrews 2013). The limitations of a generic approach to capacity development highlights the need for a better understanding of the concept of capacity. In this vein, this paper aims to enhance our knowledge about administrative capacity by reviewing the different terminologies used by scholars, their definitions, frameworks, and indicators used to operationalize and evaluate the extent of administrative capacity achieved.

Understanding and measuring capacity: frameworks and indicators
A multitude of frameworks attempts to conceptualize capacity in relation to other concepts, such as state, policy, and politics. The continued recognition of the importance of capacity for policy success and economic performance (Thomas 2006;World Bank 1997 has also led to notable increase in the use of indices that compare performance across countries (Polidano 2000). This section presents some of the most widely used frameworks and indicators for understanding and measuring administrative capacity.

Capacity frameworks
Complementing the definitional frameworks (Painter and Pierre 2005;Wu, Ramesh, and Howlett 2015) discussed in our introduction, Honadle (1981) identifies organizational functions or activities to enhance understanding of what constitutes capacity at the organizational-operational level (see Figure 1). According to him, capacity involves the ability to anticipate change; develop programs; attract, absorb, and manage resources; evaluate activities; and apply lessons to future activities. Therefore, the ability to perform the administrative practices associated with each organizational function would determine the extent of an organization's capacity. Capacity building, therefore, entails enhancing each function and its associated practices. Standards, benchmarks, and/or indicators, can allow assessment of capacity (or the lack of it). The emphasis on developing mechanisms for anticipating and adapting to change is especially important, given the likelihood of reform efforts failing altogether, because of organizational resistance to change (El-Taliawi 2018a).
Further, according to Painter and Pierre (2005), capacity building requires designing and implementing proper support systems that enhance government's implementation and delivery capabilities. However, normative values underlying the reform efforts should be first identified, followed by support systems and indicators for evaluation (see Table 2 for illustration) (Painter and Pierre 2005;Rotberg 2014). Examples of support systems cited in the literature include external auditing, management by results, strategic planning, e-government, management controls, and others (Ingraham 2007;Farazmand 2009Farazmand , 2004Denhardt 2004;Pollitt and Bouckaert 2017;Jreisat 2004

Capacity indices
One of the most widely used indices by practitioners and scholars are the World Bank's "Worldwide Governance Indicators" (World Bank 2017), that link up various dimensions of capacity to good governance. The index comprises of six sub-indicators: government effectiveness; control of corruption; political stability; regulatory quality; rule of law; and voice and accountability. Using these indicators to track country performance over time, and set standards and benchmarks to measure and subsequently enhance organizational-operational capacity can be helpful. However, a drawback is the aggregated nature of these indices, which does not allow for comparison at subnational levels across sectors and organizations. More so, critics have argued that the indicators are biased as they steer every country "on the road to Denmark" (Grindle 2004) and therefore inherently disadvantage developing countries, as they pay insufficient attention to the fact that countries have very different starting points and histories, and therefore different baseline scores in terms of their capacity.
Another key example is the World Bank's Index of State Credibility (1997). This indicator measures the credibility of a State's institutional framework as perceived by firms, which rank their countries according to criteria, such as predictability of the judiciary and merit-based recruitment, on a scale from "extremely problematic" to "no problem". While the indicator is based on perceptions, the authors of the index find strong correlations between states' credibility rating and their economic performance in terms of growth and investment.
An alternative indicator for measuring administrative capacity is the Pew Center on the States' "Government Performance Project" (Pew Center on the States 2008). The GPP was developed to assess the quality of management in state governments in the United States, according to four criteria: information (e.g. whether the state supports evidence-based policy-making), people (e.g. whether the state retains and develops its workforce), money (e.g. whether budget decisions are based on a long-term perspective), and infrastructure (e.g. whether the state conducts regular analysis of infrastructure needs). The GPP was used in four rounds of surveys across 50 states in the US, in which states were graded on a scale from A to D. These rounds generated subsequent literature in which increased managerial capacity was found to enhance government performance (Zhao and Guo 2011), corroborating the importance of focusing on the organizational-operational dimension of capacity.
More recent indices focus on the resilience and future-preparedness of country's public sectors, with KPMG's change readiness index (KPMG 2018) being a key example. Usual suspects feature in the top rungs of the government capability dimension, although various more recently developed countries such as the United Arab Emirates, Romania, and Bhutan, also increasingly rank highly. Two aspects of this index are noteworthy for the purposes of this paper. First and contrary to many traditional indices, it does not (directly or indirectly) imply that Western-style, competitive democracy is a precondition for countries' ability to display high organizationaladministrative capacity. Such presuppositions have shown to be problematic in assessing China's capacity and performance in recent years (cf. Fukuyama 2013;Rothstein 2015). Second, it compares countries within a bandwidth of peers in terms of their GDP per capita, allowing for a fairer comparison. Moreover, it assesses how countries "punch above their weight" in terms of capacity and resilience vis-a-vis their economic status, with Rwanda being the number one country in that category in the latest edition.

Assessing key components of organizational-operational capacity
Thus, administrative capacity can be conceptualized and measured in different ways depending on the most relevant or feasible components in a particular situation, with some indices and measures focusing more explicitly on the organizational-operational dimension than others. Clearly, almost any country in today's worldboth developed and developinghas been engaging in reform and modernization efforts to enhance the effectiveness, efficiency and capability of its public sector machinery (Pollitt 2011), often driven in part by a need to improve their standing on global rankings that may be directly linked to development aid and investor appetite. This section reviews recent evidence on proposed mechanisms for and components of enhanced organizationaloperational capacity.

More integrated and collaborative structures and arrangements
Key to the organizational-operational dimension of capacity is obviously how the machinery of government is organizedin a top-down, traditional bureaucratic fashion, in a more competitive and arm's length way, or in a networked, collaborative, and participatory modeand how it operatesthe extent to which structure and culture incentivize collaboration between both internal and external actors, adequate performance measurement and management, and a citizen-centric focus in implementation and service delivery.
Undoubtedly, from the 1980s to the early 2000s, the New Public Management or NPM (Hood 1991) was the most widely used, but also the most widely criticized umbrella term for a multitude of attempts to make public sector operations more "businesslike", accountable, competitive, and meritocratic. Analyses have shown how seemingly similar and consistent reform paradigms have played out very differently in countries with different administrative histories and cultures (Cheung 2005;Lynn 2006;Pollitt and Bouckaert 2017).
Two clear trends in many recent reform attempts are "whole-of-government" and "collaborative governance"; both often characterized by different labels such as "new public governance". They seek to make public sector organizations and operations more integrated and coordinated, and more collaborative, citizen-centric and participatory. Co-creation and co-production are buzzwords here (Brandsen and Honingh 2016). Often introduced in response to some of the shortcomings of NPM reforms with its businesslike focus on competition and performance, these reform initiatives seek to enhance collaboration and coordination between different departmental units to address wicked policy problems that cut across traditional boundaries, such as ageing, climate change, and terrorism, and to deal with issues that single organizations, and in fact even governments cannot address alone or in isolation from partners (Agranoff and McGuire 2001;Halligan, Buick, and O'Flynn 2011;Osborne 2010;World Bank 2018;Ballintyne and Mintrom 2018).
Both Parsons (2004) and Wu, Ramesh, and Howlett (2015) argue that weaving together the functions of modern governments by joining up multiple organizations may enhance capacity by presenting a more "coherent policy fabric". Accordingly, better tri-sectoral coordination between government, non-governmental and private organizations can enhance organizational-operational capacity.
Examples of such trends include e-government and "one-stop shops" for citizen service delivery (El-Taliawi 2018b). They were first introduced in frontrunning countries such as Estonia, the UK, and Singapore to remedy the fragmentation in service delivery that accompanied previous reforms (Efficiency Unit 2009). Similar initiatives were also implemented on various scales in countries such as Australia, Finland, and New Zealand (Christensen and Laegreid 2007). According to a UN survey conducted in 2012, the number of developing countries introducing one-stop shops rose from 63 countries in 2004 to 135 countries in 2012, including Egypt, Angola and Costa Rica (UNPAN 2012), with the UNDP award-winning Huduma (Swahili for service) centers in Kenya; being a particularly successful example (Van der Wal 2017a).
The Kenya example shows how benchmarking against, and strategic replication of tested initiatives is a great way for developing countries in particular to enhance organizational-operational capacity. Innovations and reforms need not be earth shattering and radical, as long as they are tailored to the local context in a way that maximizes public value creation.
Evidence about the impact of whole-of-government initiatives on capacity has been mixed (Christensen and Laegreid 2007;Efficiency Unit 2009;Halligan, Buick, and O'Flynn 2011). To illustrate, whole-of-government initiatives in the UK such as the Social Exclusion Unit and the Strategic Communication Unit were found to have improved cost-effectiveness by eliminating duplication and realizing economies of scale (Kavanagh and Richards 2001). Similar initiatives in New Zealand, aimed at tackling domestic violence through the creation of "circuit-breaker teams" for improved coordination and alignment of government funding, were also considered a success (Efficiency Unit 2009).
However, other initiatives in the developed world did not prove to be as successful. Carlin and Guthrie (2003) found the accrual output-based budgeting system implemented in Australia, for instance, to be more "rhetoric than actual action". Furthermore, the re-organization of Homeland Security in the United States (US), post 9/11, considered to be a major coordinative effort in a fragmented system (Kettl 2004), encountered difficulties in getting the participating subordinate administrative units to cooperate (Christensen and Laegreid 2007). Moreover, Ireland's implementation of project "REACH", as a unit within the Department of Social and Family Affairs, to bring more coherence to social services provision, also proved far too complex a system; with the result that only a few services ended up being delivered (Efficiency Unit 2009).
It will be intriguing to see how the horizontal and collaborative reform flow will impact less developed countries over the coming years, and whether local variations will differ widely from one another, just like with NPM implementations (cf. Lynn 2006). Clearly, in more authoritative and traditional settings with largely unquestioned leadership, room for capacity development associated with more citizen and civil society involvement and stakeholder collaboration may be limited (cf. McGuire and Agranoff 2011); in fact, it may be frowned upon or even penalized (Van der Wal 2017a).

Administrative autonomy and impartiality
The political-administrative dichotomy has long been a key tenet of the field of Public Administration (Wilson 1887;Goodnow 1900;Weber 1922). Although scholars debate whether a strict dichotomy between both spaces and roles is realistic and desirable (Svara 2001), sufficient autonomy of public managers and frontline workers and a merit-based appointment system devoid from politicization are widely considered beneficial to capacity in the public administration (Peters 2015;Fukuyama 2013). Indeed, in their study of 52 countries, Rothstein and Teorell (2008) found evidence that impartial government institutions are linked to higher economic growth. Impartiality in the exercise of authority is crucial for good governance and administrative capacity by corollary.
Once more, the sparse empirical evidence about NPM and autonomy comes from developed countries. In her study, Carboni (2010) examined how NPM-reforms implemented in the Italian civil service in the 1990s affirmed the principle of separation of powers between politicians and bureaucrats by assigning them specific roles in the law. While this was supposed to increase the proficiency of the civil service, she argues that increased autonomy of senior bureaucrats paradoxically resulted in a more politicized senior civil service. Conversely, in a study conducted in the Dutch public administration, Hart and Wille (2006) found that following administrative reforms, political-bureaucratic relations in the core executive exhibited growing complementarity and professionalism, and less hostility.
Recent reforms emphasize how bureaucrats redefine some of their roles and capabilities to remain effective, giving rise to intriguing questions about administrative autonomy in a context of more direct engagement of stakeholders and citizens, and contracting out and co-production of services with other players (Van der Wal 2017a).

Management controls
Management controls, or risk management tools and practices, aim to ensure that resources used to achieve organizational and government objectives are not abused or mismanaged (OECD 2017), and the aforementioned autonomy does not morph into a surplus of discretion. Such controls aim to improve administrative performance and operational capacity (Wu, Ramesh, and Howlett 2015;Peters 2015). Various management control practices have been implemented in different countries and in different areas of the public sector, such as social welfare, road construction, and overseas aid delivery. Developing quantitative performance measures, and linking incentive structures and rewards to performance outcomes, in organizations have become a common practice, especially in the context of NPM reforms (Behn 2003;Peters 2015).
The OECD (1996) has compared management control practices implemented in countries, with varying institutional arrangements, and constitutional and cultural backgrounds. The results show considerable diversity between countries. In Germany, for instance, management controls introduced in the road construction sector proved difficult to implement primarily due to sub-optimal design of project contracts. Similarly, controls introduced by the Department of Health in the UK to regulate employment termination payments also proved unsuccessful; resulting in heavy budget losses. Alternatively, the Swedish National Audit Office was able to successfully implement controls involving external auditing, to include all activities in central government agencies and to enable managers to monitor and control activities continuously.
More recently, the OECD (2017) has looked at similar practices in Africa and the Middle East. The findings from developing settings corroborate the European findings: clear legislative oversight of management controls and the design of management control standards are essential ingredients for success. Further, in his study of management controls implemented in Nigeria as part of the country's NPM reforms, Sa'id (2010) argues that such controls failed to deliver on their intended objectives, due to political cooptation. This rather mixed evidence shows how both political will and administrative autonomy are prerequisites for management to be successful in enhancing organizational-operational capacity.

Other supporting factors
While the support systems and reforms reviewed in this section are by no means exhaustive, they are regarded as prime factors that can help build organizational-operational capacity (Wu, Ramesh, and Howlett 2015). In addition, they serve as an illustration of the inconclusive evidence on their success or failure in enhancing such capacity. Other support systems that have been proposed, and which future studies may consider, include decentralization, judicial independence, watchdog bodies, deliberation councils, competitive service delivery, responsive feedback mechanisms, and "hybrid" institutional arrangements (World Bank 1997; UNDP 2011).

Advancing capacity research
This section outlines some of the prevalent issues in the conceptualization and measurement of administrative capacity. While this article has synthesized theoretical and empirical literature on the topic, the review of the literature is not intended to be exhaustive. However, the following observations may pave the way for further advancing the state of the literature: While building administrative capacity is often considered a prime objective of government reform initiatives, there is limited systematic knowledge and reliable information about it (Jreisat 2012). The term is also often used either too narrowly to be understood as the simple training of employees, or too broadly to include the entire functioning of government activities (Schiavo-Campo and McFerson 2008;Jreisat 2012). So far, there seems to be no agreement on the use of terminology, and there are conflicting interpretations of its meaning, as discussed previously (Gargan 1981). More systematic mapping of studies on administrative capacity, with a specific focus on the organizational-operational dimension, are needed to enhance conceptual and empirical clarity (cf. Wu, Ramesh, and Howlett 2015). Evidence about the specific success of components of organizational-operational capacity, such as management controls or collaborative skillsets, and causal links between components and increased capacity are sparse (Painter and Pierre 2005). Most evidence beyond rankings and indices is qualitative, consisting of both comparative and single case studies (Cheung 2005;Brans, Pelgrims, and Hoet 2006;Gualmini 2008;Tsao 2009;Carboni 2010;Pollitt and Bouckaert 2017). The fact that cross-country comparisons lack a common framework or methodology, and most measures are implemented differently, inhibits comparability. Since most evidence comes from developed countries, except for some recent World Bank and United Nations reports, the impact assessment of NPM reforms in developing countries has been ambiguous and non-systematic (Lynn 2006;Manning 2001). How developing countries can specifically enhance administrative capacity remains unclear. An integrated framework is needed to compare and benchmark across contexts. One promising avenue would be to expand Pollitt and Bouckaert's (2017) seminal comparative reform study to developing settings. Also, work similar to that of the Quality of Government Institute (Charron and Lapuente 2018; Rothstein 2011) can increase both internal and external validity of global data sources and facilitate theoretical and empirical development. Finally, while examining administrative capacity, scholars should distinguish between the different tiers of government being examined, such as capital cities and their rural counterparts, which may have different capacities and needs, even within the same country (Polidano 2000). The use of mechanism design theory may help governments to identify their objective and to design the most relevant structures and incentives needed to implement such objectives successfully in their own context (Maskin 2008). Additionally, since there is no established optimum for administrative capacity, the use of benchmarks and indices may be most useful when units of analysis are compared within comparable settings or bandwidths.

Enhancing capacity in practice
Since administrative capacity depends on each country's historical trajectory, path dependency, organizational culture, and other institutional and environmental factors (Andrews 2008;Caiden 1973;Grindle 2004), providing conclusive answers on how to build it is complex. However, in this section we provide six lessons for practice with a particular emphasis on developing settings: 1. Governments aiming to increase administrative capacity should realize that optimal or ideal systems or end states do not exist, but that systems need to be continuously upgraded. Hence, the process of administrative capacity development should be viewed as a dynamic and continuous, rather than a static, process.
2. Assessing whether a certain reform effort has succeeded or failed takes time.
Additionally, reforms can have unintended consequences that are often difficult to undo (World Bank 2018). Therefore, countries should diagnose their systems and develop tailored reforms based on objective baseline evaluation, which includes a historical appreciation of the reform context, including previous efforts. This is especially important since the success or failure of reform efforts and policies is usually path dependent on previous policies (Howlett and Rayner 2006). 3. While there has been a rejection of modernization theory in the development space, development organizations continue to advocate panaceas or universal solutions that may work in some areas, but not in others (Ostrom Janssen, and Anderies 2007). This tendency pushes developing countries to engage in isomorphic mimicry (Grindle 2004). When this process or "game" reaches equilibrium, it captures countries in a capability trap, where no substantial progress is achieved (Andrews 2008;Araral 2009;Pritchett, Woolcock, and Andrews 2013). This may explain the poor results achieved so far in building capacity in developing countries. Therefore, governments of developing countries should contextualize their reforms by diagnosing their problems, learning from international experiences, and designing mechanisms or support systems that fit their unique circumstances (Pahl-Wostl 2015). For example, just venturing into more collaborative modes of service delivery or implementation may not enhance (experienced) capacity in any way, if the foundational bureaucratic infrastructure itself is inadequate. 4. Since administrative capacity intertwines and interplays with other capacity dimensions, including policy and state capacity, components should not be addressed individually or in isolation. Addressing this interdependence is key to designing reform efforts in which enhancement of capacity elements is complementary (Painter and Pierre 2005;Wu, Ramesh, and Howlett 2015). 5. Just as capacity at the individual level presupposes a supportive environment, administrative capacity depends on overarching system-wide institutions. Capable public sectors need to rely on strong regulatory and oversight systems (Grindle and Hilderbrand 1995). Therefore, the political will to engage in sustained and serious reform is key to ensuring strong regulatory systems are not only put in place, but able to function with autonomy and impartiality. 6. Finally, given the differences in timelines, actors, and institutional contextswithin and between countriesgovernments aiming to build their administrative capacity should identify capacity gaps in skills, inter-organizational communication, and overarching monitoring and legislative mechanisms, and develop the human capital and support systems necessary to bridge those gaps by harnessing the administrative and local ownership necessary for success.

Disclosure statement
There is no conflict of interest and no funding information to disclose.