Making Performance Budgeting Reform Work: A Case Study of Korea

Performance budgeting is considered one of the essential elements of public financial management reform and has been adopted in many countries. However, it continues to present a significant challenge within the budget process. This case study draws from change management theories to tell the inside story of performance budgeting reform in the Republic of Korea, including the background of the reform, the formation of the reform team, and how the team overcame resistance among stakeholders to implement and institutionalize the performance budgeting system. Policy implications include the importance of (1) strong support from top decision makers and (2) customization of the performance budgeting system to accommodate a country's cultural and socio-economic characteristics. The Korean case also demonstrates that stable, sustainable performance budgeting reform requires capacity building of relevant stakeholders and, sometimes, significant restructuring of organizations to accommodate the new systems.


Policy Research Working Paper 6353
Performance budgeting is considered one of the essential elements of public financial management reform and has been adopted in many countries. However, it continues to present a significant challenge within the budget process. This case study draws from change management theories to tell the inside story of performance budgeting reform in the Republic of Korea, including the background of the reform, the formation of the reform team, and how the team overcame resistance among stakeholders to implement This paper is a product of the Leadership and Governance Unit, World Bank Institute Vice Presidency. It is part of a larger effort by the World Bank to provide open access to its research and make a contribution to development policy discussions around the world. Policy Research Working Papers are also posted on the Web at http://econ.worldbank.org. The author may be contacted at jchoi6@worldbank.org. and institutionalize the performance budgeting system. Policy implications include the importance of (1) strong support from top decision makers and (2) customization of the performance budgeting system to accommodate a country's cultural and socio-economic characteristics. The Korean case also demonstrates that stable, sustainable performance budgeting reform requires capacity building of relevant stakeholders and, sometimes, significant restructuring of organizations to accommodate the new systems.

I. Introduction
Since the middle of the 1980s, performance budgeting (PB), sometimes called performanceoriented or performance-based budgeting, has been in full swing in many countries-in emerging economies as well as in countries belonging to Organization for Economic Cooperation and Development (OECD). An OECD survey (Curristine 2005) shows that more than 70 percent of OECD countries include non-financial information in their budget documents, and the majority of emerging and developing countries recently introduced various forms of PB and management.
While many countries claim that they have introduced PB, it is not certain how many have successfully integrated performance information into the budget process. Within the World Bank, a major multilateral development bank, there is considerable concern about the success rates of PFM reforms (IEG 2008). One reason for the low rate of success is misguided reform programs that do not fit the specific country's existing conditions. Another part of the problem relates to political economy factors that may constrain or hinder reform efforts (Bunse and Fritz 2011). In light of these concerns, it is all the more important that reformers understand the specific country's institutional conditions, including infrastructure and political economy factors, in order to design a proper reform program.
Recently, World Bank researchers have focused on the process and efficacy of public sector reforms. Among many efforts, Andrews (2011) and Bunse and Fritz (2011) developed conceptual models, and several country teams are now exploring political economy factors that influence those reforms. This paper intends to contribute to this line of study by analyzing PB reform in the Republic of Korea (sometimes called South Korea; hereafter, Korea).
Korea embarked on a comprehensive PB reform program around 2003, after an unsuccessful attempt to introduce PB in the late 1990s. The Korean case is expected to provide useful policy implications for other countries. First, Korea provides a benchmark case of successful PB reform. Korea can be considered a model country with effective public financial management system. The country is known for its sound fiscal condition with records of successful recovery from two recent financial crises. In particular, Korea's reformed PB system is known for its successful integration into the regular budget process. The Korean case has been drawing attention from many emerging and developing economies that share some of Korea's traits: a recent transition to democratic society, experiences with economic development planning, frequent and regular rotation of civil servants, and a lack of capacity among civil servants.
Second, the Korean case may provide useful lessons for developing countries trying to implement a PB reform program within a short timeframe. Relatively quickly, Korea was able to develop a comprehensive PB system with decent success. Thus, the Korean case may offer useful policy lessons for countries adopting a "big bang" approach. This approach carries benefits and risks: it can create irreversible momentum for reform but requires huge resources that many countries lack. This study shows how PB reform was implemented as one component of a PFM reform package.
Third, the Korean case may be useful for countries considering a PB system focused on resource allocation. In some cases, PB is used for program improvement rather than improvement in resource allocation. Depending on the focus of the PB system, a different approach may be required. The Korean PB system can be a benchmark case for countries trying to develop a PB system focused on improving allocative efficiency.
Fourth, the Korean case demonstrates that technical assistance and capacity building are essential elements of PB reform. Few disagree with the goals of PB reform, but the difficulties of PB design and implementation have created many skeptics. The Korean case illustrates how the Korean government handled this skepticism through technical assistance and capacity building for the bureaucracy.
Fifth, the Korean case may offer hope to countries that struggle with implementing effective PB. Korea suffered from a previous failed PB reform program in the late 1990s, extending into 2002. In the early stages of its second attempt, Korea still had not come up with a comprehensive blueprint, and reformers experienced several starts and stops. The Korean case provides useful lessons on developing a customized PB system through trial and error.
This study first describes the current PB system in Korea (Section 2). Section 3 chronicles the rise and fall of first PB attempt (1999)(2000)(2001)(2002). Section 4 tells the change management success story of the second PB attempt, which began in 2003. Because important changes to the reform are still underway, Section 5 summarizes PB developments between 2006 and 2010. Chapter 6 concludes with a summary of the paper, some evidence on the impact of the PB reform, and policy implications.

II. The Current PB System in Korea
The Korean PB system consists of three different layers: monitoring, review, and in-depth evaluation. 1 The monitoring system, called the Performance Goal Management System, gives the first feedback on the performance of budgetary programs, drawing attention to programs that may not be meeting their goals. The strategic review system, or Self-Assessment of Budgetary Programs, collects comprehensive information on the performance of budgetary programs and prioritizes the programs for budgeting based on their ratings. The monitoring system applies to every program, and the strategic review system applies to one-third of the programs, every year. By contrast, the In-depth Evaluation of Budgetary Programs system selects a small number of programs and subjects them to in-depth evaluation. 2 Figure II-1 shows the performance management system framework for budgetary programs in Korea.
1 It is the result of the evolution not by initial design. The evolution of PB system was discussed in the later part, which shows how prioritization process and the in-depth evaluation process developed 2 For the assessment of budgetary programs, the reasons why only 1/3 of programs are evaluated each year is the following: (1) to relieve administrative burden (2)

Monitoring: Performance Goal Management System
The monitoring system requires that line ministries submit annual performance plans and reports, which became mandatory with the enactment of the National Finance Act in 2007. Line ministries submit these reports to the Ministry of Strategy and Finance (formerly the Ministry of Planning and Budget), which, in turn, submits them to the National Audit Office. Finally, they are passed on to the National Assembly. The annual performance plan includes a hierarchical structure of the budgetary programs in each line ministry: their mission, strategic goals, performance goals, and project goals. The mission corresponds to the highest goal of each line ministry; strategic goals match the office-level goals; performance goals link to bureau-level goals; and project goals correspond to team-level goals. Performance indicators and targets are assigned to each performance and project goal. The annual performance plan also contains strategies to achieve the stated goals and an outline of outside factors that are expected to affect the goals.
The monitoring system is designed to be the starting point for performance management of budgetary programs. First, it helps to identify redundancies and deficiencies in existing policy portfolios. By analyzing and developing a hierarchical policy structure, officials can examine whether a particular program is properly designed to serve a policy goal, or conversely, whether a particular policy goal is served by a proper policy portfolio. Second, the monitoring system provides a comprehensive list of performance indicators in relation to policy goals. The system works like a traffic signal, warning when the performance of budgetary programs may be falling short. Although it may not be able to provide sufficient performance information for decision-making purposes, the system serves as a platform to initiate discussion about the effectiveness of budgetary programs.
The engagement of top administrators is the key in utilizing this performance information. The basic overview generated by the monitoring system is made broadly available-to top decision makers and middle managers in the organization, to politicians in the legislature, to outside experts, and to civil service organizations. However, because further communication and information, in addition to the monitoring reports, are needed in order to make effective decisions, the monitoring system is best suited to internal management. Top decision makers within each organization can use the monitoring system to effectively initiate internal communications and to guide their management decisions. So far, the annual performance plan and report from the monitoring system have been developed for informational purposes and have not been used in a systematic way.

Reviewing: Self-Assessment of Budgetary Programs
Of the three PB oversight layers, the review process is the most actively and systematically used for decision-making purposes. As mentioned, information from the monitoring system does not provide enough information to guide the central budget authority in decision making. At the same time, since in-depth evaluation covers only about ten programs each year, its systematic impact on budget allocation is limited.
The Self-Assessment of Budgetary Programs is a review process implemented through a checklist developed by the central budget authority. 3 The checklist contains questions on program planning, management, and results, and each question requires evidence. The review process usually takes about four months to complete, and a third of government programs are evaluated every year. After the central budget authority issues review guidance, line ministries complete the checklist for their programs and submit them to the central budget authority with supporting evidence. The central budget authority then reviews the submissions and assigns each program a rating: effective, moderately effective, adequate, ineffective, or very ineffective. 4 The review results are summarized in Table II

Program Evaluation: In-depth Evaluation of Budgetary Programs
The program evaluation process seeks to measure the relevance, efficiency, and effectiveness of a program. Although ad hoc evaluation activities had previously existed in the central budget office to cope with major policy questions, there was no systematic or rigorous evaluation system prior to PB. Thus, the In-depth Evaluation of Budgetary Programs institutionalized the in-depth evaluation process into the PB system. Every year, an evaluation panel is created for each of the 10 programs selected for evaluation. Panel members, selected for their expertise, mostly come from public research institutes and universities. With the support of the Ministry of Strategy and Finance and line ministries, the evaluation panel collects data, conducts surveys, analyzes the data, and writes a report. Since many stakeholders are involved in any given program, the evaluation panel holds frequent meetings to collect feedback. The panel submits their completed evaluation reports to the central budget office, who decides whether to reflect these results in resource allocation changes or program consolidation.

PB System Features
The PB system in Korea has several context-specific features. First, its focus on linking performance information and budget allocation is mainly driven by the fact that the budget authority is confined to budgeting functions, while the prime minister's office and other ministries perform general policy and personnel management. Given the narrow scope of its mission, the budget authority tries hard to use performance information in actual budget allocation. Otherwise, it cannot find a rationale for running a performance management system. For example, the budget authority issued the guideline that a budget cut of at least 10 percent would be imposed on the programs rated as "ineffective" by the Self-Assessment of Budgetary Programs review.
Second, like other PFM reform measures in Korea, PB is implemented mainly by the budget authority within the executive branch, without much consultation with the legislature. However, the voice from the legislature has become stronger as time has passed, and the nature of PB has evolved accordingly.
Third, Korean PB focuses on producing useful information for decision-making. In practice, that means applying the review process at the sub-program level. Since external factors are more active on the program level, and also because whole programs are too big for prioritization based on performance information, the Korean central budget authority conducts PB on the sub-program level.

III. The First Reform Effort and Its Demise
The first PB reform began as a pilot project in 1999 but never reached a full-scale implementation and ended with the end of the incumbent administration in 2002. This section describes those preliminary reform attempts and their contribution to subsequent reform.
External pressure unquestionably facilitated the introduction of the PB system in Korea. The first attempt in the late 1990s responded to a perceived need for public sector reforms. The Asian financial crisis in 1997-98 raised fundamental questions about Korean socioeconomic systems and resulted in a dramatic restructuring of the private sector to reduce redundancies. This included consolidation of private firms, reforms concerning lifetime employment practices, and massive downsizing. In a parallel attempt, the newly inaugurated administration of Kim Dae-jung adopted government financial reform as a keystone of its political agenda in 1998. The Government Reform Office in the Ministry of Planning and Budget initiated the PB reform, as part of efforts to privatize many public entities and to improve effectiveness and accountability in government spending. With the backing of Korean academics and experts, who had widely discussed the new public management model in 1990s, officials attempted PB as a means of public sector reform.
An "Introduction Plan for PB" was developed by the Government Reform Office at the Ministry of Planning and Budget in January 1999. Because of the prevalence of skepticism within the ministry as well as among line ministries about the effectiveness of PB, a cautious approach was adopted: they would start with a pilot project. In retrospect, this first attempt was doomed from the beginning because no inside stakeholders were confident about the validity of PB. PB reforms will not succeed without the strong support from top decision makers and the effort of a core reform team; PB reforms require the participation of a wide range of stakeholders, such as the central budget office, the budget office in each line ministry, program managers in line ministries, the legislature, civil society organizations, and so on.
The Ministry of Planning and Budget issued guidelines for the pilot project in March 1999. The ministry selected sixteen among forty seven line ministries and agencies, which would submit annual performance plans for fiscal year 2000. The ministry measured those plans against the standard given in the pilot project guidelines, and then promoted the best practices from the submitted performance plans to other ministries and agencies. The selected sixteen ministries and agencies were also required to submit annual performance reports, comparing their performance to their plans. In 2001, the Ministry of Planning and Budget increased the number of participating ministries and agencies to 39.
Despite such efforts, this first pilot project exhibited many limitations. First, it failed to draw attention from high-level decision makers, because the scope of the project was limited to selected offices within a small number of ministries and agencies. Second, there was no specific plan to use the obtained performance information in the budget formulation process. The information in the annual performance plans and reports came in the form of performance indicators and targets for performance goals-not as standards by which organizations could gain or lose funding. By contrast, the extant budget formulation process focused on individual projects. Thus, the PB reformers failed to clarify how to use the gathered performance information in budget formulation. Third, there was an issue of overlap with the already existing policy evaluation system run by the prime minister's office. Both systems looked to the United States' Government Performance and Result Act as a benchmark, and each required similar performance plans and reports from line ministries. This issue of redundancy remained unresolved.
Most important, the Ministry of Planning and Budget failed to put forward government-wide reform agendas. Officials viewed the reforms as pertaining to the "working" rather than the "ministerial" level; that is, the tasks of planning and reporting were left to low-level bureaucrats, with little involvement from top managers. With other major government reforms taking place in Korea simultaneously, PB was unable to gain sufficient attention from high-level decision makers. 5 Lacking momentum to make PB a comprehensive system, these first reform efforts remained a pilot project until the end of the Kim Dae-jung Administration. The project officially ended with the inception of the new administration in 2003.

IV. Making Elusive PB Work: The Second Attempt
Fortunately, the first PB attempts are not the end of the story. This section discusses how the second reform was initiated, managed, and institutionalized with success. Generally, there are not many objections to PB on a normative ground, but making PB really work is a very different issue. In other words, even though stakeholders agree on the rationale of PB reform in principle, the reform cannot succeed without proper change management. Change management includes helping stakeholders accept the need for reform, building capacity to implement reform, and establishing incentives and accountability among stakeholders.
Stakeholder acceptance of PB reform can decline easily if the reform is not backed up by a well-designed PB program and implementation strategy. In the Korean case, the demise of the first reform effort spread skepticism among the central budget authority and line ministries. Not only would reformers need to work around this skepticism, they would need to find better solutions for the technical problems those skeptical officials would encounter in implementing PB. Thus, developing a legitimate technical solution to PB is intimately connected to change management issues. This chapter discusses how technical issues were handled in relation to successful change management.

Beginning the Second PB Reform in the Midst of Skepticism
The second PB reform effort was heavily influenced by strong support from academics, experts, and international organizations. The new administration, inaugurated in 2003, developed a reform agenda focused on public financial management, and in particular on introducing a medium-term expenditure framework (MTEF) with top-down budgeting. MTEF enjoyed priority status because the vice minister of the Ministry of Planning and Budget had a strong interest and expertise in the framework. 6 In contrast, regarding PB, officials in Planning and Budget did not have confidence in its usefulness, and they knew firsthand from previous attempts the difficulty of the technical issues PB might create. Outside supporters of PB offered normative ideas rather than practical know-how, and ministry officials were not sure about practical implementation. Nevertheless, the ministry decided to include PB, out of normative recognition that PB should be an important part of the PFM reform agenda.
Initially, it seemed as if the second PB reform might be doomed like the first attempt: the budget office within the Ministry of Planning and Budget was indifferent towards PB; cynicism was widespread among line ministries; the duplication issue was not entirely cleared up; and a lack of capacity among the Ministry of Planning and Budget and line ministries was obvious from the first reform failure. Consequently, PB came to be viewed as just one part of the reform agenda-and an unpopular part, at that. Considering the negativity surrounding the PB reform, it was a daunting challenge to make PB reform work.

Struggling without New Ideas
In May 2003, 22 lead organizations were selected to develop performance goals and indicator systems for 30 percent of their spending programs. The initial design and plan of this PB was not much different from the pilot project in early 2000s. Once again modeled after the Government Performance and Results Act in the United States, the new PB system was named the "Performance Goal Management System." 7 The only difference from the previous effort was that the new focus did not include programs with smaller budgets consisting mostly of staff salary costs. Instead, the system focused on significant budgetary programs. This differentiation was made in order to avoid redundancy with the performance management system run by the prime minister's office. Still, the fundamental approach to PB was largely the same, with performance indicators and targets and no explicit prioritization process. The Ministry of Planning and Budget expressed no concrete ideas as to how performance information would affect decision-making processes.

Moving from Status Quo toward Real PB Reform
As PFM reform progressed, line ministries' complaints became stronger because they could not see where PB was heading; they had produced an annual performance plan from 2003, but could not see how the information would be used in decision-making purposes. This reinforced existing cynicism among line ministries that had already seen the first PB reform go nowhere.
At the same time, it became increasingly clear to budget officials that PB was an essential element of PFM reform. In particular, the Ministry of Planning and Budget recognized that MTEF with top-down budgeting required PB as a feedback mechanism to ensure the accountability of line ministries in return for their increased autonomy.
This recognition provided additional motivation for the PB reform. The ministry explained to line ministries the need for PB as a way of securing ex post accountability in return for greater autonomy. Faced with this rationale, it was hard for line ministries to object the introduction of PB, even if they were not comfortable with the additional burden of producing performance information.
From the viewpoint of the central budget office, there was another recognition of the PB reform. The introduction of MTEF with top-down budgeting required the central budget to cut the budget request from line ministries if they are beyond given budget ceilings. They felt the PB reform gave them a tool to deal with the situation.
In October 2004, The Ministry of Planning and Budget officials reached agreement that a new approach was necessary in order to make PB really work. Public research institutes undertook further studies to find a viable way of implementing PB. The minister elevated the status of the PB reform team and added to its personnel. Up until then, PB had been handled by a single official within the initial reform team. In 2005, an independent bureau was established to deal with PB reform. This upgrade signaled that PB reform was a serious attempt-not just for show.

Building a Team for Successful Reform
Building a team that will develop and adjust reform programs is the first stepping stone to initiating any kind of reform. Within the Korean PB reform attempt, reform teams were established in two stages. The first reform team reflected the administration's priorities: the team was mostly responsible for MTEF with top-down budgeting, and only one official was intensively engaged in developing the PB reform program. This first reform team followed in the footsteps of the previous PB reform, with a framework based on an annual performance plan and report. As discussed above, the reform team did not have a clear idea as to how the annual performance plan and report would be integrated into budget process. As this limitation became clearer, a reinforced second reform team was established. The second reform team's mission was identifying and developing a new approach to integrate PB into the budget process. The second team, called the "Planning Unit for Performance Management System Improvement" and chaired by the director general of the fiscal planning office, was established in January 2005. The task force included officials responsible for PB, officials experienced in budgeting or evaluation, and researchers from the Korea Institute of Public Finance. With this reinforcement, a dedicated team for PB reform was finally set. Team members from the Korea Institute of Public Finance advised the second reform team to come up with a new review process, which greatly helped to integrate performance information into the budget process.
PB reform requires not only administrative capacity but also analytical capacity within the central budget office and line ministries. Unfortunately, this capacity is often a scarce resource within the administration. In Korea, public research institutes played an important role in supplementing and developing this capacity of the administration. They provided not only advice on policy directions but also practical guidance and training. The Center for Fiscal Analysis was established within the Korea Institute of Public Finance to support the PFM reform; within the center, a team conducted studies on PB. The team was later expanded and established the Center for Performance Evaluation and Management. This center provided advice on developing the review process and developed training programs for budget officials and program managers. It created a manual to help line ministries develop their performance metrics. In addition, the center reviewed the checklists submitted by the line ministries. The center also hosted many workshops and meetings to build consensus among various stakeholders on policy direction. It also helped the Board of Audit and Inspection and the National Assembly to understand the PB system and identify their own roles. Finally, the center engaged in various policy exchange activities organized by international organizations such as the OECD, the World Bank, and the Inter-American Development Bank.
Together, the reinforced reform teams within the Ministry of Planning and Budget and the dedicated public research institute played a key role in making the PB reform work despite initial indifference within the central budget office and line ministries. These teams customized the PB reform system and solved technical issues 8 on a constant basis. Through their efforts, numerous questions and complaints about PB were resolved gradually, if not easily and completely. These efforts overcame many hurdles and brought sustainability to PB reform.
The biggest hurdle was to develop a clear blueprint of reform. This effort drew from the team's expertise in order to understand reform goals and customize the best practices to policy needs. If PB reform had aimed for producing performance information for informational purposes only, PB reform might have remained a shell, without being integrated into the decision-making process. Specifically, during the first PB attempt, performance goals were aimed at a higher level while budget allocation was made at a lower level. Using performance information for resource allocation would have not been easily feasible. A formalized prioritization process is necessary to make performance information work for decision-making purposes. While the production of quality performance information requires a certain type of expertise, designing and implementing prioritization process demands further practical expertise.
Unfortunately, expertise on the aforementioned issues was not readily available. While outside experts and academics had general ideas on PB and supported the PB reform, they did not have expertise on practical issues. Thus, in retrospect, establishing a dedicated reform team consisting of budget officials and policy researchers from the Korea Institute of Public Finance was an important milestone for the success of the PB reform in Korea. The team had expertise on practical as well as theoretical issues and came up with innovative solutions to integrate performance management into the budgeting process.

Finding Solutions and Continued Efforts for PB Reform
Team expertise did not take away all challenges in PB implementation, of course.
Reformers worked together to formulate practical solutions to many difficulties, including the technical and change management task of implementing prioritization and evaluation plans.

Integrating PB into the Budget Process: Introducing the Prioritization Process
As 2004 started, there was concern that the current efforts might not be able to integrate performance information into the budget process. In particular, the current performance information included in the annual performance plan and report was a collection of performance goals, indicators, and targets. It was not clear how to use this information for budget allocation purposes. Since the PB was run by the central budget office and its role was limited to the budgeting issues, it was crucial to find ways of linking performance information and budget allocation. Although program improvement is one of the important purposes of PB, it could not provide enough rationale for the Ministry of Planning and Budget to implement PB.
In October, an internal meeting chaired by the Planning and Budget minister was convened under the title "Future Directions for the Performance Management of the Budgetary Programs." The meeting reached a consensus that more comprehensive performance information on each program was needed to integrate PB into the budget process. To assist in guiding this effort the Korea Development Institute, in cooperation with the World Bank, gathered information on the recent trend of PB. The Korea Institute of Public Finance conducted a case study on the PB system in the U.S., with a focus on the Program Assessment Rating Tool. The experience of the U.S. provided a very useful lesson: performance information without prioritization might not be useful for decision making purposes. In the U.S., the Program Assessment Rating Tool produced more comprehensive information on each program and provided a way of explicitly setting priorities for spending. After reviewing the case study, the Ministry of Planning and Budget bought into this recommendation.
To improve the existing PB system, the ministry established the Planning Unit for Performance Management System Improvement in January 2005. As mentioned in the previous section, this task force included officials responsible for PB, officials experienced in budgeting or evaluation, and researchers from the Korea Institute of Public Finance. The team conducted an intensive study for four months to develop a Korean version of the Program Assessment Rating Tool, named the Self-Assessment of Budgetary Programs. Rather than directly copying the U.S. tool, the team customized the criteria and questions to reflect the characteristics of the Korean system. In addition, the team developed a program checklist, which the ministry viewed not only as an evaluation tool for the central budget office but also as a self-examination tool for line ministries.
To introduce the new self-assessment process, the Planning Unit for Performance Management System Improvement solicited feedback from the budget office within the Ministry of Planning and Budget, and also from the line ministries. The line ministries were not entirely approving of the new review process, because they viewed it as an additional and duplicative performance evaluation process. In response to this lukewarm response, the planning unit explained why the process was not redundant, and why performance information would be useful for decision-making purposes. Eventually, line ministries agreed that the self-assessment would be implemented in 2005. In April 2005, at a cabinet meeting chaired by the president, the Ministry of Planning and Budget reported that the Self-Assessment of Budgetary Programs would be introduced that year. One-third of budgetary programs is evaluated every year; in other words, each program should prepare for evaluation every third year.

Implementing the Prioritization Process
As a first step to implementation, line ministries conducted a self-assessment between April and June 2005. In July and September, the Ministry of Planning and Budget conducted its own review. The ministry held workshops for the line ministries, and the Korea Institute of Public Finance opened training programs to support line ministries' assessment process.
Despite these efforts, the Ministry of Planning and Budget did not find the line ministries' self-assessment satisfactory. Initially, one round of communication with line ministries was planned to reach consensus on the review results, but in the end, three rounds of communication were needed. The Ministry of Planning and Budget's review was a joint effort performed by various groups, including ministry's own Performance Management Bureau and Fiscal Planning Unit, 9 as well as the advisory committee for performance management and the Korea Institute of Public Finance. The ministry hoped that this combination of external and internal information sources would add objectivity to the review process.
Once the ministry set the policy direction, implementation quickly followed. The Self-Assessment of Budgetary Programs was introduced in just four months. In part because of the speedy implementation schedule, line ministries viewed the project as merely another pilot test for PB and assumed that the review results would not be used for decision-making purposes. They were very surprised when the Ministry of Planning and Budget imposed a minimum 10 percent budget cut for programs rated as "ineffective." 10 This was the first time that performance information was explicitly linked to budget allocation. In effect, the cuts shocked line ministries into action-although in other settings it might be controversial to institute such a strong link between evaluation results and budget allocation. The shock dramatically changed the attitudes of the line ministries towards the Self-Assessment of Budgetary Programs. One line ministry developed its own manual for the review process, and some line ministries used the review results for their own internal performance management.
Implementing the review process was not easy. First, line ministries were not ready for the review process: performance information had not been routinely gathered and, as a result, line ministries struggled to come up with acceptable performance information. Not surprisingly, most countries that adopt PB face this lack of reliable performance information. Korea was no exception, and this obstacle created disagreements between the Ministry of Planning and Budget and line ministries. A particular point of tension came from the 9 In May 2005, a drastic organizational restructuring occurred. The Ministry of Planning and Budget was divided into two parts: (1) The Fiscal Planning Unit, similar to the previous budget office, was responsible for the sectoral issues, such as education, defense, local finance, trade and so on; and the other division (2) was comprised of offices responsible for functional issues such as medium-and long-term fiscal strategy, MTEF, annual budgeting, and performance management. The organizations in this second division were called the Fiscal Strategy Office, Fiscal Management Office, and Performance Management Bureau. All three offices were supposed to work with the Fiscal Planning Unit, comprising eleven sectoral teams. 10 Some ministries and agencies use the review results for their personnel management, which is entirely up to line ministries. ministry's requirement that line ministries present outcome information to show program effectiveness.
A second challenge came from within the ministry. It was not easy to convince budget officials in the ministry's Fiscal Planning Unit to participate fully in the review process, as the new reforms created a particularly heavy burden for them. 11 However, cooperation from the Fiscal Planning Unit was very important, because only Planning Unit officials had dealt with the budget issues of individual programs.
Information from the Planning Unit could help to correct asymmetric information problems with line ministries.
To facilitate the cooperation of Fiscal Planning Unit officials, the ministry adjusted the timing of the review process to avoid the budget office's busy season. However, a lack of incentives made it difficult to fully incorporate budget officials into the review process. 12 Thus, the Performance Management Bureau used its own resources to collect as much information as possible, using information from the Board of Audit and Inspection, the National Assembly, and the National Assembly Budget Office; reports from research institutes and civil society organizations; and media sources.
To overcome the performance information problems, training and consulting services were provided to line ministries. The Korea Institute of Public Finance customized those services and also produced a manual to help line ministries develop proper performance indicators. The manual included general principles of developing good indicators and then, for each program type, listed concrete guidelines and customized examples of indicators.
After the first year, a taskforce for improving the Self-Assessment of Budgetary Programs was formed to identify problems and find ways to improve the review process. The team consisted of officials from the administration, including the Ministry of Planning and Budget and line ministries; researchers from the Korean Institute of Public Finance; and academics. The task force suggested a multi-faceted agenda to guide the next year's review process. First, the review process for earmarked funds, formerly a separate process, was merged into the Self-Assessment of Budgetary Programs. This reduced administrative burdens on line ministries and the Ministry of Planning and Budget. Second, a re-evaluation process was established to reflect significant changes made to previously reviewed programs. Since onethird of program was evaluated every year, a program should wait for three years to get next evaluation. To reflect significant improvement in a particular program, line ministries could opt for additional evaluations rather than wait for regular three-year evaluation cycle. Third, more efforts were put into developing recommendations for program improvement. This made clear that the purpose of the Self-Assessment of Budgetary Programs was not only for resource allocation but also for program improvement. Fourth, some clarifications and corrections were made to the checklist, based on user feedback.

Completion of the Korean PB System with Formalized Evaluation
By adding the Self-Assessment of the Budgetary Programs to the existing Performance Goal Management System, the Ministry of Planning and Budget recognized that there could be multiple tools for PB and that each tool might serve different purposes. The Performance Goal Management System consisted of performance goals and indicators that would detect the first signs of program underperformance. The Self-Assessment of Budgetary Programs, with its checklist to collect more comprehensive performance information, would facilitate periodic review, resulting in performance ratings. But when and how would in-depth evaluation be conducted?
Even before PB reform, the central budget office conducted some in-depth program evaluations, mostly in response to problems with specific programs. Although the methodology of this evaluation may have been less than rigorous, the existence of in-depth evaluation activities in the central budget office allowed the Ministry of Planning and Budget to incorporate those activities as a component of PB.
As formalized by outside experts and line ministry officials, the in-depth evaluation process would apply to selected budgetary programs only, as identified, in part, by results of their self-assessments. In 2006, about nine programs were selected for in-depth evaluation. By integrating the in-depth evaluation process into PB, reformers equipped the Korean PB system with three layers of performance information: monitoring, review, and in-depth evaluation.

Strategic Communication with Stakeholders
Communication with stakeholders is a key element of successful change management. In establishing PB, communication within the Ministry of Planning and Budget was the first challenge to overcome. Although top decision makers in the ministry cautiously supported PB reform, the budget office was not cooperative at first. The reform team tried, without much success, to solicit budget officials in the review process. In meetings with the reform team, budget officials complained that they were too busy to be a part of the review process. Thus, the reform team adjusted a time schedule to minimize overlap with the annual budgeting process and conducted the review by themselves, with the help of the Korea Institute of Public Finance. Only at the final stage of the review was the budget officials' opinion solicited.
To facilitate communication with line ministries, the Ministry of Planning and Budget sponsored an annual idea-sharing workshop, then sought to reflect line ministry feedback in the next year's review process. Feedback included suggestions to adjust the review period, clarify the checklist, modify evaluation criteria, and change how review results were used in budget allocation. In addition, the reform team visited line ministries every year and provided orientation sessions. Sometimes the Ministry of Planning and Budget scheduled special sessions, right after the announcement of review results, for line ministries that received poorer ratings than other ministries. These efforts show that constant and intense communication with line ministries is required in order to keep all stakeholders aligned.
Another important group of stakeholders is the legislature, who has final authority over the budget. Communication with the legislature is crucial: If the legislature ignores or misuses performance information, all administration efforts toward PB become futile. In Korea, the legislature raised no serious disagreements over PB, mainly because lawmakers were not aware of the implications of PB, at least initially, and they did not question the objectivity of performance information produced by the administration. However, with the passage of the National Finance Act in 2006, it remains to be seen how the National Assembly will engage in PB.
Communication with a third stakeholder group, civil society, was achieved by inviting representatives of media and civil society organizations to advise on PB, and sometimes to participate in the review process. Review results are made available to the public and the media. So far, the media have reported review results and have sometimes pinpointed line ministries and programs that showed poor performance. Civil society organizations' interest has been more focused on resource-waste incidents than on PB itself. To address this issue, the Ministry of Planning and Budget opened the Center for Anti-Budget Waste, which receives, via Internet and phone, citizens' reports of government waste. Generally, media and civil society organizations have been supporters of PB.

Cooperation with International Organizations
International organizations played an important role in informing policy makers of best practices for PFM reforms and of other countries' experiences. For example, the World Bank provided policy advice in many conferences with public research institutes in Korea and with the Ministry of Planning and Budget. Cooperation with these international organizations also helped to convince stakeholders that PFM reforms were necessary to move towards the best practices in other countries. The secondees from the Ministry of Planning and Budget to the World Bank played an important role in facilitating knowledge sharing between the two organizations. In particular, they helped to clarify for the World Bank the ministry's policy questions, and then conveyed relevant information to the Korean government. In the process, these secondees developed their own views on the reform process and published research output for other countries' reference. 13 As the PB reform progressed, many questions arose among policy makers and researchers in Korea. Other countries' experiences proved useful in addressing these difficult questions and policy issues. For example, the OECD Senior Budget Officials' Network on Performance and Results facilitated valuable discussions with policy makers in various countries. The case of Korean PB reform was received with interest abroad, and Korean leaders learned much about PB implementation from other advanced countries. Researchers from the Korean Institute of Public Finance became regular attendees at international meetings and built extensive networks with policy makers and experts from many countries and international organizations. These networks proved to be an invaluable source of policy exchange about PB.

Legislation and Working with the Legislature
Establishing a legal basis for PB reform is crucial to ensuring the reforms withstand political regime change. Politicians have a tendency to be more interested in starting new reform programs than refining existing management systems. Often an administration's new management system is never really used in the decision-making process. In contrast, a management system that has a legal foundation tends to last longer and enjoys a better chance of continued improvement. Because it takes time for the management system to affect the decision-making process, stability is very important. Thus bureaucrats sometimes respond to reform initiatives by saying, "They will be gone when the current administration goes." Without proof of reform stability, officials tend to just make a show of compliance, rather than change their behavior and decision making.
The PB reform in Korea proceeded at first without legal foundations, but gained them with the enactment of the National Finance Act in October 2006. It was the first overhaul of existing laws related to public finance in 45 years. Before 2006 there had been no major changes to the Budget and Accounting Act of 1961 and the Earmarked Fund Management Act of 1991, although there had been a few revisions.
Other attempts at financial legislation had taken place over the years. In December 2000, the majority and opposition parties agreed to replace the existing Budget and Accounting Act and the Earmarked Fund Management Act with the Restoring Fiscal Soundness Act, which reflected the changed policy environment after the Asian financial crisis. However, the legislature failed to agree on some key provisions, and the legislation efforts ended with only a revision of the Earmarked Fund Management Act. These efforts did not include clauses on PFM reforms, because the PFM reforms started in 2003.
With no ongoing discussion of budgeting and public finance-related issues in the legislature, the Ministry of Planning and Budget in 2003 started to examine the possibility of reorganizing the public finance-related laws. In June the ministry prepared the first draft, which renamed the existing Budget and Accounting Act as the National Finance Act and included clauses not only about budgeting and accounting, but also about PB. After consulting with other stakeholders and making some corrections, the ministry submitted its draft to the National Assembly in October 2004. In response, the opposition party within the National Assembly developed its own version of the act: The party submitted the National Sound Finance Act in December 2004. Neither act passed.
Deliberations in 2006 allowed the legislators to voice their opinions on the new PB reform. Up until then, the reform had been introduced and implemented almost solely by the executive branch, with little communication with the National Assembly. Through the legalization process, however, the legislature joined the dialogue and gave a new momentum to the reformed budget process.
Although legislators voiced no serious disagreements over the contents of the PB system, disagreement on other issues related to public finance delayed and interrupted the enactment process. In particular, opinions about covering government debt brought heated debates to the National Assembly.
Legislators discussed only a few issues related to PB reform. The primary question: Should annual performance plans and reports be mandatory documents for the head of public organizations in the central government? The opposition party's draft specified that such documents be submitted annually to the Planning and Budget minister and to the National Assembly, while the ministry's draft stated that only that the Planning and Budget minister can require such documents. Understanding that ministries were still in the very early stages of developing their performance plans and reports, the legislators agreed that the requirement to submit the documents to the National Assembly would not take effect for one year.
National Assembly involvement added rigor to the PB reforms. Prior to legislation, the Ministry of Planning and Budget viewed the reforms as internal management tools and showed the inclination to use them with flexibility. By contrast, the National Assembly considered the reforms as tools for enhancing administrative accountability and wanted them to be formal and binding. These contrasting perspectives illustrate how different the viewpoints of the administration and the legislature can be. Reformers must find ways to balance conflicting demands without losing reform effectiveness.
After reaching agreements on other issues, the National Finance Act was passed with no objection and one withdrawal in September 2006-one year and eleven months after the administration's draft was submitted to the National Assembly. Thus PB reforms became official law in October 2006.

V. Continued Improvement in the PB System: Recent Developments
The enactment of the National Finance Act in 2006, which brought legal recognition, was the culmination of the PB reform effort. This legislation greatly helped maintain the stability of the program and to ensure its continued development. Although a new administration began in 2008, the framework of the PB system has not changed, and officials have sustained their efforts towards continued PB improvement.
Thanks to the influence of various stakeholders, including the new Ministry of Strategy and Finance, 14 the National Assembly, and the Korea Institute of Public Finance, the PB system now operates more efficiently. Budget and performance goal documents are produced in tandem, resulting in better alignment. The review process has become increasingly strict, with new incentives to encourage performance. Each of these developments is discussed below.
First, 2009 revisions to the National Finance Act addressed misalignment between programs' budget structure and the annual performance plan and report. Previously, budget and performance documents were developed by different units in the Ministry of Strategy and Finance, without much collaboration. Additional inconsistencies arise because performance goal structure tends to be more flexible than program budget structure: While performance goals are often modified internally, changing the budget structure requires agreement between the central budget office and line ministries, and the central budget office prefers to keep budget structures consistent over time.
Cumulatively, the inconsistencies created frustration for legislators, who were unable to match the numbers in the annual performance plan and report with those in budget documents. Thus, in consultation with the Ministry of Strategy and Finance, the National Assembly revised the National Finance Act with an additional specification: the administration should do its best to align the budget structure and the structure in the performance plan and report. Initially, the National Assembly wanted to mandate a one-to-one correspondence. But a Korea Institute of Public Finance expert warned that 100 percent consistency would not be feasible and recommended that the legislators ask only to be able to match the information in the annual performance plan and report to the information in the budget documents. The National Assembly accepted this opinion, and a compromise was included in the 2009 revision.
It is important to note that this PB improvement was achieved due to the involvement of the legislature. In a previous report to the Ministry of Strategy and Finance, experts at the Korea Institute of Public Finance had pointed out the same problem; but because it was not an immediate concern to the administration, no immediate action was taken. With the pressure from the legislature, however, this problem received attention and is now in the process of being corrected: the Ministry of Strategy and Finance and the line ministries are revising their budget structures based on a draft developed by the Korea Institute of Public Finance. This improvement is expected to enhance the usefulness of the annual performance plan and report by establishing close ties between budget structure and performance information.
A second improvement involved changes to the performance review process. Specifically, the Ministry of Strategy and Finance modified the performance rating scale and incentive schemes associated with annual reviews. Ministry officials expressed concern when evaluation results for 2007 showed that only 5.1 percent of programs received an "ineffective" rating-down from 11.3 percent of programs rated "ineffective" for 2006. While some of this change was likely due to actual improvements in program management, the Ministry of Strategy and Finance believed the higher program ratings came mostly from the line ministries' improved ability to convincingly complete their checklists. In order to raise the standard and make the reviews more meaningful, the ministry introduced an additional rating, "very ineffective." Under the new ratings, the proportion of line ministries rated as "ineffective" or "very ineffective" increased to 26.8 percent for 2008. In the Ministry of Strategy and Finance perspective, keeping the proportion of programs subject to budget cuts high maintains pressure on line ministries to continuously improve their performance. Since the rating changes, that proportion has stayed higher than 20 percent.
Threats of budget cuts can be problematic, however. Reviewing entitlement and other politically important programs with objectivity proved to be a difficult task; these programs were especially politically sensitive and it was not easy to impose budget cuts on them. Yet the Ministry of Strategy and Finance needed some way to keep the line ministries in charge of these programs actively engaged in improving performance. In 2010, Strategy and Finance implemented a new idea: since it wasn't feasible to cut the budget of the programs themselves, officials would instead impose punitive budget cuts on the operating costs of organizations responsible for these programs. In other words, line ministries whose programs underperformed would have fewer personnel or fiscal resources to oversee those programs in the future.
Of course, this solution presents problems of its own. Rather than incentivizing future success, could cutting oversight resources actually decrease the line ministry's oversight ability in the future? Line ministries would likely prefer that budget cuts affect their programs rather than their organization, itself-resulting in perceptions of harsh punishment. Should a particular organization be held responsible for politically ill-conceived programs? Perhaps it would be better to kill the program than to penalize the organization assigned to the program's management. In some cases, however, line ministries develop programs on their own initiative to meet popular political demand, and somehow those ministries should be held accountable. Obviously, pros and cons complicate this new incentive scheme.
How the incentives will work in practice and then evolve and remains to be seen.
The overall significance of these attempts is that Korean officials have continued to enhance the PB system even after the legislation was passed and the administration changed. An ideal PB system cannot be developed in one stroke; but it is fair to say that the Korean PB system has benefitted from collaborative innovations and has gained stability with continued improvement.

VI. Summary and Policy Implications
The reform of public finance in Korea was incremental but dramatic, and was achieved in a relatively short timeframe. Important actors from within the administration joined academics, public finance experts, and international organizations to build consensus and resolve stakeholder concerns; eventually the National Assembly played a role in improving and stabilizing PB reform. The story of initial failure, followed by success, offers lessons in both persistence and change management. It is hoped that the implications of Korea's PB reform may guide similar efforts elsewhere.

Trial and Error
The first lesson may be that policy failure can set the stage for eventual success. Initially, the Korean government was not convinced of the effectiveness of incorporating PB in the budgeting process. The Ministry of Planning and Budget started PB as a pilot project for a limited number of ministries and agencies in 1999 but failed to institutionalize it for decisionmaking purposes. Three major factors contributed to the demise of the pilot project in 1999: (1) a lack of attention from high-level officials, (2) the absence of a clear-cut way of incorporating PB into the budget process, and (3) redundancy with a similar performance management system in the prime minister's office.
Despite this failure, the new government started another attempt to introduce PB in 2003. During this second attempt, the redundancy issue was addressed from the start by clarifying that the focus of this new review process was on the management of the budgetary programs rather than overall policy management. However, the administration still did not know how to incorporate the review process into decision making. Thus, the second attempt began with a framework very similar to that of the first attempt: a monitoring system with an annual performance plan, and a report consisting of performance goals and indicators.
Why did the Ministry of Planning and Budget embark on this second effort, notwithstanding its lack of confidence in PB? One answer is political pressure. The widespread support for PB from experts and academics in Korea clearly influenced the policy makers. PB was included in PFM reforms, despite the administration's skepticism, due to outside stakeholders' support for PB. Another answer is the inclusion of PB within broader policy changes. It appears that policy makers worried less about PB, itself, because PB was included as one component of PFM reforms; the focus was on MTEF with top-down budgeting. The rationale that PB was a necessary accountability scheme in return for the increased autonomy of the line ministries under the top-down budgeting seems to have helped line ministries accept the extra work and new constraints the system implied. If the policy makers' interest had been solely on PB, they might have been hesitant in trying it again without clear blueprint. Ironically, as a secondary and comparatively familiar component of the PFM reforms, PB was largely accepted. A third important answer is found in the president's support for the reform package. His broad endorsement of MTEF, and of PB in particular, gave momentum to the reform effort. The president showed strong PB support by requiring ministers to use objective performance information in their annual plans and reports. His insistence convinced government officials that there would be no turning back under the incumbent president.

Budget Ties and Effective Implementation
Important to reform success was finding a way to incorporate PB into the budget decisionmaking process, despite weak interest in PB on the part of politicians and bureaucrats. Although PB began with program performance goals and indicators, policy makers soon realized they needed to move beyond goals and indicators. In order to be an effective decision-making tool, PB needed a prioritization process. Fortuitously, United States officials had developed the Program Assessment Rating Tool, and the Ministry of Planning and Budget benchmarked this tool, with Korea-specific adaptations. By using a standardized checklist, each program was given a rating and budget allocation was guided accordingly.
In developing the prioritization process, called the Self-Assessment of Budgetary Programs, the Ministry of Planning and Budget realized it could incorporate multiple tools that would serve unique purposes in the review process. The ministry found that the annual performance plan and report could be used for the purpose of monitoring and the selfassessment could serve the role of periodic reviews. The ministry also realized that the existing ad hoc program evaluation could be a part of the PB system, providing in-depth evaluation for selected programs. As a result, the Korean government is now equipped with a three-layer PB system.
The introduction of PB brought about the many changes. 15 First, performance information now affects budget allocation: Programs with poor ratings are more likely to receive budget cuts. Park (2011) empirically showed a significant correlation between a program's rating and the probability of a budget cut, although the significance and strength of this correlation varies over time. Second, performance information is reflected at every stage of the budget process. It is used by the line ministries, the central budget authority, and the legislature. The patterns of interaction may be changed in the future as the legislature increases its engagement, but the link between performance information and budget allocation will likely be maintained. Third, there is a tendency to move away from incremental budgeting. Park (2011) also found that once programs are subject to the review process, the sizes of their budget changes are bigger, compared to changes for programs that did not go through the review process. This suggests that budget officials become more aggressive in budget adjustments when they have new performance information at hand.
Notwithstanding their initial skepticism, budget officials within the line ministries are beginning to rely on the new performance information. According to a survey conducted in Korea by the National Assembly Budget Office, 70 percent of budgeting personnel within line ministries responded with neutral or positive answers to the question, "Are assessment results from Self-Assessments of Budgetary Programs useful in the preparation of budget requests?" Considering their tendency for critical evaluations, line ministries show quite a high satisfaction level with the process. Many budgeting personnel in line ministries indicated that the availability of assessment results made their job easier, because performance information facilitates communication between them and program managers.
Not only do line ministries utilize results from self-assessments in their budget formulation, but they also change their program management to enhance the effectiveness of service delivery. For example, since 2000 the Ministry of Health, Welfare, and Family has been running a program designed to provide assistance to unemployed low-income families. However, less than 3 percent of recipients were found to have become "economically independent" through the program. Once such poor performance was highlighted, particularly after the introduction of PB, the ministry began to reform the program. More significant, on their own initiative, ministry officials used reform strategies grounded in PB.
Originally, the assistance program was managed without regard for performance. Funds ranging between US $1,200,000 and $1,900,000 were allocated to 242 regional centers; there was only one center in each locality, and the amount of funds was decided by the number of dependent families. Thus, there was no competition among the centers, and flexibility was very limited. Recently, however, the ministry began a pilot program based on performancebased management. The pilot created a separation between buyers and suppliers: the service provider (or assistance center) is selected based on submitted proposals, and eligibility to become a service provider has been extended to private firms as well as non-profit organizations. In other words, the contract is now performance-based. The ministry makes fixed payments, proportional to the number of families a center serves. In addition, two more incentives have been introduced: (1) If a recipient finds a job meeting a certain standard, a bonus is given to the service provider; and (2) If a recipient moves off the government subsidy, the provider receives an additional bonus. To discourage providers from skimming the best clients, differentiated fixed payments are allocated according to the characteristics of the family. This reform initiative aims to improve the effectiveness of service delivery. The pilot program was introduced in 2008 and will be evaluated in 2012 to see whether nationwide implementation is warranted. Although PB in Korea is still in the early stage, it is gaining momentum and changing program management within line ministries.

Implications for Future Reforms
Korea's success story offers policy implications for other countries. First, strong support from top decision makers is crucial in drawing attention from stakeholders, such as the central budget office and line ministries. These top decision makers should include those who deal with issues across ministries, such as the president, prime minister, and minister of finance, in order for PB to cover the entire government. In the Korean case, the support of top decision makers was crucial in sustaining the second attempt at PFM reform all the way through the institutionalization of PB. In contrast, Korea's first effort towards PFM reform could not last without such endorsement.
Second, cultural and administrative aspects of the government must be considered in any reform effort. In many countries that have adopted PB, officials face little pressure to use performance information explicitly for budget allocation purposes, and it is hard to find evidence that performance information affects budget allocation. In Korea, by contrast, bureaucratic and political culture now creates significant pressure towards the explicit use of performance information. Recognizing that the right political culture was essential to reform, the Ministry of Planning and Budget was very cautious in its implementation of PB until top decision makers supported the reform.
Opportunely, the Ministry of Planning and Budget embraced innovations that were not in their initial blueprint: The ministry developed a prioritization process and found a way to formalize an existing in-depth evaluation process for budgetary use. These actions helped the PB system became well-integrated into the budget process and resulted in greater depth of performance information. In turn, the high stakes of budgeting and the usefulness of performance information have enhanced the perceived value of performance information. If reformers in other countries, like those in Korea, encounter pressure to clearly explain the use of performance information, they need to consider moving beyond performance indicators and targets and find ways to integrate performance information into the budget process. It may not be feasible for those countries to have a PB system only for informational purposes.
A third implication is that organizational restructuring and capacity building may be required to make PB reforms work. In the Korean case, it was not initially clear that capacity among officials was not adequate; clearer was the inadequacy of their willingness to adapt. Thus, organizational restructuring was implemented to overcome resistance from the central budget office and to empower the budget offices in the line ministries. Since organizational restructuring can add confusion in the midst of the PFM reform process, it should be pursued with caution. When adopted with prudence, restructuring may increase momentum for the reform and help the organization fit the reformed system. Before, throughout, and after the reorganization, bureaucrats in Korea received training from many sources, private, public, and administrative. In any scenario, capacity building among budget officials and program managers is an essential element for the success of PB reform. Capacity building can be achieved by upgrading the budget office with additional personnel and more authority, or by providing consulting services and training programs, or both. Since PB reform requires budget officials and program managers to have planning and evaluation capacity-scarce resources in many governments-serious effort should be put into solving the capacity-building issue.
A fourth implication is that a big-bang reform approach may have pros and cons. By adopting a big-bang approach, the Korean government made the reforms irreversible, while it overburdened the government and brought about some inconsistencies among reform programs. This drastic approach signaled strong support from the top decision makers and created an irreversible momentum for the reform. However, there were some technical issues that could have been minimized if a gradual approach with right sequencing had been adopted. In selecting a reform strategy, political and institutional factors should be considered.
A fifth implication is that strong and competent knowledge-providers are essential for the design and implementation of reform programs. Public research institutes in Korea played such role by conducting cross-country studies to identify best practices in other countries and providing many capacity building programs. If such option is not available, international organizations' role looms large. In addition, experts from other advanced countries can assist the reform efforts. However, it would be best to have local expert groups that have knowledge of reform program and local institutional context.

Conclusion: Leadership and Change
Perhaps the most important component of PB success was effective leadership throughout the change. After the demise of the first reform, the Korean government developed a vision and strategy, based on several case studies commissioned to external public research institutes. Top administrators understood that PB reform would require not only administrative capacity but also analytical capacity within leading and line ministries. Thus they assigned these research institutes an important role in supplementing and developing this capacity. Over time, ministries have built up in-house capacity and knowledge regarding PB processes. Because bureaucratic turnover is low, this information is transferable to from administration to administration.
In addition, the Korean government created a sense of urgency among the line ministries and scaled up the reform program as quickly as possible. The Ministry of Planning and Budget stipulated in the budget draft guideline that at least a 10 percent budget cut should be considered for "ineffective" programs. This shocked line ministries into quick action. Another incentive for bureaucrats was public pressure. PB review results were made available to the public, and the media called attention to underperforming organizations. This publicity motivated government officials to honor the new system and, at the same time, helped the reform gain social acceptability among civil society organizations and citizens.
Establishing a dedicated reform team consisting of budget officials and policy researchers from the public institutes allowed reformers to merge expertise on practical and theoretical issues. The team came up with innovative solutions to integrate performance management into the actual budgeting process. The public institutes hosted many workshops and meetings to build consensus on policy direction and coalition among various stakeholders, and at the same time also provided training and consulting services to support line ministries.
Communication with stakeholders is a key element of successful change management. To facilitate additional communication with line ministries, a workshop was held every year to collect their feedback, and their opinion was reflected, as much as possible, in the next year's PB process.
In part because of effective change management, the second reform program did gain the support of middle managers, often the strongest opponents of change, as well as high-level officials. Soon the legislature wanted a say in the reforms, as well. With the National Finance Act of 2006 and revisions to that act in 2009, PB is in Korea to stay.