How to Read This Report
What This Report Is
This report documents Korea’s government performance evaluation system as it existed in 2012–2013 — the product of five decades of evolution from a simple program assessment unit in 1961 to a comprehensive system covering central ministries, local governments, and public institutions. The author, Prof. Mann Hyung Hur of Chung-Ang University, describes the current system in meticulous operational detail: who evaluates whom, using which indicators, on which schedule, with what feedback and incentive mechanisms.
The report’s central argument is that Korea’s system is distinguished by three features: self-evaluation (agencies evaluate their own performance using standardised indicators), integrated evaluation (all public-sector evaluation activities unified under a single legislative framework), and performance management orientation (evaluation results are directly linked to budget, personnel, and incentive systems). These features were built incrementally — Korea’s first evaluation system (1961) was rudimentary, and the current framework was only enacted in 2006.
Read this report as an operational reference, not as a prescription. Korea’s system is complex, institutionally mature, and the product of decades of iteration. The Companion identifies which specific elements are foundational and transferable immediately, which require significant institutional prerequisites, and where the report’s own candid assessment of problems should temper enthusiasm for wholesale adoption.
- Explain Korea’s three-paradigm evolution (program assessment → policy evaluation → performance evaluation) and the institutional and political drivers of each shift.
- Describe the GPEC-centred framework: the roles of the Government Performance Evaluation Committee, ministry self-evaluation committees, and the Policy Analysis and Evaluation Office, and how they interact.
- Distinguish self-evaluation from top-down (teukjung) evaluation — their purposes, targets, procedures, and feedback linkages — and assess which would be feasible as a starting point in your country.
- Identify the report’s own acknowledged problems (fragmented operation, GPEC capacity limits, absence of long-term evaluation) and use them as a basis for critical assessment of the system’s transferability.
How This Companion Is Organised
Practitioner path: 🔍 Diagnose → 📋 Type Guide (your type) → confirm First Action
Course preparation path: 📚 Read the Report → 🔍 Diagnose → 📋 Type Guide (all types) → 🔬 Critical Reading
Critical reading path: 🔬 Critical Reading → Check Understanding → Scenario → Further Reading
Korea's Government Performance Evaluation System: A 6-Chapter Summary
Ch.1 Introduction — The Development of Korea's Policy Evaluation System
Why Evaluation? Korea's Starting Point
Korea’s GNI per capita was $70 in 1960 — lower than the Philippines ($170). By 2011 Korea’s GNI had reached $19,830, fourteen times that of the Philippines ($1,410). The report situates policy evaluation as an underappreciated contributor to this transformation: the program assessment system under the Prime Minister’s Office enabled the government to “readily alter the planned targets with appropriate adjustments in the budgetary priorities and financial facilities so that the private sector could make better use of new market opportunities.” Korea’s achieved economic growth exceeded its planned rate for 15 consecutive years (1962–1977).
Three Paradigm Shifts: Program Assessment → Policy Evaluation → Performance Evaluation
| Paradigm | Period | Key Institution | Focus | Driver |
|---|---|---|---|---|
| Program Assessment (PPAS) | 1961–1994 | PMO (1961–81) then EPB (1981–94) | Economic efficiency of Five-Year Plan projects; 90 university professors reviewed progress annually | Economic growth ideology; Wolf (1961) advisory report recommending independent evaluation |
| Policy Evaluation | 1990–2006 | PMO | Individual ministries’ key public policies; institutional evaluation of ministry performance; Policy Evaluation Committee (30 members) rated and ranked ministries | Korea’s social diversification (1988 Olympics era); abolition of EPB in 1994; Kim Dae-jung’s 2001 Government Performance and Evaluation Act |
| Performance Evaluation | 2006–present | GPEC (Government Performance Evaluation Committee) | Self-evaluation + top-down evaluation; integrated under one legal framework; directly linked to budget, personnel, and incentive systems | Roh Moo-hyun administration’s adoption of New Public Management (NPM) theory; 2006 Government Performance Evaluation Act; influence of US GPRA (1993) |
The 2006 Act: The Foundation of the Current System
The 2006 Government Performance Evaluation Act defines two types of evaluation for central ministries: (1) self-evaluation, conducted by each ministry’s own self-evaluation committee using standardised indicators across three areas (policy processes, financial performance, administrative capacities); and (2) top-down evaluation (teukjung evaluation or specific evaluation), conducted directly by the Government Performance Evaluation Committee on key national policy areas. The Act also extends the evaluation framework to local governments and public institutions, making Korea’s system one of the most comprehensive in scope among OECD countries.
Korea’s evaluation system evolved through three paradigms over 45 years — it was not designed comprehensively from the start. The PPAS (1961) was rudimentary; the 2006 system is the product of iterative reform in response to identified failures. Countries attempting to import the 2006 system without the institutional learning of the preceding 45 years are importing the mature product without the developmental process that produced it.
Ch.2 The Framework of the Performance Evaluation System
Three Core Institutions
Korea’s government performance evaluation operates through three interdependent institutions. The Government Performance Evaluation Committee (GPEC) is the primary institution — co-chaired by the Prime Minister and a civilian representative, with 15 members total (4 government: Prime Minister, Minister of Strategy and Finance, Minister of Public Administration and Security, Minister of Prime Minister’s Office; 11 private sector: 10 college professors from diverse disciplines, 1 accounting professional; 2-year terms). GPEC supports and supervises ministry self-evaluation, directly conducts top-down evaluations, supervises public institution and local government performance evaluation, and may re-evaluate any self-evaluation result it finds unreliable.
The Self-Evaluation Committee is organised by each central ministry individually. It consists of 10–30 members appointed by the Minister (mostly external, some internal; 2-year terms). It meets at least twice a year. Sub-committees handle specific evaluation areas. It is the operational unit of annual self-evaluation — GPEC sets the framework and standards; the ministry committee conducts the actual evaluation.
The Policy Analysis and Evaluation Office (headed by an Assistant Minister under PMO) technically supports GPEC: develops evaluation standards and indicators, prepares basic and execution plans, and handles working-level coordination. Critically, this office has only 35 government officials, approximately one-third dispatched from other ministries — a personnel constraint the report identifies as a significant operational limitation.
Philosophical Background: NPM and Self-Evaluation Theory
The report grounds Korea’s system in New Public Management (NPM) theory and Wildavsky’s (1972) concept of the “self-evaluating organisation.” Under NPM, government employees are held accountable not just for following due process (“doing the right thing”) but for achieving results and citizen satisfaction. Self-evaluation is the mechanism through which agencies internalise this accountability — rather than being evaluated by external inspectors, agencies measure their own performance against pre-established criteria and use the results for internal improvement.
Three Distinctive Features of Korea’s System
Self-evaluation: Agencies evaluate their own performance using standardised indicators developed by GPEC. This is not unmonitored self-reporting — GPEC (under the Roh administration) verified results through meta-evaluation; the Lee administration replaced this with “operational status evaluation” of self-evaluation activities. Integrated evaluation: Before 2006, multiple ministries ran their own evaluation programmes for all other ministries (PMO evaluated policy goals; MOPAS evaluated administrative capacities; MOSF evaluated budget projects; Ministry of Information and Communication evaluated IT capacities). Each ministry prepared for all of these separately — “additional tasks, frequently overburdened.” The 2006 Act unified them. Performance management orientation: Evaluation results are not filed and forgotten — they feed directly into budget appropriations, personnel promotions, and incentive pay.
The GPEC’s civilian majority (11 of 15 members) mirrors the Regulatory Reform Committee design. But like the RRC, these members are government-appointed, serve part-time, and rotate every two years. The report acknowledges that “most members serve only one term, and therefore, they do not have enough time to understand the mechanism of government performance evaluation system.” The civilian majority is structural, but its functional independence and capacity are limited by appointment terms and part-time participation.
Ch.3 Types of Government Performance Evaluation
Self-Evaluation: The Foundation
Self-evaluation covers three target areas. Policy processes: supervised by PMO; evaluates annual goal achievement of individual ministry programmes; results linked to individual employee performance management (incentive pay). Financial performance: supervised by Ministry of Strategy and Finance; evaluates budget project outcomes and R&D programme management; results reflected in next year’s budget appropriations. Administrative capacities: supervised by Ministry of Public Administration and Security; evaluates organisation management, personnel management, and e-government capacity; results linked to work appraisal system for employee promotions.
Each area has its own set of self-evaluation indicators, originally developed by GPEC with input from the supervising ministry, then disseminated to individual agencies for adaptation. Self-evaluation indicators were restructured under the Lee Myung-bak administration: five areas (policy processes, organisation management, personnel management, financial performance, information management) were consolidated into three (policy processes, administrative capacities, financial performance) when the Civil Service Commission and Ministry of Information and Communication were merged into MOPAS.
Top-Down Evaluation (Teukjung Evaluation)
Top-down evaluation is conducted directly by GPEC — not by the evaluated ministry itself. Under the Lee Myung-bak administration, seven target areas were evaluated: key national policies (3 evaluation sections, 9 items, 15 indicators); job creation policies (3 sections, 6 items, 9 indicators); green growth policies (3 sections, 4 items, 9 indicators); policy management capacities (4 sections, 8 indicators); policy PR (3 sections, 10 indicators); regulatory reform; and public satisfaction (customer satisfaction survey conducted by Korea Institute of Public Administration on GPEC’s behalf). Results are announced publicly in December each year.
Public Institution Performance Evaluation
GPEC is the primary institution but MOSF conducts the actual evaluation. Five types of public enterprises and quasi-governmental agencies are evaluated on three dimensions: leadership/responsibility management, management efficiency, and key business programmes. Six grades (S, A, B, C, D, E) are awarded within each institution type. Performance bonuses: public enterprises graded S, A, or B receive maximum 300% performance bonus; quasi-governmental agencies graded S, A, or B receive 100% bonus. Grade C or D triggers at least 10% budget cut.
Local Government Performance Evaluation
GPEC primary institution; MOPAS conducts the evaluation. Target: 6 metropolitan cities and 9 provinces. 38 programmes (2012) categorised into 9 policy areas. Individual area results graded A, B, or C. Grade A receives special revenue sharing; individual achievers receive special incentive bonuses and certificates of merit. A joint evaluation committee conducts the actual evaluation and publishes best practices annually.
The distinction between self-evaluation and top-down evaluation is not just procedural — it reflects a deliberate division of purposes. Self-evaluation is appropriate for annual programme performance (manageable scope, routine feedback). Top-down evaluation is used for key national policies requiring external objectivity and cross-ministry coordination. Korea’s problem, identified in Chapter 6, is that this division has created fragmented governance — neither type covers long-term strategic performance effectively.
Ch.4 Operations of the Government Performance Evaluation System
Annual Self-Evaluation Calendar
| Month | Activity | Responsible Party |
|---|---|---|
| January | GPEC prepares and distributes annual performance plan guidelines | GPEC |
| February | Central ministries prepare annual performance plan; submit to GPEC via Policy Analysis and Evaluation Office | Each ministry |
| May | Ministries prepare annual self-evaluation plans per guidelines | Each ministry |
| July | Self-evaluation committees conduct first-half process evaluation | Ministry self-evaluation committees |
| Aug–Nov | On-the-spot checks for annual self-evaluation (field verification) | Self-evaluation committee members |
| March (following year) | Ministry self-evaluation committees report results to PMO | Each ministry |
| December | Top-down evaluation results announced publicly | GPEC |
Self-Evaluation Indicators — The Operational Core
Chapter 4 is the most detailed section of the report, providing actual indicator sets for each of the three self-evaluation areas. The chapter documents the indicators used across administrations, showing how they were modified. Key structural features of self-evaluation indicators: (1) they are standardised across all ministries, not ministry-specific; (2) they cover both process and outcome measures; (3) they are developed by the supervising ministry in consultation with GPEC; (4) they are revised as national priorities change between administrations.
The chapter also documents financial performance indicators in detail, showing how budget project evaluations (covering research projects, social overhead capital, welfare programmes) are integrated with the self-evaluation framework. The number of target projects per ministry and the evaluation weighting across project types are specified — this level of operational detail is the report’s most practical contribution for countries designing their own indicator systems.
Top-Down Evaluation Operations
Each top-down evaluation type has its own evaluation schedule, indicators, and methods. Key national policy evaluation: 3 sections (planning, implementation, performance outcome), 9 items, 15 indicators; process and outcome measures combined. Green growth evaluation: GPEC assigns evaluation to relevant expert organisations per policy area; results integrated for final GPEC determination. Regulatory reform evaluation: GPEC delegates to Korea Institute of Public Administration; both comprehensive evaluation (all target regulations) and concentrated evaluation (individual ministries’ critical regulations) conducted. Customer satisfaction evaluation: survey of actual service recipients, not ministry self-report.
Ch.5 Feedback Management
The Fragmented Incentive Architecture
One of the report’s most revealing structural findings is that Korea’s government performance evaluation is integrated at the evaluation stage but fragmented at the feedback and incentive stage. Despite GPEC being the primary evaluation institution, it directly administers incentive systems only for: policy process self-evaluation; operational status evaluation of self-evaluation activities; and all seven top-down (specific) evaluation areas. Other incentive systems are administered by other ministries — MOSF administers financial performance and public institution incentives; MOPAS administers administrative capacities and local government incentives. This means the central evaluation institution does not control the rewards flowing from its own evaluations.
Feedback Linkages by Evaluation Type
| Evaluation Type | Feedback Mechanism | Who Controls |
|---|---|---|
| Policy processes (self-eval) | Linked to individual employee performance management — evaluation results automatically reflected in employee performance records | PMO with GPEC |
| Administrative capacities (self-eval) | Linked to work appraisal system — used for promotion decisions | MOPAS |
| Financial performance (self-eval) | Reflected in next year’s budget appropriations | MOSF |
| Top-down evaluation (all 7 types) | GPEC incentive system: recognition, ministerial commendations, performance bonuses | PMO with GPEC |
| Local government performance | Grade A: special revenue sharing; individual achievers: bonus + certificate of merit | MOPAS + joint evaluation committee |
| Public enterprises (grade S, A, B) | Maximum 300% performance evaluation bonus | MOSF |
| Public enterprises (grade C or D) | Minimum 10% budget cut | MOSF |
Individual vs. Organisational Incentives
The report documents that approximately 40 ministries have established their own incentive systems linked to self-evaluation results: 34% of senior executive officers are affected through incentive pay; 31% of director-level officers; 21% of general government employees. This penetration of performance-based pay down to individual employee level is a significant institutional achievement — but it also means evaluation accuracy has direct financial consequences for individual employees, creating incentive to game evaluation results.
Evaluation without feedback is administration for its own sake. Korea’s system explicitly treats feedback as “the first step of performance management,” not the last step of evaluation. But fragmented feedback governance — where GPEC evaluates but MOSF and MOPAS control the incentives — weakens the coherence of the system. Countries designing performance evaluation systems should map the feedback and incentive architecture before the evaluation architecture: if you cannot control what happens to the results, the evaluation process will not change behaviour.
Ch.6 Government Performance Evaluation: Performance and Problems
What Has Worked — Five Achievements
The report claims five genuine achievements after seven years of the 2006 system. First: institutionalisation of self-evaluation culture — ministries have internalised the practice of producing long-term strategic plans, annual performance plans, performance reports, and remedial action plans. Second: full-cycle policy documentation — understanding the evaluation system means understanding the whole mechanism of policy activity in the public sector. Third: monitoring capability — officials have developed the skill to compare progress against pre-set goals continuously, not just at year-end. Fourth: starting point of incentive systems — self-evaluation is not just an evaluation tool but the trigger for performance-linked pay and promotion decisions. Fifth: prospective evaluation capacity — the self-evaluation framework can be used to assess likely policy impacts before implementation.
What Has Not Worked — Five Problems
The report is notably candid about structural problems. First: system is integrated in appearance, fragmented in operation — supervising agencies (MOPAS, MOSF) run their own evaluation activities independently of GPEC; GPEC cannot coordinate what it does not control. Second: GPEC lacks executive power — it is a “deliberative assembly” that can make decisions but cannot spend budget or take follow-up actions; its authority ends at the decision stage. Third: Policy Analysis and Evaluation Office is understaffed — only 35 officials (approximately 11 dispatched from other ministries); not well-trained in social science research methods; a “personnel shortage applies to almost all evaluation activities in the public sector.” Fourth: GPEC members have insufficient time and tenure — part-time members, 2-year terms, most serve only one term; they cannot fully understand the system they oversee. Fifth: no long-term performance evaluation — all 40+ ministries produce long-term strategic plans, but no evaluation tracks whether long-term goals are actually achieved; self-evaluation is annual; top-down evaluation covers key national policies but not individual ministry strategic programmes.
Recommendations for Developing Countries
The report offers three sequenced recommendations: (1) Adopt self-evaluation as the primary framework — it is the most feasible starting point, builds analytical capacity in government employees, and creates direct linkage to incentive systems. (2) Adopt top-down evaluation as a supplementary tool — specifically for long-term strategic programmes and key national policies that self-evaluation cannot adequately cover. (3) Develop a systematic incentive system — evaluation must be the first step of a performance management cycle, not an end in itself; results must feed back into budget, personnel, and incentive mechanisms to change behaviour.
The report’s most important recommendation for developing countries is implicit rather than explicit: start with self-evaluation, not with the whole system. Korea’s integrated system — covering central ministries, local governments, and public institutions under one legal framework — is the product of 45 years of iterative development and requires institutional capacity that most developing countries do not yet have. The self-evaluation mechanism (agency evaluates its own performance against standardised criteria, results linked to incentives) is the foundational element. Everything else can be added incrementally.
Before designing any new evaluation system: audit your current evaluation practices. List every evaluation programme currently operating across central ministries — who conducts it, who is evaluated, what happens to the results, and what it costs. If you find, as Korea did in 2005, that multiple ministries are each running separate evaluations of all other ministries, the immediate priority is integration — consolidating the evaluation burden before adding new evaluation requirements.
What Performance Evaluation Problem Am I Trying to Solve?
Diagnostic Tool — Find Your Type in Three Questions
Korea's Performance Evaluation Experience Mapped to Your Problem Type
Type A No System — Where to Start
‘We have no formal evaluation. Ministries produce annual reports but nobody evaluates whether the goals in those reports were actually achieved. The finance ministry looks at budgets, not outcomes. Nobody is measuring policy performance.’
Korea’s Experience
Korea’s starting point in 1961 was similar: a simple advisory report by economist Wolf recommended a separate evaluation unit independent of the planning office. The Korean government organised the Government Policy Evaluation Board — 90 university professors who reviewed economic development plan progress annually. This was not a sophisticated system: it focused on economic programme efficiency only, operated semi-officially, and functioned as monitoring during implementation rather than evaluation after it. But it worked — it enabled the government to alter plans when market forces changed, and Korea’s economic growth exceeded its planned targets for 15 consecutive years.
The lesson is not to build the 2006 system from scratch. The lesson is to start with a single, focused, legally-grounded evaluation activity that produces results officials actually use. Korea’s PPAS worked because it was: (1) independent of the planning office; (2) staffed by credible external experts; (3) directly linked to adjustment of plan targets; and (4) focused on a limited scope (economic development plans only, not all government activity). The 2006 integrated system came 45 years later.
The first evaluation system need not be comprehensive. Korea’s 1961 PPAS evaluated only economic development plan programmes. The most important features of a starting system are: independence from the evaluated entity; linkage to decisions that officials care about (budget, plan targets); and a manageable scope that enables credible evaluation with limited resources.
| Priority | Section | Why read it |
|---|---|---|
| 🔴 Essential | Ch.1 §2 (PPAS to Policy Evaluation evolution) | Korea’s starting point — a simple program assessment that preceded the current system by 45 years |
| 🔴 Essential | Ch.6 §3 (Suggestions for developing countries) | The report’s direct recommendations: self-evaluation first, top-down second, incentive systems third |
| 🟡 Recommended | Ch.2 §1 (Significance of performance evaluation) | The NPM theoretical background and the case for evaluation as management tool, not compliance exercise |
| ⚪ Optional | Ch.1 Table 1-2 (Transition of Policy Evaluation System) | Visual summary of all paradigm shifts — useful for explaining the evolution to senior officials |
The PPAS model — external experts reviewing progress against a national plan — is directly transferable to any country with a national development plan. It requires: (1) a designated evaluation secretariat (even small); (2) a credible pool of external reviewers; (3) a legal or administrative mandate connecting evaluation to plan adjustment. These are achievable before the institutional infrastructure for a full self-evaluation system exists.
Select two or three highest-priority government programmes (ideally from your national development plan) and commission a one-year pilot evaluation by an independent academic or research institution. Require the evaluation report to specify: (a) what the programme intended to achieve; (b) what it actually achieved; (c) what should change in next year’s plan targets or budget. Use the pilot to build evaluation literacy among ministry officials before mandating self-evaluation across all agencies.
Type B Fragmentation — Too Many Separate Evaluations
‘The finance ministry evaluates our budget performance. The home affairs ministry evaluates our administrative capacity. The sector ministry evaluates our programme targets. The anti-corruption commission evaluates our ethics. We spend more time preparing for evaluations than doing actual work.’
Korea’s Experience
This was exactly the problem the 2006 reform solved. Before 2006: PMO evaluated policy goal achievement; Ministry of Information and Communication evaluated all ministries’ IT capacity; Ministry of Government Administration evaluated administrative capacities; Ministry of Planning and Budget evaluated budget project outcomes. Each ministry prepared separately for all of these with different indicators, schedules, and formats. The 2005 advisory committee found individual ministries “overburdened with these preparations” and decided to “integrate these individual evaluation measures under the government performance evaluation system.”
The 2006 Act unified all central ministry evaluations under a single legislative framework, with three self-evaluation areas (policy processes, financial performance, administrative capacities) replacing the multiple separate evaluations. Each area was assigned to a coordinating ministry (PMO for policy processes; MOSF for financial performance; MOPAS for administrative capacities) but all operated under GPEC oversight and a common annual schedule. The report acknowledges that this integration worked at the evaluation stage but not at the feedback stage — incentive systems remain fragmented across the three coordinating ministries.
Integration of evaluation burden requires a legal anchor — the 2006 Act — that formally prohibits duplicate evaluation requirements and requires all central ministry evaluations to operate within the unified framework. Without a legal basis, each ministry’s evaluation programme survives through bureaucratic inertia. Korea’s integration was enabled by the legal mandate, not by voluntary coordination.
| Priority | Section | Why read it |
|---|---|---|
| 🔴 Essential | Ch.1 §3 (Birth of Government Performance Evaluation System) | The 2005 advisory committee process that diagnosed the fragmentation problem and designed the integrated solution |
| 🔴 Essential | Ch.2 §1 (Integrated evaluation feature) | How the three-area structure (policy processes / financial performance / administrative capacities) replaced multiple separate evaluations |
| 🟡 Recommended | Ch.6 §1.2, Problem 1 | Why integration in form does not guarantee integration in operation — the residual fragmentation problem |
| ⚪ Optional | Ch.6 Table 6-1 (Evaluation types and agencies) | Visual map of which agencies administer which evaluation types — shows the structural fragmentation that persists |
The integration concept is directly transferable. The legal mechanism (a Government Performance Evaluation Act that explicitly supersedes separate evaluation programmes) is straightforward to design. The challenge is political: the ministries that lose their separate evaluation powers (and the compliance authority that comes with requiring all others to submit to their evaluation) will resist. Korea’s integration required a presidential advisory process and a new law to overcome this resistance.
Conduct an evaluation audit: list every evaluation programme imposed on central ministries by any central agency. For each, record: who conducts it, who is evaluated, what indicators are used, what the results feed into, and what it costs in ministry staff-days per year. Present the total compliance burden to the Prime Minister’s office as the case for integration. The audit itself, without any new policy, typically generates significant political momentum for consolidation.
Type C Evaluation Exists But Results Are Not Used
‘We produce 200-page annual performance reports. They are submitted to the planning ministry in March. Nobody reads them. The same budget is allocated next year regardless of results. Ministers know the evaluation has no consequences, so they treat it as a paperwork exercise.’
Korea’s Experience
Korea’s feedback architecture is the most operationally distinctive feature of its system. The 2006 Act explicitly linked each evaluation area to a specific consequence: policy process evaluation results are automatically reflected in individual employee performance records and incentive pay; administrative capacities results feed into work appraisal systems used for promotion decisions; financial performance results are reflected in next year’s budget appropriations. The report notes that approximately 34% of senior executive officers, 31% of directors, and 21% of general employees are affected by performance-based incentive pay linked to self-evaluation results.
This feedback design is intentional and structural, not incidental. The Korean system was explicitly modelled on the US GPRA (1993), which requires agencies to engage in goal-setting, measuring results, and reporting progress — and which links these reports to the congressional budget process. Korea’s version goes further by linking results not just to budget but to individual personnel decisions, creating personal financial stakes in evaluation accuracy.
Feedback must be designed before evaluation, not after. The question “what will we do with the results?” must be answered before the evaluation system is designed, because the answer determines what indicators to use, what evaluation frequency is appropriate, and what institutional authority the evaluating body needs. Korea’s three feedback linkages (performance pay → employee records; promotions → work appraisal; budget → appropriations) each required different institutional connections designed into the system from the start.
| Priority | Section | Why read it |
|---|---|---|
| 🔴 Essential | Ch.5 §1.2 (Framework of feedback management) | The complete architecture: which evaluation type feeds into which consequence, administered by which agency |
| 🔴 Essential | Ch.5 Table 5-1 (Evaluation types and incentive agencies) | Visual map of all evaluation-to-incentive linkages — the most transferable design reference in the report |
| 🟡 Recommended | Ch.6 §1.1, Achievement 4 | “Self-evaluation in Korea is a starting point of incentive systems” — the report’s clearest statement of the feedback-first philosophy |
| ⚪ Optional | Ch.5 §2.1.1 (Self-evaluation feedback system detail) | Percentage of officers affected by incentive pay at each grade level — useful for designing penetration depth of feedback linkages |
The concept of linking evaluation results to consequences is directly transferable and highly important. The specific linkages (budget appropriations; civil service performance management; promotion appraisal) require different institutional connections in different country contexts. Budget linkage is usually the most achievable first step — it requires coordination only with the finance ministry and does not require changing civil service rules. Personnel linkages are typically more politically sensitive and come later.
For the next budget cycle: require that each ministry’s budget submission for any programme that was evaluated in the previous year include a one-page summary of: (a) the evaluation finding; (b) the budget implication of that finding (increase, decrease, or redirect). This creates the minimum budget-evaluation linkage without requiring changes to civil service rules. If the finance ministry accepts evaluations that recommend budget cuts as well as increases, the feedback loop begins to function.
Type D Self-Evaluation Not Credible — Agencies Overrate Themselves
‘Every ministry gives itself a high rating. There is no external check. We know the self-evaluation scores are inflated but the system has no mechanism to verify them. The scores are meaningless for management decisions because no one trusts them.’
Korea’s Experience
Korea confronted this exact problem and tried two different solutions. Under the Roh Moo-hyun administration, GPEC conducted a meta-evaluation — evaluating each ministry’s self-evaluation results for reliability and objectivity. If GPEC found a self-evaluation result questionable, it re-evaluated the ministry independently. This was labour-intensive and was criticised as creating a new layer of evaluation on top of self-evaluation. Under the Lee Myung-bak administration, meta-evaluation was replaced by operational status evaluation — evaluating the quality of each ministry’s self-evaluation process rather than its results. This shifted the accountability from “were your results accurate?” to “did you conduct evaluation properly?”
The top-down (teukjung) evaluation system serves as an independent external check on the most important national policy areas: GPEC directly evaluates whether ministries are effectively executing key policies, providing a benchmark against which self-evaluation scores can be compared. For public institutions, the evaluation board (150 professionals) conducts external evaluation, and results directly determine performance bonuses — creating strong incentives for accuracy.
Self-evaluation credibility requires external verification, but the form of verification matters. Korea tried meta-evaluation (check the results) and moved to operational status evaluation (check the process). Neither fully solves the overrating problem, but the combination of: (1) standardised indicators that are difficult to manipulate; (2) external process verification; (3) top-down evaluation of key policies as a credibility benchmark; and (4) financial consequences for major overrating creates a credible system without requiring full external evaluation of every programme.
| Priority | Section | Why read it |
|---|---|---|
| 🔴 Essential | Ch.6 §1 (Performance — first achievement) | How self-evaluation settled in — the transition from meta-evaluation to operational status evaluation and why |
| 🔴 Essential | Ch.4 §1.3 (Evaluation on operational status of self-evaluation activities) | The operational status evaluation mechanism: what it checks and how it works as a credibility tool |
| 🟡 Recommended | Ch.3 §1 (Overview of self-evaluation) and Ch.3 §3 (Top-down evaluation) | How the two evaluation types serve complementary credibility functions |
| ⚪ Optional | Ch.2 §2 (GPEC functions) | GPEC’s authority to re-evaluate questionable self-evaluation results — the legal basis for external verification |
The operational status evaluation approach (verify the process, not just the results) is more transferable than full meta-evaluation because it requires less evaluation capacity. The key transferable design principle: make self-evaluation indicators specific and outcome-focused enough that inflating scores requires falsifying data, not just generous interpretation. Indicators that ask “did you conduct a stakeholder consultation?” (easily inflated) are less credible than indicators that ask “what was the citizen satisfaction score for service X?” (harder to fabricate without external data).
Select five indicators from your current self-evaluation system (if one exists) or from the system you are designing. For each indicator, ask: can this indicator be self-reported without any external data? If yes, how easily can it be inflated without being detected? Replace the two most easily inflated indicators with outcome-based indicators that require either external data (citizen surveys, inspection results, court records) or third-party verification. This is the minimum credibility improvement that requires no new institutional infrastructure.
Type E Long-term Strategic Performance Not Evaluated
‘We have five-year strategic plans for every ministry. But we only evaluate annual performance. Nobody knows whether the five-year goals are being achieved. In year three of a five-year plan, there is no way to know if we are on track to achieve the year-five targets.’
Korea’s Experience
This is Korea’s most candidly acknowledged problem — and it persists despite seven years of the 2006 system. All 40+ Korean ministries are legally required to produce long-term strategic plans. But the report states plainly: “No self-evaluation indicator directly indicated toward the achievement of long-term performance goals. Moreover, long-term goals have not been assessed even as a form of top-down evaluation in the last seven years.” Korea’s self-evaluation is entirely annual; top-down evaluation covers key national policies but not individual ministry’s strategic programmes.
The report recommends that GPEC and PMO’s Office of Policy Analysis and Evaluation conduct evaluation on individual ministries’ long-term strategic programmes, conducted every other year. This would add a third evaluation mode to the existing system: annual self-evaluation for programme performance + periodic top-down evaluation for key national policies + biennial strategic programme evaluation for long-term goals. At the time of writing, this remained a recommendation, not an operational component.
Annual evaluation systems are structurally mismatched to long-term strategic goals. If a ministry sets a 10-year goal and is only evaluated annually on short-term outputs, annual evaluation creates perverse incentives: optimise for this year’s metrics at the expense of long-term trajectory. Long-term evaluation requires different institutional design: longer evaluation cycles (every 2–3 years), different indicators (progress toward 5-year milestones rather than annual outputs), and evaluators with sufficient tenure to understand the programme’s history.
| Priority | Section | Why read it |
|---|---|---|
| 🔴 Essential | Ch.6 §1.2, Problem 5 | Korea’s candid admission: seven years of the system, no long-term evaluation conducted — the most honest diagnosis in the report |
| 🔴 Essential | Ch.6 §2 (Model for better evaluation practices) | Korea’s proposed solution: biennial top-down evaluation of individual ministry strategic programmes |
| 🟡 Recommended | Ch.6 §3 (Suggestions for developing countries) | Top-down evaluation as a supplementary tool specifically for long-term strategic programmes |
| ⚪ Optional | Ch.6 §1.1, Achievement 5 (Prospective evaluation) | How self-evaluation’s goal-setting process can be used as a prospective evaluation tool — a partial substitute for formal long-term evaluation |
This type is the most transferable as a warning, not as a model. The lesson from Korea’s experience is: when you design an annual self-evaluation system, explicitly include a biennial or triennial long-term strategic programme evaluation component from the start. Korea did not — and seven years later, strategic plans are produced but not evaluated. Countries designing performance evaluation systems now can avoid this gap by design rather than discovering it after years of operation.
For each of your government’s top five strategic programmes (from the national development plan or equivalent): write down the five-year goal, the current year’s expected milestone, and the actual current achievement. If you cannot write down the “actual current achievement” column without investigation — if this information is not available and monitored routinely — that absence is your diagnosis. The first action is to establish a routine (quarterly or annual) tracking mechanism for five-year milestone achievement before designing a formal evaluation system for long-term performance.
Reading Against the Grain
5.1 🔬 Success Bias
This report is unusually honest — it dedicates an entire chapter to acknowledged problems and explicitly states that GPEC lacks executive power, the secretariat is understaffed, and long-term evaluation has never been conducted. But even an honest report has selection biases. Reading what is omitted alongside what is said is essential for critical assessment.
| What the Report Says | What the Report Does Not Say | Critical Question |
|---|---|---|
| Self-evaluation has “settled in” and government employees have changed their attitudes from rule-of-thumb to rational analysis | The report provides no systematic evidence of this attitude change. It cites the author’s own 2009 paper but does not present before/after comparison data or independent verification. The claim that self-evaluation changed evaluation culture is asserted, not demonstrated | What evidence would you need to verify that self-evaluation has genuinely changed decision-making behaviour, rather than simply producing more sophisticated compliance theatre? |
| Korea’s 1961 PPAS contributed to the government’s ability to alter plans when market conditions changed, enabling 15 years of economic growth exceeding targets | Causality between evaluation and economic growth is not established. Korea’s growth during 1962–1977 also reflected: US military assistance, cheap labour, export market access, Japanese reparations investment, and Park Chung-Hee’s authoritarian ability to override vested interests. The PPAS operated in a political system where the President could simply direct changes regardless of any evaluation result | In a democratic system without authoritarian enforcement mechanisms, would the same evaluation system produce the same adjustment capacity? What would be different? |
| The integrated self-evaluation system reduced evaluation burden on ministries compared to the pre-2006 multiple separate evaluations | The report acknowledges in Chapter 6 that the system is “seriously fragmented in its operation” and that GPEC cannot coordinate supervising agencies that run evaluation activities independently. The burden reduction at the ministry level may have been offset by the added complexity of managing results across three different supervising ministries with different incentive systems | Has the integration of evaluation inputs (what ministries prepare) been matched by integration of evaluation outputs (what decision-makers use)? How would you measure this? |
| GPEC is the primary institution of government performance evaluation, co-chaired by the Prime Minister | GPEC is described as a “deliberative assembly” without executive power or budget. Its 11 civilian members serve part-time, rotate every two years, and most serve only one term. The Prime Minister’s co-chairmanship signals priority but does not guarantee the PM attends or that decisions are enforced | In your country, would a co-chairmanship by the Prime Minister guarantee active engagement, or would it become a ceremonial title delegated to a deputy? |
| The financial performance self-evaluation is reflected in next year’s budget appropriations | The report does not provide data on how often evaluation results have actually resulted in budget cuts. The claim that poor evaluation leads to budget reduction is structurally plausible but the report does not verify it with budget outcome data | If budget cuts based on evaluation results are politically difficult to implement (as they are in most countries), how does the budget feedback link actually function in practice versus in design? |
5.2 ✅ Check Your Understanding
5.3 ✏️ Scenario Writing
Read the scenario, examine the tensions, then write your own response in your notes. There is no single correct answer.
The Scenario: The Minister Wants Evaluation Results to Look Good
Director Park Eunhee is the head of the Performance Evaluation Division of the Ministry of Planning in a Southeast Asian country that introduced a self-evaluation system three years ago. The system was modelled on Korea’s framework: each ministry evaluates its own programmes using standardised indicators, submits results to the central evaluation office, and receives a score that feeds into its budget allocation for the following year.
The Minister of Health has called Director Park to ask for clarification on how scores are calculated. Two of the Ministry of Health’s programmes received low scores on outcome indicators (vaccination coverage rate: 61% against a 75% target; maternal mortality reduction: 3% against a 10% target). The budget implications of these low scores would be significant cuts to both programmes.
The Minister of Health argues: the targets were set when the government’s health sector budget was expected to grow by 15% per year. The budget actually grew by only 3%. Achieving 61% vaccination coverage with only 3% budget growth is, the Minister argues, better performance than achieving 75% with 15% budget growth would have been. He is asking Director Park to allow the Ministry of Health to resubmit its evaluation with adjusted targets that reflect the actual budget constraint, which would change both scores from low to high.
Director Park knows this request is happening simultaneously in four other ministries facing similar budget-constrained situations. If she approves the Health Ministry’s request, she will be flooded with similar requests. If she denies it, she may be right procedurally but the Health Minister is correct that the targets were set under assumptions that did not hold.
Core Tensions
| Tension | Value 1 | Value 2 |
|---|---|---|
| Procedural integrity vs. substantive fairness | Allowing retroactive target adjustment destroys evaluation credibility — if targets can be changed after results are known, evaluation is meaningless | Holding ministries to targets set under assumptions that proved false punishes ministries for factors outside their control, creating perverse incentives for target-setting |
| Budget consequences vs. evaluation learning | Budget cuts based on evaluation results create strong incentives to take evaluation seriously — the feedback mechanism only works if it has real consequences | Budget cuts imposed on health programmes with genuinely constrained budgets may worsen the very outcomes being evaluated, without improving anything except score consistency |
| System consistency vs. contextual adjustment | A standardised system that applies equally to all ministries enables comparison and prevents special pleading | Targets that do not account for differential budget constraints produce unfair comparisons between ministries with different resource levels |
- The target-setting design question: What does this scenario reveal about how performance targets should be set? Should targets be fixed at the start of the year regardless of budget changes, or should they include an automatic adjustment formula when budget assumptions change by more than a threshold amount?
- The Korea parallel: Korea’s GPEC has authority to re-evaluate ministry self-evaluation results it finds questionable. Is this scenario an argument for GPEC-style meta-evaluation? Or for better target-setting design? Or both?
- The feedback dilemma: If Director Park refuses the adjustment and budget cuts follow, and vaccination coverage falls further next year as a result, who bears responsibility? How does this affect your answer about whether to allow the adjustment?
- Your own context: Has your country’s evaluation system faced similar pressure to adjust results or targets retroactively? What happened, and what were the long-term consequences for evaluation credibility?
5.4 📝 Assignments
- a.For each feature: is it present (fully, partially, or not at all) in your country? What specific evidence supports your assessment?
- b.Which feature is most important to develop first given your country’s current institutional capacity? Why not the other two?
- c.Korea developed these three features sequentially over decades, not simultaneously. What sequence would you recommend for your country, and why does sequence matter?
- a.Identify the specific structural feature that produces fragmentation (Ch.6 problems 1 and 4)
- b.Explain why integrated evaluation at the input stage (what ministries prepare) did not produce integrated evaluation at the output stage (what decision-makers use)
- c.Propose one structural change that would reduce this fragmentation without requiring a new law or new institution
- ①Situation diagnosis — which of the five problem types (A–E) applies to your country, with evidence
- ②One element to adopt directly — with specific reference to the Korean mechanism and evidence of its effectiveness
- ③One element to adapt — with explicit statement of what must change to fit your country’s context
- ④One element not to adopt — with honest reasoning about why it does not fit your situation (institutional prerequisites not yet met, or Korea’s own problems are reason enough to avoid it)
- ⑤First action in 30 days — a concrete action that does not require new legislation or budget
5.5 📚 Further Reading
- Wildavsky, A. (1972), "The self-evaluating organization," Public Administration Review 32(5): The foundational theoretical paper behind Korea’s self-evaluation philosophy. The report cites it explicitly and repeatedly. Reading the original reveals how much of Korea’s system is a direct application of Wildavsky’s concept — and what Wildavsky actually meant by an “evaluative man” versus what Korea’s system has achieved.
- US Office of Management and Budget, Annual GPRA Performance Reports: The American model Korea explicitly borrowed from. Comparing US GPRA implementation experience (its achievements and criticisms) with Korea’s implementation provides a controlled comparison of how the same theoretical model plays out in different institutional contexts. The US Government Accountability Office’s assessments of GPRA effectiveness are particularly useful.
- OECD, "Government at a Glance" (annual): Comparative data on government performance management systems across OECD and partner countries, including Korea. The chapters on performance budgeting and performance information use allow you to benchmark Korea’s system against international peers and identify where Korea’s approach is distinctive versus standard practice.
- Boyne, G.A. et al. (2004), "Toward the self-evaluating organization? An empirical test of the Wildavsky model," Public Administration Review 64(4): The paper the report cites as its theoretical foundation for self-evaluation in public organisations. Provides the empirical evidence base for when and why self-evaluation works — and when it does not. Reading this gives the conditions under which the Korean model is most likely to be transferable.
- Kim, S. (2010), "The Performance Evaluation System in the Korean Public Sector," OECD Working Paper on Public Governance: An external OECD assessment of Korea’s performance evaluation system, independent of the KSP report. Provides a comparative perspective that the self-authored KSP report cannot provide — particularly on where Korea’s system compares favourably and unfavourably to OECD norms.
- Pollitt, C. and Bouckaert, G. (2011), "Public Management Reform: A Comparative Analysis – New Public Management, Governance, and the Neo-Weberian State" (3rd ed., Oxford University Press): The standard comparative public management textbook that contextualises NPM-based performance evaluation within the broader history of public management reform. Provides the analytical framework for understanding why Korea’s system took the form it did and how NPM-inspired reforms have fared differently in different national contexts.
This Companion is a learning aid for Prime Minister’s Office · KDI School, KSP Knowledge Sharing Program — Korea’s Government Performance Evaluation System and Operating Experience (2013). Use alongside the original report.