Do Evaluations Follow Policy? Mission-Oriented R&D and the Limits of Output Accountability in Korea

• In Event: Governance Ecosystems: Innovation, Institutional Design, and Policy Outcomes

Abstract

Governments increasingly claim to embrace mission-oriented innovation — actively shaping technological trajectories rather than correcting market failures. Whether public performance evaluation systems have been redesigned to match that ambition remains largely untested. When a government adopts Entrepreneurial State logic, do the evaluation frameworks governing its public research institutions follow?
Korea is a timely case. Since 2022 it has reorganized national science and technology policy around 12 National Strategic Technologies, explicitly assigning its government-funded research institutes (GRIs) a directive role in shaping new technological trajectories — the posture Mazzucato (2013) theorizes as the “Entrepreneurial State.” The paper first frames this as a paradigm shift: from New Public Management-era output accountability, under which GRIs were evaluated on measurable research outputs, toward a model demanding directionality, mobilization of private actors, and transformative systemic impact.
Against this backdrop the paper presents a systematic content analysis of performance indicators drawn from the annual evaluation frameworks issued between 2012 and 2026 and from the institutional assessments of 22 of the 23 GRIs under Korea’s National Research Council of Science and Technology (NST) — the complete population, excluding only one security-sensitive institute whose evaluation documents are not disclosed. Each indicator is coded for alignment with either the traditional output-accountability paradigm or the Entrepreneurial State paradigm. The analysis reveals a patterned misalignment: indicators remain heavily concentrated in attributable, institution-level outputs — publications, patents, and licensing revenue — while direct measures of mission directionality, private investment leverage, and systemic transformation are structurally absent across institution types.
The paper then asks why the gap persists, proposing three mechanisms and assessing each against the indicator record. First, path-dependent measurement infrastructure built during the NPM era generates inertia: goal-level panel data show indicators carried over even through the 2021 and 2026 redesigns of the evaluation system itself. Second, the attribution problem inherent to mission-oriented research makes systemic outcomes hard to link to individual institutions, creating incentives to retreat to bounded metrics; the near-complete absence of shared-attribution formulas is itself the evidence. Third, a shared preference for defensible, quantifiable measures biases selection against diffuse, long-horizon outcomes — though the 2021 withdrawal of quantitative goal-attainment assessment suggests that quantitative accountability was relocated rather than relaxed. The record supports the first two mechanisms and constrains the third without confirming it.
Building on these findings, the paper proposes a normative framework for mission-aligned evaluation design. Drawing on OECD guidance on mission-oriented innovation policy assessment and recent work in research impact methodology, it articulates design principles rather than prescriptive indicators, organized around four dimensions: mission directionality, actor mobilization, systemic transformation, and adaptive learning capacity. Without deliberate realignment along these dimensions, performance management systems risk incentivizing misaligned behavior regardless of how ambitious the stated policy goals are. The findings carry implications for any government navigating the tension between accountability demands and the uncertain, long-horizon nature of mission-oriented public investment.

Author

• Hye Yeon Kwon, Korea Institute of Science and Technology
  Presenting Author