Comparing Like With Like: Dynamic Peer Comparison in Indonesia
Most sub-national indices score Indonesia’s 514 local governments against a single national benchmark, which tends to reward structural advantage over governance quality. The Regional Government Success Scorecard addresses this through Dynamic Peer Comparison, giving every local government a fairer footing.
Imagine two local governments in geographically diverse Indonesia, both being scored on an index. It would be difficult to directly compare a resource-rich regency with better connectivity with another that has limited resources and higher connectivity costs. A ranking that places them on the same scale and holds them to the same benchmark is not really measuring how well a local government does its job. It is measuring, in large part, where each government happens to be.
This is the challenge that Dynamic Peer Comparison (DPC) is designed to address. The Regional Government Success Scorecard (RGSS) uses DPC to compare each of Indonesia’s 514 regencies (kabupaten) and cities (kota) against structurally similar peers rather than a single national standard. In the Indonesia pilot, two structural factors form the basis of that comparison: geographic conditions and natural resource endowment, including terrain, isolation, and disaster risk. The factors a government cannot change are accounted for. What remains is a clearer picture of governance quality.
The logic: making the right comparisons
Think of two runners who both finish a race in four hours. On the face of it, they performed identically. But one ran on a flat sea-level course; the other ran a high-altitude mountain route. The same finish time, under very different conditions, represents a very different achievement. Race organisers have long known this: when ranking runners across events, you do not simply compare times. You compare each runner against others who ran the same course under the same conditions.
DPC applies this logic to local governments. Geography and natural resources are the two factors a local government genuinely cannot change: a regency cannot lower its earthquake risk, and cannot create mineral wealth the ground does not contain. How well a government manages its budget, how effectively it uses technology, and how well public services function: these are within reach of good governance. DPC separates the former from the latter, so the quality of governance can be assessed on its own terms.
Rather than measuring each kabupaten or kota against a single national standard, DPC compares each one against structurally similar peers: local governments within the same class, either kabupaten or kota, with comparable geographic conditions and natural resource endowments. A disaster-prone regency in eastern Indonesia is still compared against all other regencies, but the comparison is weighted: regencies with more similar geographic conditions and natural resource endowments carry more weight in the benchmark, and those that are structurally very different carry less. The weighting is continuous rather than categorical. The more structurally similar two regions are, the more weight each carries in the other’s benchmark. This comparison is constructed dynamically across all 514 local governments.
What the data shows
In the RGSS Indonesia pilot, the impact of DPC is most visible in the gap between kota and kabupaten. Before DPC is applied, the average kota sits at the 63rd percentile of the national score distribution, while the average kabupaten sits at the 47th. That 16-percentile-point gap reflects structural advantage, not governance quality. After DPC, both groups average the 50th percentile, with their distributions overlapping almost completely. The structural noise is removed. What remains is a more meaningful basis for comparison.
The overall ranking order does not change dramatically. When DPC is applied, the overall ranking order stays largely intact, with most local governments moving only a few places. DPC is not an equaliser that flattens all differences. It is a corrective that removes the structural component from scores so genuine governance differences can surface.
For about one in five local governments, 103 of the 514, the shift is substantial. They moved more than 50 places in the rankings once DPC is applied. Among those rising are predominantly kabupaten that scored modestly on raw national measures but emerge as genuine high performers against their structural peers.
DPC in practice: the case of Kabupaten Nias
Kabupaten Nias illustrates the impact of DPC. Sitting off the western coast of Sumatra, Nias is one of Indonesia’s most geographically constrained regencies. It ranks 504th out of 514 on geographic conditions, reflecting its island isolation, difficult terrain, and history of natural disasters. On natural resource endowment, it ranks 367th. Without accounting for these structural realities, Nias ranked 312th overall, solidly in the lower half of the national table.
Once DPC is applied, Nias rises to 210th, a jump of more than 100 places. Nias is now assessed against local governments facing similar structural conditions, and against those peers, its performance is considerably stronger than the raw ranking suggested. For a regency government working under genuine constraints, that is a meaningful recognition.
What this means for leaders and policymakers
For a Regent (Bupati) or Mayor (Walikota) reading their RGSS score, a positive DPC-adjusted result carries a specific meaning: their government is outperforming places facing the same structural conditions. That is a genuine achievement. It suggests strong performance under constraint, and one that a raw national ranking would not have surfaced.
For national and provincial governments, as well as development partners and donors, DPC changes what the Scorecard is useful for. A ranking built on raw scores may effectively identify where the well-resourced places are. A DPC-adjusted ranking identifies where the better-performing places are. Those are not always the same.
The RGSS website allows users to explore the scores and ranks of each dimension and to filter local governments by administrative and governance types, GDP per capita, population, and structural characteristics such as archipelagic status. This last filter is particularly relevant for Indonesia’s unique ‘Archipelagic’ local governments. The term is defined as those with a sea area larger than their land area. These archipelagic local governments continue to average significantly below non-archipelagic peers even after DPC adjustment, suggesting their geographic circumstances may impose governance costs that the current methodology reduces but does not entirely eliminate. This is an area for refinement as richer data becomes available. The RGSS goal is not comparison for its own sake, but to give governments and their partners a more honest and actionable basis for understanding where governance quality is strongest, and where support is most needed.
Methodological Note
Foundational Environment factors: In the Indonesia pilot, DPC is based on two structural factors: Natural Resources, proxied by the ratio of natural resource revenue sharing to regional revenue (Ministry of Finance); and Geographical Condition, proxied by the ‘Construction cost index’ (Badan Pusat Statistik, the National Statistics Office) and the ‘Disaster risk index score’ (Badan Nasional Penanggulangan Bencana, the National Board for Disaster Management).
Pre- and post-DPC scores are drawn from the April 2025 pilot dataset covering all 514 Indonesian local governments. The Spearman correlation between pre- and post-DPC overall rankings is 0.962. Full technical detail on the DPC methodology, including the construction of the Foundational Environment peer-grouping signal and the inverse-distance weighting formula, is available in the RGSS Implementation Plan.
About the Authors
Damien Huang is the Deputy Director (Data Analytics & Research) at the Chandler Governance Group.
Wen Haoyu is the Manager (Data Analytics & Research) at the Chandler Governance Group.
Related Articles
