Policy space dashboard
01 — Research · Room to Respond
Two indices of policy space
The Policy Buffers Index and the Effective Space Index, from Room to Respond. Pick countries from the map or the search box, choose an index, and set the year range to compare them over time.
Preliminary research output — please read before using these numbers
- This is not final work. Room to Respond is an unpublished working paper. Both indices, their inputs, and the estimates behind them may change materially before publication. Nothing here should be cited as a settled result.
- The views expressed are those of the author and do not necessarily reflect the views of the International Monetary Fund, its Executive Board, or IMF Management.
- Not a country risk assessment. Neither index is a rating, a forecast, a measure of creditworthiness, or advice of any kind. They are research measures built to answer one narrow question about crisis response.
- Adjusted and subject to revision. The values plotted here have been smoothed and coarsened, and may be revised at any time without notice.
The underlying data is not published yet
The dashboard reads /files/policy-space.csv. That file is not in the
repository, so there is nothing to plot — deliberately, rather than showing invented
numbers.
The expected format is one row per country-year:
iso3,country,year,pbi,esi
USA,United States,2019,26.0,38.4
PHL,Philippines,2019,57.2,61.0iso3 is the ISO 3166-1 alpha-3 code, pbi and esi
are within-year percentile scores from 0 to 100, and either may be left blank where a
country is unranked in that year. Drop the file in and this page comes to life with no
other change.
How the indices are built, and what they cannot see
Policy Buffers Index (PBI). An equal-weighted average of six observable components: the inverse debt ratio, the fiscal balance, the real policy rate, the current account balance, foreign reserves over GDP, and the inverse long-rate spread over the United States as a market-access measure. Each is signed so higher means more room, converted to a within-year percentile rank before averaging. No estimation is involved, so anyone can audit it.
Effective Space Index (ESI). The within-year percentile rank of an effective-space score from an honest causal forest trained on banking-crisis severity, using the Laeven and Valencia (2026) chronology over 1970 to 2019. It changes both the recipe and the ingredients relative to the PBI.
Coverage. Just under 200 economies, from the late nineteenth century to 2024, with far thinner coverage in the earlier decades than in recent ones.
What is plotted here is not the raw estimate. The series on this page have been adjusted and smoothed, and are coarser than the estimates in the paper. They should be read for shape and broad position rather than for precise values, and anyone needing the underlying figures should work from the paper.
How to read these numbers
Both indices are within-year percentile ranks, so a score is a country's position among those reporting in that year rather than a quantity of fiscal space, and a country can move simply because others did. Coverage is uneven and thinner in earlier years and outside the advanced economies, inputs arrive with error and reporting lags, and countries can be scored on partial information, so neighbouring ranks are not meaningful distinctions. The indices can only reward what their inputs measure, which leaves out capacities such as reserve currency status, standing swap lines, deep domestic investor bases and institutional strength — an omission that mechanically favours some countries over others and produces individual rankings the paper itself treats as implausible. The estimated index carries materially more uncertainty than the observable one, is sensitive to specification, and is presented as a companion to it rather than a replacement. The series shown here have been adjusted and smoothed, so they are coarser than the underlying estimates, and rankings will not line up with the tables in the paper, which restrict the sample. All of it is preliminary and may be revised at any time, without notice and without the charts on this page being marked as changed. Read the paper before drawing conclusions from any single country's series.
Source: author's calculations, from Room to Respond. Country outlines from Natural Earth (public domain); boundaries and names do not imply any judgement on the legal status of any territory. Northern Cyprus, Somaliland and Kosovo are absent from the basemap as they carry no ISO 3166-1 alpha-3 code.