About & Methodology
This platform is a model, not an oracle. Data is sourced from official Australian government publications, peer-reviewed institutional sources, and clearly-flagged estimates. This page explains how the scores are calculated, where the data comes from, and what the limitations are.
About the Author
Christopher J. Howlett is the author of Unprepared: Australia in an Age of Chinese Power. The book represents a 215,000-word argument for comprehensive national transformation across eight interdependent pillars of strategic policy.
This platform is the book's companion — making the argument visible, interactive, and continuously updated with real-world data.
How Scores Are Calculated
Variable Score (0-100)
Each tracked variable is scored by its progress toward its programme target. If the current value equals the target, the variable scores 100. If the current value is at the worst-case baseline, the variable scores 0. Values worse than baseline are clamped to 0.
Pillar Score (0-100)
Each pillar score is a weighted average of its tracked variables. Variables with high confidence receive full weight. Estimates receive 80% weight. The weighting within a pillar reflects the relative importance of each variable to the pillar's core objective.
Overall Preparedness Score (0-100)
The overall score is a weighted average of all eight pillar scores. Pillars 1, 2, and 6 (Resource Sovereignty, Fiscal Engine, Maritime Defence) receive higher weights as foundational pillars. A score of 50 represents adequate deterrence. Current score: 9/100.
Data Sources
| Source | Update Frequency | Pillars | Confidence |
|---|---|---|---|
| ABS (Australian Bureau of Statistics) | Quarterly/Annual | 1, 2, 4, 5 | High |
| Defence Budget Papers | Annual (May) | 2, 6 | High |
| ANAO Major Projects Reports | Annual | 3, 6 | High |
| Dept of Climate Change — Fuel Monitor | Quarterly | 6 | High |
| Pentagon Annual Report on Chinese Military Power | Annual | 6, 7 | Medium |
| SIPRI Military Expenditure Database | Annual | 2, 6, 7 | High |
| IISS Military Balance | Annual | 6, 7 | High |
| DFAT Trade Statistics | Annual | 5 | High |
| NDIA Annual Report | Annual | 2 | High |
| Budget Papers / MYEFO | Annual + mid-year | 2 | High |
| RBA Statistical Tables | Monthly | 2 | High |
The datasets above are the FACT baseline — the world as it is — refreshed from their sources. Live developments reach the platform two ways: a news-intelligence pipeline (RSS, Reddit and Google News, classified by Claude) and GDELT press-coverage monitoring (descriptive coverage volume + tone, queried from BigQuery; see the Signals page). Both only propose changes into a human-approval queue — neither moves a score automatically. Programme targets, by contrast, are fixed to the book's argument (PRESCRIPTION) and do not move with the news.
Limitations
- · Scenario probabilities are qualitative estimates, not statistically derived forecasts.
- · Chinese military data is inherently uncertain. PLA figures use Pentagon estimates as a baseline.
- · The fifty-year trajectory model uses simplified assumptions. Real outcomes will differ.
- · Live developments are surfaced automatically — by a news pipeline (RSS, Reddit, Google News, classified by Claude) and by GDELT press-coverage monitoring — but every change is human-reviewed before it moves a number, so some lag between an event and a platform update is expected.
- · This is a partisan argument as well as an analytical one. The programme prescriptions reflect the author's judgement about what Australia should do.
What's Been Built
Every layer below is built and live.
- 8-pillar tracking + 15 scenarios
- News ingest + AI classification
- Threat tracker & nation comparisons
- Admin panel + human approval flow
- Strategy Simulator (50-year model)
- All platform data DB-backed
- Automated weekly digest (Substack post + data summary)
- Email capture
- Score-history charting
- Dynamic social share cards
- Public API
- Source-credibility tiers
- Semantic duplicate detection
- Manual article entry
- Live-updating monthly report, trend analysis & historical comparison
- Descriptive coverage monitoring — volume + tone — via BigQuery (the Signals feed)
- Human-gated GDELT → proposal loop
- Forecasting: rigorously tested, not validated — parked, nothing published
Contact & Corrections
If you identify a data error, an outdated figure, or a methodological flaw, please reach out via the Substack.
powerprogressaustralia.substack.com