Complete transparency. All formulas verifiable. All sources cited. Download everything.
The Deflation Index tracks technological cost reductions across four fundamental sectors: computing, communications, energy, and transportation. Each sector is measured using cost-per-performance metrics that capture the true force of technological progress.
The DI is a weighted geometric mean of these four sector indices, showing how technological deflation compounds over time.
Why geometric mean? Technological improvements multiply (compound), they don't add. A geometric mean properly captures this multiplicative relationship.
Where each sector represents fundamental technological progress:
Current Scope (v3.0.3): Four sectors covering approximately 40% of technological impact. Future versions will add healthcare, education, and housing sectors with rebalanced weights. See expansion roadmap →
Note on Future Expansion: v4.0 will add healthcare, education, and housing sectors. Current v3.0.3 composition uses four sectors with normalized weights totaling 100%.
Why geometric mean? Technological improvements multiply (compound), they don't add. A geometric mean captures this multiplicative relationship.
Sector weights (29.4%, 23.5%, 29.4%, 17.6%) reflect three factors:
See detailed BEA analysis and complete methodology on GitHub →
To ensure our findings aren't artifacts of weight choices, we tested multiple methodologies:
| Methodology | Weights (C/Co/E/T) | Cumulative Deflation |
|---|---|---|
| Multi-Factor (Primary) | 29/24/29/18 | -96.25% |
| Equal-Weighted (Baseline) | 25/25/25/25 | -95.04% |
Key Finding: Both methodologies show massive technological deflation (95-96% cumulative). The core finding is robust regardless of weighting choices.
See complete sensitivity analysis on GitHub →
All data comes from authoritative, publicly available sources. Every data point includes source citations in the Excel files.
| Source | What We Use | Reliability |
|---|---|---|
| Federal Reserve (FRED) | M2 money supply, CPI data | 100/100 (A+) |
| Bureau of Labor Statistics | CPI components, historical pricing | 100/100 (A+) |
| IRENA | Solar LCOE, renewable energy costs | 95/100 (A) |
| BloombergNEF | Battery costs, EV economics | 95/100 (A) |
| AI Impacts | Computing power costs (FLOPS) | 90/100 (A-) |
| DOE | LED efficiency, transportation data | 95/100 (A) |
| FCC | Broadband pricing, network costs | 90/100 (A-) |
| Backblaze | Hard drive cost tracking | 85/100 (B+) |
Average source reliability: 92/100 (A-grade). We prioritize government agencies, international organizations, and industry leaders with transparent methodologies.
Why provide an alternative? To demonstrate that our core finding holds regardless of weighting assumptions. We provide our Multi-Factor methodology plus an equal-weighted baseline—the simplest possible approach with zero subjective choices.
Methodology: Balances three factors: GDP contribution, enabling effect (how much other sectors depend on it), and deflationary force (magnitude of cost reductions).
Weights: Computing 29.4% • Communications 23.5% • Energy 29.4% • Transportation 17.6%
Use case: Recommended default. Most comprehensive approach accounting for both direct and indirect economic impact.
Download PrimaryMethodology: Simple equal weighting across all sectors. No assumptions about relative importance—pure mathematical baseline with zero subjective choices.
Weights: Computing 25% • Communications 25% • Energy 25% • Transportation 25%
Use case: Zero-assumption baseline. If you disagree with our weighting methodology, this shows the finding still holds.
Download Equal-Weighted| Methodology | 2024 DI | Cumulative | Annual Avg |
|---|---|---|---|
| Multi-Factor (Primary) | 3.74 | -96.25% | -9.21% |
| Equal-Weighted (Baseline) | 4.96 | -95.04% | -8.45% |
Key Finding: Both methodologies show massive technological deflation (95-96% cumulative). The core finding is robust regardless of weighting choices.
When we expand to 7+ sectors (adding healthcare, education, housing), we plan to implement rigorous expenditure-weighted and GDP-weighted methodologies derived directly from BLS Consumer Expenditure Survey and BEA GDP-by-Industry data. We welcome collaboration from economists interested in helping develop this methodology.
We're transparent about where the Deflation Index is currently limited. Better to acknowledge constraints openly than to overstate capabilities.
Sectors Covered (v3.0.3):
Coverage: The current four sectors represent approximately 40% of measurable technological deflation, not 100%. We prioritize defensible measurement over comprehensive coverage—better to measure four sectors perfectly than twenty sectors poorly.
Expansion Timeline: v4.0 (2026-2027) will add healthcare, education, and housing sectors with rebalanced weights across all seven sectors.
Weighting System:
Index Construction:
Conservative Assumptions:
Why These Limitations: We prioritize verifiable, reproducible data over comprehensive coverage. Each limitation represents an opportunity for future improvement and is documented for transparency.
Current:
Planned:
These limitations don't invalidate the core finding: massive technological deflation occurred (-96.26% cumulative) while consumer prices rose (+155%). This divergence exists regardless of methodological choices.
What changes with better methodology: The precise magnitude of the gap. Not whether it exists.
We welcome critique: Challenge our assumptions, propose alternatives, suggest improvements. The goal is the best measurement possible, not defending any particular method.
For comprehensive technical details, see our GitHub repository:
All data is public, all calculations are transparent, all assumptions are documented. Download the Excel files and verify every number.