The DataSkinny modeling system is structured into three distinct layers, moving from high-level economic forces down to hyper-granular operational realities:
- Layer 1: Sector Level (Public) – High-level macro indicators mapping out roughly 70 major U.S. economic sectors. This establishes the structural baseline.
- Layer 2: Commercial Line-of-Business Level (Public) – The "commercial grade" sweet spot. This layer tracks 400 more granular types of economic activity, delivering highly credible, rigorously modeled insights built for strategic decision-making.
- Layer 3: Granular Micro-Analytics (Proprietary/Non-Public) – An ultra-deep, restricted layer that unpacks the underlying mechanics inside specific lines of business for more highly focused institutional analysis.
The specific sectors and lines of business covered in Layers 1 and 2 are fully documented in the attached PDF.
A quick note on Layer 3: I currently use this ultra-granular coverage internally to backstop and stress-test my broader research. At this depth, micro-noise increases significantly, which is why I keep these models proprietary. While I treat these deep-tier analytics with a healthy dose of professional skepticism, they provide invaluable context. In the future, I may open access to select portions of Layer 3, accompanied by the appropriate framework for handling such volatile, hyper-local data.
Raw Data Sources
The DataSkinny model relies on a robust network of primary, authoritative federal data sources. To bridge the gap between high-level macroeconomic trends and granular micro-level realities, the modeling system processes data through two distinct lenses: benchmarking (building the structural foundation) and time series (updating the system dynamically to reflect the present day).
Structural Benchmarking Sources
These foundational datasets establish the baseline economic structural relationships, inter-industry links, and demographic baselines for the model:
- Annual & Benchmark Input-Output Accounts (Bureau of Economic Analysis) – Maps how industries interact, buy from, and sell to one another.
- U.S. Economic Census (U.S. Census Bureau) – Provides the comprehensive 5-year bedrock baseline for American business activity.
- Annual Business Survey (U.S. Census Bureau) – Supplies critical annual updates on firm demographics, innovation, and operational characteristics.
- The National Employment Matrix (Bureau of Labor Statistics) – Details the precise occupational staffing patterns across different industries.
Dynamic Time Series Sources
These high-frequency datasets inject real-time momentum into the model, continuously updating the structural benchmarks to reflect current economic conditions:
- Industrial Production Indexes (Federal Reserve Board) – Tracks real-time changes in output across manufacturing, mining, and utilities.
- Producer & Wholesale Price Indexes (PPI) (Bureau of Labor Statistics) – Monitors inflation pressures and commodity price shifts before they hit consumers.
- Import / Export Price Indexes (MXP) (Bureau of Labor Statistics) Captures international trade price dynamics, currency impacts, and import inflationary shocks.
- Consumer Price Indexes (CPI) (Bureau of Labor Statistics) – Monitors inflation pressures at the consumer level.
- Employment, Earnings, & Cost Indexes (ECEI) (Bureau of Labor Statistics) – Measures labor market tightness, wage growth, and total employer compensation costs.
- Current Employment Statistics (CES) State & Area Program (Bureau of Labor Statistics) – Delivers highly specific, monthly data on nonfarm payrolls, hours, and earnings.
