Warsh & the Fed

Replication repository for the deep dive Inheriting Discretion: What Monetary Rules Prescribe for the Warsh Fed.

R pulls quarterly macroeconomic series from FRED and scores the actual federal funds rate against four monetary-rule benchmarks. Python reads the scored data and generates Chart.js widgets and static figures. The pipeline mirrors the Taylor-rule deep dive repo so the two projects read as one research program.

Data vintage

FRED revises historical series, including GDPPOT (CBO potential output) and GDP (nominal, subject to benchmark revision). Results reproduce exactly against the vintage retrieved for the published piece; subsequent pulls may shift rule prescriptions. The sample window is locked at 2018 Q1–2026 Q1 in R/config.R (START_DATE, END_DATE).

Requirements

• R (≥ 4.0); packages auto-install via R/setup.R: fredr, dplyr, zoo, lubridate
• Python (≥ 3.10): numpy, pandas, matplotlib
• A free FRED API key — https://fredaccount.stlouisfed.org/apikeys

Setup

cp .Renviron.example .Renviron   # edit and paste your FRED_API_KEY
pip install numpy pandas matplotlib

R loads .Renviron from the project root automatically. Run both commands from the repo root; run_all.R will error if called from a subdirectory.

Run

Rscript run_all.R        # FRED pull → rule scoring → output/data/rules_export.csv
python python/charts.py  # rules_export.csv → widgets + figures

run_all.R also prints a key-numbers block (latest quarter rates, peak deviations) for use in prose.

FRED series

Series ID	Description	Transformation
FEDFUNDS	Effective federal funds rate	levels
PCEPILFE	Core PCE deflator	YoY% (pc1)
GDPC1	Real GDP	levels
GDPPOT	CBO potential real GDP	levels
GDP	Nominal GDP	levels
M2SL	M2 money stock	levels

Monthly series convert to quarterly by taking the last observation of each quarter.

Rule specifications

Rule	Formula	Key parameters
Taylor (baseline)	r* + π + 0.5(π − π*) + 0.5·y_gap	r* = 0.5%, π* = 2%
Taylor (original)	same	r* = 2.0% (Taylor 1993)
Meltzer k-percent	M2 YoY growth vs. fixed target	k = 4%
Selgin productivity norm	NGDP growth vs. potential GDP growth	NGDP trend = 4%/yr from 2019 Q4

Taylor rules are floored at zero. Output gap = 100·(GDPC1 − GDPPOT)/GDPPOT. The NGDP level path compounds at 1% per quarter from the 2019 Q4 base (NGDP_BASE_DATE in R/config.R).

Where things can drift

GDPPOT is revised with each CBO Budget and Economic Outlook, shifting the Taylor output-gap term. GDP benchmark revisions alter both the Meltzer M2 deviation and the NGDP gap. The Meltzer and Selgin rules are more stable to revision than the Taylor variants. The companion piece documents the vintage used.

Layout

Script	Role
R/01_data.R	FRED pull → quarterly conversion → output/data/df_raw.rds
R/02_rules.R	Rule scoring (Taylor, Meltzer, Selgin, NGDP path) → df_rules.rds
R/03_export.R	CSV export + prose key-numbers → output/data/rules_export.csv
python/charts.py	Chart.js widgets + PNG figures from rules_export.csv
R/config.R	Sample window, π*, r*, k-percent, NGDP trend
R/setup.R	Package management

License

Source data is public, served by FRED and subject to its terms. The code and generated figures are licensed under CC BY 4.0 — reuse with attribution.