FRED – US Consumer Price Index (CPI) Annual
- Rows
- 1,365
- Columns
- 2
- Data range
- 1913-01-01 to 2026-09-01
- Auto-refresh
- Re-checks its source daily
- Last refreshed with new data
- 2026-10-14
Annual US Consumer Price Index for All Urban Consumers (CPI-U, not seasonally adjusted) from 1913 to present, tracking inflation over more than a century.
AI analysis
FRED US Consumer Price Index (CPI) – Dataset Analysis
1. Dataset Overview & Value for Correlation Studies This dataset contains the US Consumer Price Index for All Urban Consumers (CPI-U, not seasonally adjusted), sourced directly from the Federal Reserve Bank of St. Louis (FRED) API at [fred.stlouisfed.org](https://fred.stlouisfed.org/graph/fredgraph.csv?id=CPIAUCNS). The FRED platform is one of the most authoritative and continuously maintained macroeconomic data repositories in the world, lending high credibility to this series. Spanning from 1913 to the present across 1,360 monthly observations (approximately 113+ years), this dataset is exceptionally valuable for long-horizon economic correlation studies. Its temporal breadth captures transformative economic eras — the Great Depression, post-WWII expansion, 1970s stagflation, the Great Moderation, and post-COVID inflation surges — making it a cornerstone variable for virtually any macroeconomic analysis.
---
2. Data Quality Observations Overall data quality is very strong for a dataset of this age and scope. There is only 1 null cell across the entire 2-column, 1,360-row dataset, confined entirely to the CPIAUCNS column (0.1% null rate), which is negligible and likely represents a single missing monthly observation rather than a systemic gap. There are no type mismatches reported, and the Date column shows 1,360 distinct values with zero nulls, confirming a clean, gapless time index with no duplicate dates. Duplicate row counts are pending final Parquet recomputation (Phase D), but given the perfect date distinctness, structural duplicates are highly unlikely. The single null in CPIAUCNS should be imputed (e.g., linear interpolation between adjacent months) before any time-series modeling to avoid downstream calculation errors.
---
3. Key Column Distributions & Statistical Insights The CPIAUCNS column tells a rich story through its statistics. The mean of 91.87 versus a median of 37.00 signals a heavily right-skewed distribution (skewness = +0.925), which is expected — CPI values were very low in the early 20th century and have accelerated dramatically in recent decades. The min of 9.7 (circa early 1913) versus a max of 333.02 (recent years) reflects over a 34x increase in the price level across the dataset's lifetime. The wide interquartile range (Q1 = 17.65 → Q3 = 161.55, spread of ~144 points) and a large standard deviation (σ = 90.47, nearly equal to the mean) further confirm that the bulk of price-level growth is concentrated in the latter half of the series. For correlation analysis, working with CPI percentage change (YoY inflation rate) rather than raw index levels is strongly recommended, as the raw series is non-stationary and will produce spurious correlations with other trending variables.
---
4. Recommended Join Key Columns The Date column is the clear and unambiguous join key, with 1,360 distinct values, zero nulls, and apparent monthly granularity. It should be standardized to ISO 8601 format (YYYY-MM-DD) before joining to ensure compatibility with other FRED series or financial datasets. For datasets with only annual granularity, dates can be aggregated to year-end (December) or annual averages to enable cross-dataset merges. Notably, the monthly frequency makes this dataset a flexible join partner — it can align with both monthly and quarterly (via resampling) economic datasets without significant information loss.
---
5. Suggested Companion Datasets for Correlation Analysis
| Dataset | Expected Correlation | Rationale | |---|---|---| | Federal Funds Rate (FEDFUNDS – FRED) | Strong negative (lagged) | Fed raises rates to combat inflation; classic monetary policy relationship | | US Unemployment Rate (UNRATE – FRED) | Negative (Phillips Curve)| Core macroeconomic theory linking inflation and unemployment | | M2 Money Supply (M2SL – FRED) | Strong positive | Monetary expansion is a leading indicator of inflationary pressure | | S&P 500 / DJIA Historical Prices | Complex/non-linear | Real vs. nominal returns; inflation erodes equity purchasing power | | US GDP (GDP – FRED) | Positive long-term | Demand-pull inflation correlates with economic growth cycles | | Crude Oil Prices (DCOILWTICO – FRED) | Strong positive | Energy costs are a primary CPI driver, especially in headline CPI | | US Average Hourly Earnings (CES0500000003) | Positive | Wage-price spiral dynamics; labor market inflation pass-through |
For the most analytically rigorous correlations, all paired series should be differenced or log-transformed to achieve stationarity before computing Pearson or Spearman correlations, and cross-correlation functions (CCF) should be used to identify meaningful lead-lag relationships rather than contemporaneous correlations alone.
Columns
- Date (date)
- CPIAUCNS (decimal)