Business Applications: Construction in the United States (BABANAICS23NSAUS) | FRED | St. Louis Fed
- Rows
- 263
- Columns
- 2
- Data range
- 2004-07-01 to 2026-05-01
Graph and download economic data for Business Applications: Construction in the United States (BABANAICS23NSAUS) from Jul 2004 to May 2026 about business applications, construction, business, and USA.
AI analysis
Analysis: Business Applications: Construction (BABANAICS23NSAUS)
1. Dataset Overview & Value
This series, sourced from the Federal Reserve Bank of St. Louis (FRED) via fredgraph.csv, tracks monthly counts of new business applications filed under the NAICS 23 (Construction) sector in the United States, unadjusted for seasonality (NSA). As a direct API pull from FRED, the data benefits from a highly reputable, regularly maintained government-affiliated source, and the stated coverage through May 2026 suggests either forward-looking projections/registrations or that this snapshot was pulled with an extended forecast window — worth verifying against FRED directly for the true "last actual" observation date. As a leading indicator of construction-sector economic activity, this series has strong potential value for correlation studies against housing starts, construction employment, mortgage rates, GDP components, or regional economic indicators, since business formation often precedes actual construction spending and hiring.
2. Data Quality Observations
The dataset is clean on the surface: zero null cells across both columns and zero missing dates in the observationdate field (263 distinct dates matching 263 rows, indicating no duplicate timestamps). However, the duplicate-row check is explicitly marked as pending ("Phase D — recomputed against Parquet"), so full deduplication cannot yet be confirmed — this should be revisited before finalizing quality sign-off. One subtle flag: the value column shows 262 distinct values out of 263 rows, meaning exactly one repeated value exists in the series — not a data quality concern per se, but worth noting for anyone doing exact-match joins or uniqueness assumptions.
3. Key Column Distributions
The core metric, BABANAICS23NSAUS, ranges from 14,654 to 62,652 applications per month, with a mean of 29,744.9 and median of 26,681 — the mean exceeding the median signals right-skew, confirmed by a skewness of 0.806, indicating a longer tail toward higher application counts (likely reflecting boom periods or post-pandemic surges). The interquartile range (Q1=21,697.5, Q3=37,155) spans roughly 15,500 units, and 2 outliers were flagged, plausibly tied to anomalous spikes (e.g., 2020–2021 pandemic-era business formation surges) or reporting artifacts. This distribution shape suggests analysts should consider log-transformation or robust statistics (median, IQR) rather than mean-based comparisons when benchmarking against other series.
4. Recommended Join Key
The observationdate column is the clear and only viable join key — it's a clean, non-null, monotonic monthly date field. For cross-dataset correlation, ensure other datasets are aligned to first-of-month timestamps (this series uses YYYY-MM-01 format) and matching frequency (monthly); resampling will be needed for daily/quarterly series.
5. Complementary Datasets for Correlation Discovery
Strong pairing candidates include: Housing Starts (HOUST) and Building Permits (PERMIT) for construction pipeline validation; Construction Employment (USCONS) from BLS/FRED to test whether business applications lead hiring; 30-Year Mortgage Rates (MORTGAGE30US) to examine rate-sensitivity of new construction ventures; Producer Price Index for Construction Materials to explore cost-driven formation trends; and Total Business Applications (BABATOTALSAUS) as a benchmark to isolate construction-specific versus economy-wide entrepreneurial activity. Regional Federal Reserve or Census Bureau construction spending data (e.g., TTLCONS) would also help distinguish application volume from actual capital deployment.
Columns
- observation_date (date)
- BABANAICS23NSAUS (integer)