---
name: us-treasury-fiscal-data-ingestion
description: "How to get U.S. Treasury Fiscal Data API data into a pandas DataFrame and Parquet: access, the request, the format, limits, quirks and terms."
---

# Load U.S. Treasury Fiscal Data into pandas and Parquet

Source entry: https://signal-lab-jet.vercel.app/datasets/us-treasury-fiscal-data
Provider: https://fiscaldata.treasury.gov/api-documentation/
Checked against the provider's documentation: 2026-10-29

## Access
- No account, registration or API key.

## Request
```bash
curl 'https://api.fiscaldata.treasury.gov/services/api/fiscal_service/v2/accounting/od/debt_to_penny?fields=record_date,tot_pub_debt_out_amt&sort=-record_date&page%5Bsize%5D=5'
```

## Format
JSON: a data array of rows plus meta (count, labels, dataTypes, total-count, total-pages) and links. Numbers and dates arrive as strings.

## Load into pandas and Parquet
Needs Python 3 with requests, pandas and pyarrow.
```python
"""Debt to the Penny (U.S. Treasury Fiscal Data API) -> pandas -> Parquet. No key."""
import requests
import pandas as pd

URL = "https://api.fiscaldata.treasury.gov/services/api/fiscal_service/v2/accounting/od/debt_to_penny"
params = {"fields": "record_date,tot_pub_debt_out_amt", "filter": "record_date:gte:2020-01-01",
          "sort": "record_date", "page[size]": 500, "page[number]": 1}
rows = []
while True:
    r = requests.get(URL, params=params, timeout=30)
    r.raise_for_status()
    body = r.json()
    rows += body["data"]
    if params["page[number]"] >= body["meta"]["total-pages"]:
        break
    params["page[number]"] += 1

df = pd.DataFrame(rows)
df["record_date"] = pd.to_datetime(df["record_date"])
df["tot_pub_debt_out_amt"] = pd.to_numeric(df["tot_pub_debt_out_amt"])  # values arrive as strings
df.to_parquet("us_treasury_debt_to_penny.parquet", index=False)
print(len(df), "rows", df["record_date"].min().date(), "to", df["record_date"].max().date())
```
Run by us on 2026-10-29 (python 3.13.5 | pandas 3.1.0 | pyarrow 26.0.0 | requests 2.34.2). Result: 1,714 daily rows, 2020-01-02 to 2026-10-28.

## Limits
- Rate limit: Not stated.
- Pages default to 100 rows (page[size]); the snippet asks for 500 and follows meta total-pages.

## Quirks
- Values are strings (for example "40551054852362.41"); convert with pd.to_numeric.
- Debt to the Penny is posted at the end of each business day with the previous business day's data, so there are no weekend or holiday rows.
- Each dataset has its own endpoint path and field names, listed in the API documentation.

## Terms and attribution
- Treasury states the data is free "without restriction" for non-commercial or commercial use. No attribution requirement is stated.
- Terms page: https://fiscaldata.treasury.gov/api-documentation/

## Sources
- https://fiscaldata.treasury.gov/api-documentation/: "Our API is open, meaning that it does not require a user account or registration for a token."
- https://fiscaldata.treasury.gov/api-documentation/: "The data is offered free, without restriction, and available to copy, adapt, redistribute, or otherwise use for non-commercial or commercial purposes."
- https://fiscaldata.treasury.gov/datasets/debt-to-the-penny/: "Debt to the Penny is updated at the end of each business day with data from the previous business day."

Information for educational purposes only. Free tiers, limits and licenses belong to the providers and may change; check the provider's own terms before using the data.