The Deficit and Interest Figures Check Out — But the Article's "Spending Is Slowing" Inference Rests on Comparing a 10-Month Number to an 11-Month Number, Plus a Revenue Item Miscounted as Spending

Summary

Every individual dollar figure in Craig Eyermann's Sept 23, 2026 Independent Institute piece checks out: the ~$2.1 trillion FY2026 deficit, the $125.2 billion tariff refund, the $117 billion (14%) rise in interest costs, and the roughly doubled, $40-trillion-plus national debt are all corroborated by Treasury/CBO data and AskAmerica's own Treasury Monthly Treasury Statement (MTS) table. But the article's central inference -- that these two items, combined ($242B), exceeded the entire $147B year-over-year rise in federal spending through August, implying underlying spending growth is actually slowing and 'might even have shrunk' -- does not hold up. The $117B interest figure and the $147B spending figure are measured over different windows: the interest figure covers the first 10 months of FY2026 (through July), while the spending figure covers 11 months (through August). In the properly time-matched 10-month window, total federal spending actually rose $308 billion year-over-year -- more than double the $147B figure used, and not exceeded by the $242B 'combined factors' at all. Separately, the $125.2B tariff refund is a revenue-side item (netted against customs receipts under standard Treasury accounting), not spending, so adding it to interest costs and calling the sum increased 'spending' is a category error. And the $147B figure itself was artificially depressed by a one-time calendar quirk -- Aug 1, 2026 fell on a Saturday, shifting Social Security/Medicare payments into July -- which Treasury itself flagged in the same week's reporting. Rated 1 Pinocchio for fidelity (sources quoted accurately, but the window mismatch is never disclosed) and 3 Pinocchios for claims accuracy (the headline inference runs opposite to what the article's own cited sources show once time-matched).

Claim-by-claim verdicts

2 true · 1 mostly true · 1 false · 4 assertions checked

Verdicts: true, mostly true, partially true, mostly false, false, not checkable here (no table carries the measure), stale vintage (the article cites a release this corpus has not loaded — a freshness gap, not a falsehood).

Claim detail

SELECT record_date, classification_desc, current_fytd_rcpt_outly_amt FROM econ.mts_outlays_by_function WHERE record_date='2026-08-31' AND line_code_nbr IN ('120','340')

SELECT record_date, current_fytd_rcpt_outly_amt, prior_fytd_rcpt_outly_amt FROM econ.mts_outlays_by_function WHERE classification_desc='Customs Duties' AND record_date='2026-08-31'

SELECT record_date, current_fytd_rcpt_outly_amt, prior_fytd_rcpt_outly_amt FROM econ.mts_outlays_by_function WHERE classification_desc='Net Interest' AND record_date='2026-07-31'

SELECT record_date, classification_desc, current_fytd_rcpt_outly_amt, prior_fytd_rcpt_outly_amt FROM econ.mts_outlays_by_function WHERE line_code_nbr='340' AND record_date IN ('2026-07-31','2026-08-31')

What AskAmerica's own warehouse data shows

Querying econ.mts_outlays_by_function (Treasury's Monthly Treasury Statement, Table 9) directly:

The $309.1B spending increase through July matches Fox Business/CBO's own reported '$308 billion' figure for that same 10-month window almost exactly -- confirming both that AskAmerica's MTS data lines up with CBO's independently, and that the $117B interest figure Eyermann quotes was drawn from that same 10-month comparison. The $147B figure he actually uses in his 'combined vs. total' comparison is the later, through-August figure -- a different window entirely.

The category error: tariff refunds are revenue, not spending

Reuters, citing Treasury (Sept 11, 2026): the Treasury 'collected $292.5 billion in customs duties, even as it has had to refund about $125.2 billion, resulting in a net inflow of $167.3 billion.' This is booked on the receipts side of the federal ledger -- confirmed in AskAmerica's MTS table, where Customs Duties (line_code_nbr 100) falls within the receipts line-code range (10-120), distinct from the outlays range (130-340) where Net Interest (line 320) and Total outlays (line 340) live. A tariff refund reduces how much revenue the government collects; it does not increase how much the government spends. Eyermann's framing -- 'those two things increased the government's spending by $242 billion' -- folds a revenue-side item into a spending-side total, then compares that blended figure against a spending-only baseline ($147B). That comparison mixes categories that don't belong on the same side of the ledger.

A confound the article doesn't mention: the August payment-calendar shift

Reuters (Sept 11, 2026), also citing Treasury: 'For August alone the budget gap was down 52% from a year earlier, but that was largely due to the fact that August 1, 2026 was a Saturday, meaning that Medicare and social security payments normally paid on the first of the month were made instead in July. Adjusting for those items... the month's deficit was $248 billion, up $7 billion from a year earlier.' This is a well-known federal-accounting quirk: benefit payments due on a weekend or holiday get paid on the preceding business day, which can push a large chunk of one month's outlays into the previous month and mechanically distort year-over-year comparisons for both months. This is very likely a major reason the YoY spending increase appears to shrink from $308B (through July) to $147B (through August) -- a timing artifact, not evidence of decelerating underlying spending growth. The article presents the $147B figure without this caveat.

Bottom line

The individual facts in this piece are accurate and properly sourced. The argument built on top of them -- that spending growth, once interest and tariff refunds are stripped out, is slowing or might have shrunk -- is not supported by a correct reading of the same sources the article cites, once time windows are matched and the tariff refund is put on the correct side of the ledger. Readers should treat the deficit, revenue, debt, and interest figures as reliable, and treat the 'underlying spending is slowing' conclusion as unsupported.

Every query behind this report

4 warehouse calls ran in this session, in order. Each is reproducible against the same snapshot.

SELECT record_date, record_fiscal_year, record_fiscal_quarter, classification_desc, line_code_nbr, current_fytd_rcpt_outly_amt, prior_fytd_rcpt_outly_amt, current_month_rcpt_outly_amt
FROM econ.mts_outlays_by_function
WHERE line_code_nbr IN ('320','340') OR classification_desc ILIKE '%Total%' OR classification_desc ILIKE '%Net Interest%'
ORDER BY record_date DESC LIMIT 30

SELECT DISTINCT classification_desc, line_code_nbr FROM econ.mts_outlays_by_function ORDER BY line_code_nbr

SELECT record_date, record_fiscal_year, classification_desc, current_fytd_rcpt_outly_amt, prior_fytd_rcpt_outly_amt
FROM econ.mts_outlays_by_function
WHERE classification_desc = 'Customs Duties' AND record_date >= '2026-06-30'
ORDER BY record_date

SELECT record_date, tot_pub_debt_out_amt FROM econ.federal_debt WHERE record_date IN (
  (SELECT MAX(record_date) FROM econ.federal_debt)
) OR record_date BETWEEN '2016-09-01' AND '2016-09-30'
ORDER BY record_date