Guides
How to check US Canada border and trade claims with CBP and Census data
A guide to checking US Canada border and trade claims with Census trade tables and CBP data, covering units, source vintages and four common failure modes.
What to take away
- A US goods trade deficit with Canada is real. A typical annual range is 40 to 80 billion dollars, and it counts goods only.
- CBP counts encounters, not people. Recent northern border totals run in the low hundreds of thousands per fiscal year, against roughly two million at the southwest border.
- A tariff rate is not a cost. Duty collected equals the rate times the customs value of covered imports. Section 232 steel and aluminum duties stand at 50 percent, raised in June 2025.
- NEXUS membership is a stock in the low millions. Crossings are a flow, and CBSA publishes them separately.
The four verdicts at a glance
Verdict
- US goods deficit with Canada of a stated figure
- Needs a qualifier: goods only
- A stated number of migrants crossed the northern border
- Fails: encounters are events
- Tariffs on Canadian steel and aluminum cost a stated sum
- Fails: import value is not duty
- A stated number of travelers use NEXUS
- Needs a qualifier: two units
Source table
- US goods deficit with Canada of a stated figure
- Census FT900, goods by country
- A stated number of migrants crossed the northern border
- CBP nationwide encounters
- Tariffs on Canadian steel and aluminum cost a stated sum
- CBP duty data, Census customs value
- A stated number of travelers use NEXUS
- CBP enrollment figures, CBSA traveller statistics
Vintage
- US goods deficit with Canada of a stated figure
- Latest monthly release, revised
- A stated number of migrants crossed the northern border
- Fiscal year to date
- Tariffs on Canadian steel and aluminum cost a stated sum
- Rate set in June 2025
- A stated number of travelers use NEXUS
- Enrollment annual, crossings monthly
Claim 1: the goods deficit with Canada
The claim is a dollar figure for the bilateral deficit. The check starts with Census FT900, which reports goods trade by country. Counts and values are not interchangeable, as statistical measures compared sets out.
Census splits general imports from imports for consumption. It also separates customs value from free alongside ship value. Monthly figures are preliminary and revised.
Recent annual goods deficits sit in the tens of billions, a typical range of 40 to 80 billion dollars. Name the year and the definition.
Verdict: needs a qualifier. The figure is checkable, but goods only. Add services and the balance narrows or flips.
Claim 2: northern border encounters
The claim says people crossed. CBP encounter tables count events. One person turned back and crossing again appears twice.
Pull the sector table, not the national total, if the claim names New York, Vermont or Washington. Family unit counts can sit in more than one table. A statistics verification checklist keeps the order of checks from slipping.
Verdict: fails as written. The restatement names the unit: CBP recorded N northern border encounters. Whether that is a lot is answered by the full series at Stats and Summaries.
Claim 3: the steel and aluminum tariff cost
The claim gives what tariffs cost Americans each year. Duty collected is the cost. Import value is not. CBP collects duties and publishes trade guidance at Trade.
Work the check in five steps.
- Identify the legal authority and the effective date. Section 232 duties on steel and aluminum were raised to 50 percent in June 2025.
- Identify product scope by tariff classification, not by common name.
- Find the customs value of covered imports in Census or CBP data.
- Multiply by the actual rate, allowing for exclusions and quotas.
- Compare the result with the claim, and state the year.
Checking a tariff cost claim
- Identify legal authority and effective date
- Identify product scope by tariff classification
- Find customs value of affected imports
- Multiply by actual rate, minus exclusions
- Compare result to claim, state year
Verdict: fails. Such claims usually quote total import value, and duty is a fraction of it. A blanket claim covering all Canadian steel and aluminum is wrong at the edges.
Scope runs by classification, with exclusions and origin rules. Mislabeled figures follow a familiar pattern, catalogued in common statistics and chart reporting problems.
Claim 4: NEXUS enrollment against crossings
The claim gives a number of travelers who use NEXUS. Program rules for pre-approved low-risk travelers are at NEXUS.
Enrollment is a stock in the low millions. Crossings are a flow, and they change with the season and the port.
Verdict: needs a qualifier. Say enrollment when you mean enrollment, and name the port and mode when you mean crossings.
Facial comparison runs at some ports and modes, so a claim that the border is fully biometric is false.
Matching US and Canadian releases
Statistics Canada publishes merchandise trade by country and commodity. CBSA publishes traveller and enforcement statistics. Match concepts, not identical numbers.
US import data and Canadian export data describe one flow under different valuation, origin and timing rules. A gap of a few percent is routine.
Exchange rates are a hazard. Canadian releases often report in Canadian dollars. Use the rate the source used, and say which one.
When the two countries diverge sharply, the cause is usually a definition. Agency records and commentary are different kinds of evidence, a distinction set out in evidence source types compared.
Four failure modes
Every claim above fails in one of four ways: the unit, the denominator, the date or the source. The unit error is most common.
Four failure modes
Failure mode
- Unit
- Encounters reported as people
- Denominator
- Seizures as share of crossings
- Date
- Prior-year figure quoted now
- Source
- Social post as evidence
Example from article
- Unit
- Denominator
- Date
- Source
Denominator errors are quieter. A seizure total counts what was found, not what crossed, so a rising trend can mean more smuggling, better detection, or both.
