Business Class Fare Check: 12 Rows in September—Verify, Not Coupon or Winner

Takeaway	Detail
37.5% does not verify the September swing	No supplied source states a May-to-September percentage gap for the requested U.S.-city comparison, so the figure cannot support Riley Quinn's larger-Japan-discount claim.
$1,195 and $1,200 need a like-for-like basis	The extracts do not normalize trip type, airline, aircraft, nonstop versus connecting service, fare brand, taxes, or baggage rules across gateways.
$720 is not tied to a Japan itinerary	Every destination in FlyExpert's visible DCA popular-destination list is European; Japan appears only in broader coverage, without a city-pair quote or nonstop claim.
$1,793 and $1,920 cannot rank MSP routes	FlyExpert lists Tokyo Narita and European routes from MSP, but supplies no consistent, dated checkout total that ranks those itineraries.

The supplied FlyExpert and Skylux Travel extracts quote neither May nor September business-class fares for the requested American-city comparison. The separate 37.5% figure therefore cannot verify Riley Quinn's contrarian claim that Japan's September discount should be larger after holiday weeks and seat-product differences are removed. A valid test must hold trip type, carrier, aircraft, stops, fare brand, taxes, and baggage rules constant.

The route pages sharpen the distinction between coverage and price. Washington Reagan's visible popular-destination list is entirely European, while Japan appears only in broader destination coverage. Minneapolis explicitly lists Tokyo Narita alongside European destinations. Neither supplied page establishes nonstop service or a matched business-class checkout total. Route availability is not a coupon, and a gateway is not automatically a fare winner.

The figures $1,195, $1,200, $1,793, $1,920, and $720 cannot rank the comparison unless their itinerary and fare-basis fields are documented. The supplied evidence does not make those links. Europe may win an individual U.S. gateway and Japan may show the larger percentage swing, but neither conclusion is proven. The defensible result is a verification warning: compare the airline checkout total, not the cheapest headline.

The Fare Engine

September is a test window, not a coupon. The engine records matched all-in fares before rendering a verdict. The thesis is supported only if September 2026 undercuts May 2026 in at least 7 of 12 markets and at least 4 of 6 markets in each region, with Japan’s median percentage discount exceeding Europe’s. A “shoulder season” label cannot manufacture any of those conditions.

Cabin C fixes scope; it does not flatten the rules. IATA code C means business, while W for premium economy and F for first are excluded. A C booking may use different fare-basis records, and an ATPCO-filed record can invoke different advance-purchase, minimum-stay, seasonal-validity, and refund conditions. The engine accepts a price only after confirming that the applicable rules fit the same flexible booking.

A business seat and a cheap business fare are separate inventories. Revenue management can close a discounted fare bucket while continuing to sell higher booking classes within C. Thus, a seat map can show business available after the low fare has vanished. Reprice through the bookable flow and capture the fare construction; never substitute W or F merely to preserve the earlier price.

Freeze the basket before searching. Use one adult, round trips, seven nights, USD pricing, taxes and carrier charges included, and paid inventory only. Search New York, Chicago, and Los Angeles to London, Paris, Tokyo, and Osaka: a 3×2×2 grid of 12 markets, evenly divided between Europe and Japan. For each market, use the same preassigned metro airports, eligible booking channels, and qualification rules for matched Sunday-to-Sunday itineraries in May 2026 and September 2026. Do not improve one observation by silently changing airports or trip type.

Normalize bags and disclose benefits. Price one carry-on and one checked bag in both windows. When a fare excludes the checked bag, add its charge to the all-in total. Record lounge, seat, and change terms as terms—not as invented dollar values—so an unpriced benefit cannot distort the price comparison.

Give routing a fixed precedence. The primary result is the lowest qualifying nonstop in C when one exists; otherwise, it is the lowest protected one-stop itinerary. Retain another routing as a labeled sensitivity check, recording its connection airport and total elapsed time. After matching bags and cabin rules, if both windows are flexible, select the lower all-in fare only when the price advantage is clear; otherwise, select the better available nonstop or protected one-stop itinerary. Never let “September” decide that branch.

Do not launder benchmarks into results. According to Skylux Travel’s August 20, 2026 USA-to-London trend, the lowest observed fare was $1,195 and the typical higher level was approximately $1,920. The supplied extract does not state whether either figure is one-way or round trip, so neither enters this round-trip basket. FlyExpert’s DCA page mentions IAD or Newark connections for long-haul Europe, and its example ranges are not guaranteed; a destination listing is not nonstop inventory. Save the bookable total, fare-basis rules, routing, and timestamp for every market before calculating the regional counts and medians.

Option	Fixed marker	Decision
Business	IATA C	Eligible when cabin and fare rules match.
Premium economy	IATA W	Exclude: wrong cabin.
First	IATA F	Exclude: wrong cabin.
Lowest qualifying nonstop	Primary fare	Wins whenever one exists.
Lowest protected one-stop	Fallback fare	Wins only when no qualifying nonstop exists.
Alternate routing	Sensitivity check	Retain the airport and elapsed time; never replace the primary.
Lower all-in fare	Clear price advantage	Select it only after bags, cabin, and both flexible windows match.

The Evidence Ledger

The ledger’s most important output is an auditable denominator, not merely a low fare. Within one 30-minute UTC retrieval window, I query Google Flights for the lowest qualifying business fare in each of the 12 fixed origin-destination markets formed by New York, Chicago, and Los Angeles to London, Paris, Tokyo, and Osaka, using both the September 2026 and May 2026 comparison periods. That means 24 displayed totals. Each row stores the carrier, flight numbers, airports, stops, duration, and its own UTC timestamp. All fare fields use one unit: an all-in round trip for the seven-night search. “Shoulder season” is not evidence.

I then reproduce all 24 searches in ITA Matrix with the business-cabin filter explicit. The Google Flights row never disappears: if ITA changes the total or product, both records remain side by side with difference codes for cabin, carrier, flight, stops, or amount. A result whose cabin cannot be verified remains visible in the raw file but is excluded from the eligible set; it is never silently replaced with a different itinerary.

For each available market-and-window result, the operating airline’s checkout becomes the confirming layer. I record the final amount due, fare basis when exposed, baggage allowance, change and refund terms, and evidence that a comparable business seat is sellable under the recorded product. The comparator must have matching cabin and included-bag conditions. A search card or seat map without that checkout record is not a confirmed fare, and an unmatched itinerary cannot qualify merely because its displayed number is lower.

Every amount carries a named source and UTC timestamp: Google Flights estimate, ITA Matrix fare, or confirmed carrier-checkout price. These are distinct observations, not interchangeable estimates to average. In a New York–Tokyo September row, for example, a Google display and an ITA fare remain separate even if an airline checkout confirms a third amount for the same itinerary. For evidentiary weight, a matching carrier checkout wins; Google Flights discovers candidates, and ITA challenges them.

Any fare below half the median for its destination and search month is quarantined as a possible error fare. I inspect the itinerary for mixed cabins, the checkout for omitted tax, and the booking state for an expired hold. A quarantined observation can re-enter the eligible set only when the named carrier confirms the same product and total. Zero-result, sold-out, and nonstop-unavailable cases likewise remain explicit raw-ledger statuses, not blank cells. They preserve missing observations and the coverage denominator, preventing an incomplete fare set from masquerading as exhaustive. Once sealed, this evidence—not a seasonal assumption—determines whether the article’s market and regional thesis survives.

Control	Required execution	Ledger disposition
Metasearch capture	12 markets × 2 windows = 24 Google Flights totals	Store itinerary fields and an individual UTC timestamp
Time control	One 30-minute UTC retrieval window	Flag collection drift outside the common window
Matrix check	24 ITA Matrix searches with business filter explicit	Preserve differences; exclude cabin-unverified products
Carrier verification	One operating-airline checkout for each available winner	Confirm matching total, product, terms, bags, and sellable seat
Error-fare screen	Fare below one-half of the destination-month median	Quarantine until the named carrier confirms product and total
Missingness	3 explicit statuses: zero-result, sold-out, and nonstop-unavailable	Retain raw rows; never substitute or report the set as exhaustive

The 12-Row Scorecard

The scorecard’s defensible result is a data-status verdict, not a winner: the supplied record cannot support a fare-gap ranking. The FlyExpert and Skylux Travel material contains no same-city, same-route May-and-September pair from which a percentage gap can be calculated, while the FlyExpert and Upgraded Points set contains no priced literal Europe-to-Japan itinerary. A shoulder-season expectation is not a fare observation.

For this scorecard, an eligible quote is a U.S.-dollar, round-trip business-class total per traveler under identical cabin and baggage rules. Let M be May and S September. Dollar gap equals M − S; percentage gap is (M − S) ÷ M, expressed as a percentage. If S exceeds M, retain the negative sign rather than erasing it. “Observed winner” records only the lower all-in total. It is not an automatic booking choice: the article’s clearance rule and the nonstop-or-protected-one-stop comparison remain separate.

The codes below are deliberate audit states: NS means no matched fare supplied; NC means not calculable; UQ means not quote-verified; NT means no timestamp supplied; and U means unresolved. In the current booking year, both the May 2026 and September 2026 cells for New York–London are NS, so its derived gaps are NC. That explicit absence is more useful than a plausible but unsourced fare.

Fixed market	May 2026 all-in fare	September 2026 all-in fare	Dollar gap	Percentage gap	Observed winner	Stops	Elapsed time	Fare basis	Baggage entitlement	Change or refund	Source timestamp
New York–London	NS	NS	NC	NC	U	UQ	UQ	NS	UQ	UQ	NT
New York–Paris	NS	NS	NC	NC	U	UQ	UQ	NS	UQ	UQ	NT
New York–Tokyo	NS	NS	NC	NC	U	UQ	UQ	NS	UQ	UQ	NT
New York–Osaka	NS	NS	NC	NC	U	UQ	UQ	NS	UQ	UQ	NT
Chicago–London	NS	NS	NC	NC	U	UQ	UQ	NS	UQ	UQ	NT
Chicago–Paris	NS	NS	NC	NC	U	UQ	UQ	NS	UQ	UQ	NT
Chicago–Tokyo	NS	NS	NC	NC	U	UQ	UQ	NS	UQ	UQ	NT
Chicago–Osaka	NS	NS	NC	NC	U	UQ	UQ	NS	UQ	UQ	NT
Los Angeles–London	NS	NS	NC	NC	U	UQ	UQ	NS	UQ	UQ	NT
Los Angeles–Paris	NS	NS	NC	NC	U	UQ	UQ	NS	UQ	UQ	NT
Los Angeles–Tokyo	NS	NS	NC	NC	U	UQ	UQ	NS	UQ	UQ	NT
Los Angeles–Osaka	NS	NS	NC	NC	U	UQ	UQ	NS	UQ	UQ	NT
FARE-GAP WINNER: UNRESOLVED—the supplied record names neither window as the observed leader. A September label must satisfy both the required market count and the lower-median all-in-fare condition; the lowest quote is not itself a booking instruction.

The regional audit is symmetrical by design. Europe’s and Japan’s median percentage gaps are both NC, so their direct comparison cannot yet support the prediction that Japan will show the larger expected September swing. Placing each median beside its minimum dollar gap, maximum dollar gap, and won-market count prevents one deep-discount itinerary from carrying a region—or the overall thesis. The fixed market set must remain unchanged when live observations arrive.

Scope	Median all-in fare	Median percentage gap	Minimum dollar gap	Maximum dollar gap	Markets won	Decision use
Europe: six fixed markets	Not calculable	Not calculable	Not calculable	Not calculable	Not measurable	Direct baseline
Japan: six fixed markets	Not calculable	Not calculable	Not calculable	Not calculable	Not measurable	Compare directly with Europe; the larger expected swing remains untested
All fixed markets	Not calculable	Not calculable	Not calculable	Not calculable	Not measurable	No auditable overall leader

The Counter-Evidence

The September thesis is calendar-sensitive, not a shoulder-season reflex. A lower card can be an artifact of moving a holiday, changing the product, or observing stale inventory; it earns a win only when those variables are held fixed.

The supplied ledger does not document Golden Week dates for the comparison year. I therefore treat any Japan itinerary that may touch Golden Week as a separate counterfactual, preserving the same city pair, bags, cabin, and nonstop or protected-one-stop standard. Domestic leisure demand can consume the inventory cushion and erase the normal late-May edge. Because no matched holiday-week fare is supplied, that control remains unresolved—not evidence for either window.

The supplied ledger does not document Memorial Day or Labor Day dates. Their placement relative to the comparison periods is therefore unresolved. When flexibility expands a trip to contain either holiday, that stay is a holiday case, not an ordinary shopping period, and belongs in a separate sensitivity row.

The supplied ledger does not document the dates of the UK’s Spring Bank Holiday or France’s Ascension Day. The London and Paris observations therefore cannot be assigned the same holiday treatment, and a combined Europe result must retain that destination-level difference.

Price risk is also not product risk. A “business” label may pair a lower fare with different lie-flat seating, lounge access, connection protection, aircraft, or departure time. I normalize those fields before comparing. When both windows remain flexible, the governing rule lets the lower all-in fare control only when its price advantage is clear after matched bags and cabin rules; otherwise, the better available nonstop or protected one-stop itinerary controls. A premium is justified only if that added product or protection survives the all-in comparison.

A single retrieval timestamp adds inventory uncertainty: a metasearch card can lag the airline database, and a fare bucket can close without any corresponding reduction in aircraft capacity. The card is an observed quote, not proof that the same bucket remains open for booking.

The U.S. Department of Transportation provides a hold or penalty-free cancellation right for a qualifying covered ticket bought directly from the airline when its advance-purchase condition is met. That right protects the reservation or refund; it does not guarantee that the captured fare will remain available. The practical close is conservative: mark unresolved calendar or product cases, preserve the direct-airline booking record, and apply the article’s fallback to itinerary quality rather than counting an unsupported September advantage.

Control	Calendar or policy fact	Primary source	Verdict and treatment
Japan holiday counterfactual	Not supplied	Japan Cabinet Office	Keep separate; no directional result without a matched fare.
U.S. holiday controls	Not supplied	U.S. Office of Personnel Management	Calendar placement is unresolved; reclassify any stay containing a relevant holiday.
United Kingdom control	Not supplied	GOV.UK	Holiday treatment cannot be assigned from the supplied ledger.
France control	Not supplied	Service-Public.fr	Do not assign a date effect; split the Europe result by destination.
Direct-airline booking protection	Qualifying advance-purchase and hold conditions	U.S. Department of Transportation	A qualifying booking gets the stated protection, not a fare lock.

Also worth reading What the new Japan departure tax · Fresno vs SFO to Yosemite 2026 · British Airways Avios pricing: 2026

The Worked Case: LAX

LAX cannot honestly be scored for the September thesis from the supplied record—not because September loses, but because the record contains no matched LAX fare. According to FlyExpert, the available Tokyo NRT item is an MSP–NRT route-page listing: it supplies no May fare, September fare, carrier, aircraft, or nonstop guarantee. It cannot be relabeled as LAX–HND or LAX–NRT. A shoulder-season reputation is not evidence for a Los Angeles checkout.

Required result field	May 2026	September 2026
Carrier and flight number	Not supplied	Not supplied
Local departure and arrival times	Not supplied	Not supplied
Stops	Not confirmed	Not confirmed
Aircraft	Not supplied	Not supplied
Seat description and fare basis	Not supplied	Not supplied
Research timestamp	Not recorded	Not recorded

Those blanks cannot be hidden behind a “similar” itinerary. Until the exact carrier-and-flight results exist, this case fails at result capture, before the fare comparison begins.

Checkout ledger	Base fare	Taxes	Carrier surcharges	Mandatory extras	Checked bag	Seat selection	Final amount due
May 2026	Not supplied	Not supplied	Not supplied	Not supplied	Not supplied	Not supplied	Not computable
September 2026	Not supplied	Not supplied	Not supplied	Not supplied	Not supplied	Not supplied	Not computable

“Not supplied” means unknown, not zero. The record also leaves the pricing unit unresolved: it does not establish one-way or round-trip. I will not silently normalize one into the other, because that would contaminate the comparison before the arithmetic begins.

Neither requested output—an exact dollar difference nor an exact percentage—is reportable. For a September win, the test is the May confirmed total minus the September confirmed total, divided by the May total. If May is lower, the absolute gap must be labeled a May saving, not a September discount.

The strongest protected one-stop alternative cannot be identified because no LAX connection result, fare, cabin, or schedule is supplied. Its checkout total, connection penalty, and elapsed-time penalty therefore cannot be placed beside the nonstop option. A lower fare alone would not make the itineraries equivalent: the canonical rule still requires matched bags and cabin rules, and an unclear price advantage does not displace the better protected itinerary.

No observed saving can be assigned to the base fare, taxes, baggage treatment, or weaker change rules because the corresponding amounts and fare-rule text are absent. September would become the cheaper basket if a documented change to its base fare, taxes, surcharges, extras, bag cost, or seat cost brought its final total below May’s. The reverse comparison applies to May. Weaker change terms are a product penalty, not a price reduction.

The publishable next action is exact: retrieve both available LAX nonstop results and the strongest protected one-stop in the same timestamped research window, then close each checkout with identical

Frequently Asked Questions

What must September 2026 achieve for the Japan-discount thesis to be supported?

The thesis is supported only if September 2026 undercuts May 2026 in at least 7 of 12 markets and at least 4 of 6 markets in each region, with Japan’s median percentage discount exceeding Europe’s.

Which 12 markets and comparison periods make up the fare check?

The screen covers fixed markets from New York, Chicago, and Los Angeles to London, Paris, Tokyo, and Osaka, using matched seven-night round trips in May 2026 and September 2026.

Can premium economy or first-class fares be used if they are cheaper?

No, IATA W for premium economy and F for first are excluded, while business IATA C is eligible only when the cabin and fare rules match.

What routing is primary when qualifying nonstop service is available?

The primary result is the lowest qualifying nonstop in IATA C when one exists; otherwise, it is the lowest protected one-stop itinerary, while any alternate routing remains a labeled sensitivity check.

How are carry-on and checked-bag costs handled?

Price one carry-on and one checked bag in both windows, and when a fare excludes the checked bag, add its charge to the all-in total.

Can the reported 37.5% swing verify that Japan has the larger September discount?

No, because no supplied source states a May-to-September percentage gap for the requested U.S.-city comparison, so the separate 37.5% figure cannot support the claim.

Quick answers

Is the September test window a coupon?	The Fare Engine September is a test window, not a coupon.
What threshold must the September business-class check meet?	The thesis is supported only if September 2026 undercuts May 2026 in at least 7 of 12 markets and at least 4 of 6 markets in each region, with Japan’s median percentage discount exceeding Europe’s.
What standardized trip and pricing setup should be used?	Use one adult, round trips, seven nights, USD pricing, taxes and carrier charges included, and paid inventory only.
How should checked-bag charges be handled when a fare excludes the bag?	When a fare excludes the checked bag, add its charge to the all-in total.
Does a listed route or gateway by itself prove a fare winner?	Route availability is not a coupon, and a gateway is not automatically a fare winner.