What is the average cost per revenue passenger mile for airlines in 2027?

What cost per revenue passenger mile actually measures

A revenue passenger mile (RPM) is one paying passenger carried one mile. An available seat mile (ASM) is one seat flown one mile, occupied or not. Load factor is RPM divided by ASM. Cost per revenue passenger mile is total operating expense divided by RPMs — the all-in cost of moving one paying body one mile.

The critical thing to understand up front is that essentially nobody reports this number directly. U.S. carriers report CASM — cost per available seat mile — in their 10-Ks, earnings decks, and DOT Form 41 filings. The Bureau of Transportation Statistics publishes the inputs. Airlines for America publishes cost indexes. IATA publishes global unit costs, usually per available seat kilometer. But cost per RPM is a derived figure you build yourself from CASM and load factor, and the arithmetic is simple:

Cost per RPM = CASM ÷ load factor

If a carrier flies at a 14-cent CASM and an 85% load factor, its cost per revenue passenger mile is 14 ÷ 0.85 = 16.5 cents. The empty 15% of seats do not vanish from the cost base; their cost gets redistributed onto the passengers who actually paid. That redistribution is the entire reason the metric exists and the entire reason it is always higher than CASM.

Why bother deriving it at all, when CASM is the reported convention? Because CASM answers a supply-side question — how efficiently do we operate the airplane — while cost per RPM answers a demand-side question: what does it cost us to serve the customer we actually got. The revenue-side twin is yield, which is passenger revenue divided by RPMs. Yield and cost per RPM are denominated identically, so subtracting one from the other gives a clean per-passenger-mile margin. You cannot do that with CASM without first converting revenue into RASM. For anyone modeling profitability per traveler, per route, or per fare class, cost per RPM is the cleaner comparison because both sides of the equation share a denominator that represents real, paying demand.

The metric is also unusually sensitive to load factor in a way CASM is not, and that sensitivity is the point. A carrier that improves CASM by 2% but loses three points of load factor has made its cost per passenger mile worse, not better. Airlines chasing unit-cost headlines through capacity growth routinely produce exactly this outcome: more ASMs, lower CASM, thinner loads, higher true cost per customer served. Tracking both numbers side by side catches that immediately.

One more framing note. The metric is an average across an enormously heterogeneous production system. A single carrier flies 400-mile shuttle segments and 7,000-mile widebody rotations in the same quarter, and the per-mile economics of those two products differ by a factor of two or more. Any industry-wide "average cost per revenue passenger mile" is a weighted blend of business models, aircraft gauges, stage lengths, and labor contracts. It is useful as a benchmark and a sanity check. It is close to useless as a target.

Building the 2027 number step by step

Because 2027 has not happened, there is no actual to look up. What exists is a defensible construction: take a recent verified base year, decompose the cost stack, escalate each component with its own driver, then reapply a forecast load factor. Here is the sequence practitioners actually run.

Step one — fix your scope before you touch a number. Decide whether you want mainline only, mainline plus regional capacity-purchase flying, domestic only, or system including international. These four scopes can produce answers 30% apart for the same carrier in the same year. Regional flying in particular is expensive per mile — short stages, 50–76 seat gauge, high cost per seat — so including it pushes a network carrier's blended cost per RPM up by roughly a cent to a cent and a half. Write the scope down at the top of the model and never silently change it.

Step two — pull the base-year operating expense. DOT Form 41 Schedule P-1.2 and P-5.2 carry operating expense by carrier and category. SEC filings carry the same figures on a GAAP basis with reconciliation to whatever "adjusted" measure the carrier prefers. Use one source consistently; mixing GAAP totals from a 10-K with Form 41 traffic data works, but mixing adjusted-excluding-special-items expense with GAAP expense across carriers does not.

Step three — pull traffic. T-100 segment and market data gives RPMs and ASMs by carrier, month, and region. Confirm the units are miles, not kilometers, before doing anything else.

Step four — compute the base CASM and base cost per RPM. OpEx ÷ ASM, then divide by load factor. This is your anchor. Everything downstream is an adjustment to it.

Step five — decompose the cost stack. For a typical U.S. network carrier the split runs roughly: labor 32–38% of operating expense, fuel 18–28% depending on crude prices, aircraft ownership and rent 6–10%, maintenance 8–11%, landing fees and other rents 5–8%, distribution and selling 4–6%, and everything else in the remainder. Ultra-low-cost carriers skew differently: fuel is a larger share because there is less labor and overhead to dilute it, and distribution costs are lower because direct booking dominates.

Step six — escalate each line with the right driver. Fuel escalates with the jet fuel price curve, not with CPI. EIA publishes kerosene-type jet fuel spot prices and forward-looking energy outlooks; a forward curve or a scenario band is the honest input. Labor escalates with the contractual rate steps already signed — pilot and flight attendant agreements are multi-year documents with published percentage increases, so the 2027 labor line is substantially knowable in advance rather than a guess. Ownership costs escalate with delivery schedules and lease rates. Airport costs escalate with the rate-setting cycles of the carrier's hub authorities. Maintenance escalates with fleet age and shop-visit timing, which is lumpy and does not respect calendar years.

Step seven — rebuild CASM and apply a forecast load factor. U.S. industry load factor has settled into the low-to-mid 80s in recent years, and the structural pressure is upward: revenue management systems are better, capacity discipline is stronger than it was pre-2010, and seat density has increased. Absent a demand shock, an 83–86% assumption for 2027 is reasonable. Model 82% and 87% as bookends.

Step eight — publish a band, not a point. A ±$0.50 per gallon swing in jet fuel moves fuel CASM by roughly 0.7–0.8 cents on a fleet averaging 60–70 ASMs per gallon, which moves cost per RPM by nearly a full cent after the load-factor divide. Any single-number 2027 forecast that does not carry a fuel sensitivity is presenting false precision.

Typical ranges, cost drivers, and what 2027 likely looks like

Here is how the archetypes stack up, expressed as CASM and then converted at an 84–85% load factor.

Ultra-low-cost carriers. Single fleet type, high seat density, minimal overhead, direct distribution, heavy aircraft utilization. Reported CASM in the high single digits — call it 7.5–9.5 cents. At 83–85% load, that becomes roughly 9–11.5 cents per revenue passenger mile. Their entire business model is engineered around this number: more seats per airframe drops the denominator's cost per seat, and longer daily utilization spreads fixed ownership across more miles.

Low-cost and hybrid carriers. Broader networks, some premium seating, more airport infrastructure, richer labor contracts than pure ULCCs. CASM in the 10–13 cent range, producing roughly 12–15.5 cents per revenue passenger mile.

Network legacy carriers, mainline only. Mixed fleets, widebody international flying that pulls average stage length up and unit cost down, but far more overhead, hub complexity, and senior labor. Mainline CASM commonly reported in the 11–14 cent band, giving roughly 13–16.5 cents per RPM.

Network legacy carriers, consolidated system including regional flying. This is the figure most people mean when they say "a major airline's cost." Capacity purchase expense for regional partners lands as a large single line item, and those regional ASMs are expensive per mile. Consolidated CASM commonly 14–17 cents, giving roughly 16.5–20 cents per revenue passenger mile.

Weight those archetypes by their share of industry RPMs — network carriers still carry the majority of domestic traffic — and a defensible 2027 U.S. industry-wide average sits in the 15–19 cent range, with the mainline-heavy center of gravity at 17–20 cents. Globally, IATA's unit-cost reporting is denominated per available seat kilometer, and converting to a per-revenue-passenger-mile basis requires both a load-factor divide and a 1.609 kilometer-to-mile multiply; skipping either step is the single most common arithmetic error in this analysis.

The stage-length effect deserves its own paragraph because it dominates cross-carrier comparison. Every flight incurs a fixed block of cost at each end: pushback, taxi, takeoff thrust, approach, landing fee, gate turn, crew report time. Spread that fixed block over 400 miles and the per-mile cost is high. Spread it over 3,000 miles and it nearly disappears into the noise. This is why a regional operator flying 350-mile segments can post a cost per RPM two to three times a widebody international operator's, without being remotely less efficient. The standard correction is to normalize CASM to a common stage length — multiply by the square root of the ratio of the benchmark stage length to the carrier's actual stage length is a rough industry convention — before comparing anything.

Fuel is the volatile line and the one that actually moves 2027. At roughly 60–70 ASMs per gallon for a modern narrowbody-heavy fleet, fuel CASM works out to price-per-gallon divided by ASMs-per-gallon. At $2.50/gallon and 65 ASM/gal, that is 3.8 cents per ASM, or about 4.5 cents per RPM at 85% load. At $3.50/gallon, it is 5.4 cents per ASM and about 6.3 cents per RPM. That single input swings the answer by nearly two cents. Fleet renewal cuts the other way: each generation of narrowbody has delivered meaningful fuel-burn improvement per seat, so a carrier taking delivery of new-generation aircraft through 2027 is structurally lowering the fuel line even at flat prices.

Labor is the sticky line. The multi-year contracts signed across the U.S. industry carry scheduled annual step increases, which means 2027 labor CASM is more forecastable than fuel but only moves in one direction. Productivity offsets — higher utilization, better crew scheduling, larger gauge aircraft — are the only lever that keeps labor cost per mile from rising in lockstep with the rate steps.

Gauge and density are the quiet lever. Upgauging from a 150-seat to a 180-seat aircraft on the same route spreads nearly the same trip cost across 20% more seats. The trip cost rises somewhat with weight and fuel burn, but nowhere near proportionally. This is why the average seats per departure across the U.S. industry has climbed steadily and why it is the most reliable structural downward force on cost per available seat mile — and, provided load factor holds, on cost per revenue passenger mile too.

Where teams get this metric wrong

Comparing CASM to cost per RPM as if they were the same number. They differ by the load factor divide, which at 85% is a 17.6% gap. Analysts who pull one carrier's reported CASM and another's derived cost per RPM produce a comparison that is wrong by roughly a fifth before any real analysis begins. Always confirm which denominator each figure uses.

Mixing miles and kilometers. U.S. DOT reporting uses RPM and ASM. IATA and most non-U.S. carriers use RPK and ASK. A cost per ASK is 62% of the same cost per ASM. Blending the two in one table produces a spread that looks like a stunning efficiency finding and is actually a unit error.

Ignoring stage length. Covered above, and it is the number one cause of nonsense benchmarking. A short-haul carrier is not inefficient because its cost per RPM is high; it is flying a fundamentally different product. Normalize or do not compare.

Using "CASM ex-fuel" without saying so. Airlines report CASM excluding fuel, and often excluding special items, because it isolates the part of the cost base management controls. It is a legitimate and useful measure. It is also 4–6 cents lower than all-in CASM. Half the confusion in published airline cost comparisons traces to one side quoting ex-fuel and the other quoting all-in.

Contaminating the denominator with non-passenger revenue. Total operating expense includes the cost of carrying cargo, running the loyalty program, and operating maintenance-for-hire lines of business. Dividing total operating expense by passenger RPMs charges all of that to passengers. For carriers with meaningful cargo or co-brand credit card economics, this overstates true passenger cost per mile. The cleaner approach is to allocate cost to the passenger segment or, at minimum, to disclose that you did not.

Averaging quarterly figures without weighting. Airline costs and traffic are seasonal. Q1 has weak loads and high per-passenger cost; Q3 has peak loads and low per-passenger cost. A simple mean of four quarterly cost-per-RPM figures is not the annual figure. Sum annual expense, sum annual RPMs, divide once.

Treating an industry average as a carrier target. The average is a weighted blend of business models that are deliberately built to different cost points. A ULCC hitting the industry average would be catastrophically uncompetitive; a network carrier hitting the ULCC number would have to dismantle the hub product its premium revenue depends on. The correct comparison set is the carrier's actual competitive peer group on comparable stage lengths.

Forgetting that low cost per RPM is not the goal. The goal is a positive spread between yield and cost per revenue passenger mile. A network carrier with 19-cent costs and 23-cent yield outperforms a low-cost carrier with 11-cent costs and 12-cent yield. Optimizing the cost side in isolation is how airlines cut the product features that were generating the revenue premium in the first place.

Building a 2027 forecast off a single anomalous base year. Any base year distorted by an unusual fuel spike, an operational meltdown, a fleet grounding, or one-time labor settlement charges will propagate that distortion through the entire forecast. Use a normalized base or a two-to-three year average, and disclose the adjustment.

Choosing the right unit metric for the question you are asking

Cost per revenue passenger mile is one tool in a small toolkit, and picking the wrong one produces confidently wrong decisions. The framework below routes the question to the right measure.

If the question is about pricing or profitability per traveler, use cost per RPM against yield. Both are denominated in revenue passenger miles, so the subtraction is meaningful. This is the right lens for questions like "are we profitable on this fare class" or "what does a one-cent yield decline cost us."

If the question is about operating efficiency or benchmarking against a peer, use CASM — ideally ex-fuel and stage-length adjusted. Fuel price is a market input neither management team controls, so including it measures the oil market rather than the airline. Stage-length normalization removes the network-shape distortion. This is the right lens for "are we running a tighter operation than our competitor."

If the question is a route or fleet decision, abandon per-mile metrics entirely and use trip cost. The decision to fly a specific segment with a specific aircraft turns on the total dollars that flight consumes versus the total revenue it generates — fuel burn for that stage, crew hours, landing fee, gate cost, marginal maintenance — against forecast segment revenue. Per-mile averages have already blended away exactly the specificity the decision requires. This is where per-mile thinking does the most damage: routes get killed because their cost per RPM is above the system average, when they were contributing positive margin above their true marginal cost.

If the question is about capacity strategy, watch cost per RPM and CASM together as a pair. Diverging trends are the signal. CASM falling while cost per RPM rises means capacity is outrunning demand — the airline is buying a unit-cost headline with empty seats. Both falling together means genuine productivity gains. Both rising means the cost base is escalating faster than the network can absorb it.

If the question is a 2027 planning figure, commit to a scenario band and a named scope. State the scope — mainline or consolidated, domestic or system — state the fuel assumption, state the load factor assumption, and publish a range. A statement like "17–20 cents per revenue passenger mile for U.S. mainline airlines in 2027, at $2.50–$3.25 jet fuel and 83–86% load factor, consolidated basis" is a usable planning input. A bare point estimate with no assumptions attached is not, no matter how precise it looks.

Related questions

How do you convert CASM into cost per revenue passenger mile?

Divide CASM by load factor. At a 14-cent CASM and an 85% load factor, cost per revenue passenger mile is 16.5 cents. The empty seats' cost redistributes onto paying passengers, so the figure is always higher than CASM.

Is cost per RPM reported anywhere officially?

No. DOT Form 41, SEC filings, and carrier earnings decks report CASM and its components. BTS publishes the RPM and ASM traffic data. You derive cost per revenue passenger mile yourself from those inputs.

Why do ultra-low-cost carriers post such low figures?

High seat density, a single fleet type, heavy daily aircraft utilization, direct distribution, and lean overhead. All five lower cost per seat, and the density lever alone can spread nearly the same trip cost over 15–25% more seats.

What single input moves the 2027 estimate the most?

Jet fuel price. A $0.50 per gallon swing changes fuel CASM by roughly 0.7–0.8 cents on a fleet averaging 60–70 ASMs per gallon, which becomes close to a full cent per revenue passenger mile after the load-factor divide.

Does a lower cost per RPM mean a more profitable airline?

Not by itself. Profitability is the spread between yield and cost per revenue passenger mile. A carrier at 19-cent costs earning 23-cent yield outperforms one at 11-cent costs earning 12-cent yield.

FAQ

What is a revenue passenger mile?

One paying passenger transported one mile. It is the standard demand unit in U.S. airline reporting. Its supply-side counterpart is the available seat mile — one seat flown one mile whether or not anyone bought it. Dividing RPMs by ASMs gives load factor. Non-revenue passengers, such as employees traveling on pass benefits, are excluded from the count, which is why the word "revenue" is in the name.

Why does everyone quote CASM instead of cost per revenue passenger mile?

CASM measures the production side of the business, which is what airline management directly controls. It is stable, comparable across periods regardless of demand swings, and it is what the reporting conventions settled on. Cost per revenue passenger mile is the derived demand-side view and is more useful when you are comparing cost directly against yield, since both share the RPM denominator.

How do I compare a U.S. carrier's number to a European or Asian one?

Convert units first. Most non-U.S. carriers report per available seat kilometer, so multiply by 1.609 to get a per-mile basis before doing anything else. Then handle currency at a consistent exchange rate and, ideally, normalize for stage length, since long-haul-heavy carriers will always show a lower per-mile cost for structural reasons rather than efficiency reasons.

What percentage of airline operating cost is fuel?

It varies with crude prices, typically landing somewhere in the 18–28% range for U.S. carriers in recent years, and it has moved outside that band during price spikes and collapses. Labor is usually the larger line at roughly 32–38%. Because fuel is the volatile component, most carriers also report a CASM excluding fuel so investors can see the controllable cost base.

Can I use the industry average as a benchmark for one airline?

Only loosely. The industry average is a weighted blend of ultra-low-cost, low-cost, network, and regional operators with fundamentally different cost structures and stage lengths. A meaningful benchmark compares a carrier against a peer group flying similar stage lengths with similar aircraft gauge and a similar product. Compare against the average only to establish rough position, never as a target.

How reliable is a 2027 forecast of this metric?

The labor portion is fairly reliable, since multi-year contracts publish their scheduled rate increases in advance. Ownership and airport costs are moderately predictable. Fuel is not predictable, and load factor depends on demand conditions nobody can forecast confidently a year out. Present the result as a range tied to explicit fuel and load-factor scenarios rather than a single point.

Sources

• https://www.transtats.bts.gov/
• https://www.bts.gov/topics/airlines-and-airports
• https://www.airlines.org/dataset/
• https://www.iata.org/en/publications/economics/
• https://www.eia.gov/dnav/pet/pet_pri_spt_s1_d.htm
• https://www.faa.gov/data_research/aviation/aerospace_forecasts
• https://www.sec.gov/edgar/search/
• https://www.bls.gov/ppi/

Related on PULSE

• [What is CASM and how do airlines calculate it?](/knowledge.html)
• [How does load factor change unit economics?](/knowledge.html)
• [What is airline yield and how does it differ from RASM?](/knowledge.html)
• [How do you build a route profitability model?](/knowledge.html)
• [Why does stage length distort unit-cost benchmarking?](/knowledge.html)