On 11 September the national average price of diesel crossed six dollars a gallon for the first time ever. It has kept going. On 1 October it stood at $6.37.
A year ago the same measure was $3.71.
That is a rise of 72 percent. It is not only crude, though Brent near $104 is part of it. The world is short of distillate specifically. That is why the G7 is releasing a hundred million barrels.
Now go to a supermarket. Prices are up, but they are not up 72 percent, or anything like it.
Both of those things are true. Here is how.
Fuel is a smaller slice than it feels
Two numbers do almost all the work, and they multiply.
The first is what freight costs as a share of a retail price. For an ordinary consumer good it runs around a tenth.
The second is what fuel costs as a share of freight. The trucking industry publishes this. Total operating cost hit a record $2.336 a mile. Fuel was 48.2 cents of it. That is about 21 percent, rising toward 30 at today's diesel.
The rest is drivers, trucks, insurance, maintenance, tires. None of it moves when the pump does.
fuel ≈ 25 percent of freight
so fuel ≈ 2.5 cents of every retail dollar
Two and a half cents. That is the thing diesel moves.
A 72 percent rise applied to two and a half cents gives about 1.8 cents. Call it a 1.8 percent rise on an average good, once it all comes through.
That is real. It is also not what people brace for when they see a record at the pump.
The six things standing in the way
And 1.8 percent is the destination, not the speed. Six separate layers sit between a refinery and a sticker, and they stack.
| Layer | What it does | How long |
|---|---|---|
| Crude to pump | Refiners and distributors work through fuel they already bought. | 1 to 3 weeks |
| Pump to surcharge | Fuel surcharges are pegged to the previous week's government diesel index. The carrier eats the gap. | about a week |
| Spot to contract | Spot freight reprices immediately. Most volume moves under contracts that reprice annually. | months |
| Inventory | What is on the shelf today was shipped at last quarter's cost. | weeks to months |
| Package before price | Shoppers punish a price rise 2.6 times harder than the same loss taken as a smaller package. | first response |
| Retail stickiness | Nobody wants to move first with store brands at a record 23.8 percent of units. | quarters |
Add them up and the answer to how long is roughly one to two quarters before this is clearly visible in a price index.
Which means the honest version of the question is not when prices rise. It is when you will be able to see prices rise, and those are different dates.
Who actually pays
Now the part that matters, and it is hiding in that word average.
Freight is not a tenth of every price. It is a tenth of the average price, and the spread around that average is enormous.
| Good | Freight as a share of retail | Why |
|---|---|---|
| Milk | 14 to 21% | Water in a jug. Heavy, cheap, refrigerated. |
| Potatoes | 8 to 13% | Dense, low value, sold by the sack. |
| Chicken | 3 to 13% | Varies with cut and cold chain. |
| Beef | 3 to 7% | Heavy, but each pound carries more value. |
| Flour, tomatoes | 2 to 5% | Light relative to what they sell for. |
The rule underneath is simple. Freight share tracks bulk divided by value.
A pallet of milk and a pallet of phones cost about the same to move. One is a fifth of the price. The other is a rounding error.
That is the finding. Run the same 72 percent through milk instead of the average and the number is not 1.8 percent. It is closer to 3 percent.
Meanwhile the electronics aisle barely moves, because a shipping container of laptops carries its freight cost in the first laptop.
Why this matters to the theory
The Toilet Paper Theory says the quantity of a necessity a household can use is roughly fixed. You use what you use.
Put the two findings together.
Necessities are disproportionately the cheap, heavy, low margin goods. Those are exactly the goods where freight is the largest share of price. And they are the goods a household cannot respond to by buying less, because it already buys the amount it needs.
So a fuel shock arrives through the one door that cannot be closed.
The household facing it has three options and two are already gone. It cannot use less milk. It cannot wait for the price to pass. What is left is trading down, which is why store brands keep setting records. And that is a one way door.
What would prove this wrong
Three things, all checkable in public data.
Say food at home inflation runs more than three points above core over the next two quarters. Then the pass-through is much larger than this arithmetic allows, and the 2.5 cent figure is wrong.
If package sizes hold steady through this shock while shelf prices move first, the fifth layer does not exist. The shrinkage finding was a coincidence of the last cycle.
And if the gap between heavy goods and light goods fails to widen, freight share is not driving the distribution. Something else is. Milk against flour is the cleanest pair to watch.
Two cautions. The 2.5 cents is a chain of averages, each with a wide range. Read it as the shape of the effect, not a measurement. And it covers only the first round. Diesel also moves farming, fertilizer and plastics, and that arrives later, through a different door, and probably larger.