LA can show where its commercial trash is buried. It can’t show where most of what’s set out for recycling or compost ends up.

The stakes

Commercial waste makes up 70% of what LA sends to landfills

RecycLA is the city’s commercial waste franchise. 8 haulers own exclusive contracts to serve 11 zones. We pulled hauler-submitted data from January 2022 through December 2024, the most recent available — 4,828,942 tons in all.

The labels on the bins

What the bin colors are supposed to mean

Every commercial customer sorts into the same three bins. Each color is a promise about where the waste goes:

Black bin is labeled for landfill; green bin for compost; blue bin for recycling.

The ideal path

Where each bin is supposed to go

Before we open the records: the chain of facilities each color is meant to travel, and what happens at every stop. Solid lines are the intended route; dashed lines are the residual that still ends up in a landfill.

Black bin: transfer station or MRF, then landfill or waste-to-energy; at a MRF, metal is picked out and sold to a commodities buyer. Green bin: transfer station or MRF, then a compost facility, an extrusion press feeding anaerobic digestion, or waste-to-energy; residual goes to landfill. Blue bin: transfer station or MRF, then a sorting MRF that bales separated commodities for buyers — cardboard, paper, aluminum, glass, HDPE, PET, and PP, whose US recycling rates run from about 96% down to 3% — with contamination going to landfill.

The gap

The records stop before the waste does

The reports count a ton as “diverted” the moment it reaches a recovery-typed first stop — like a recycling sorting plant (MRF). But the first stop rarely decides where waste ends up; the second hop does, and the reports usually go quiet there. That silence covers 1,114,692 tons — 23.1% of everything collected. So we joined the reports to CalRecycle’s and LA County’s disposal records and graded every ton:

CONFIRMED — the records show it directly.

EST. CEILING — the state logs what a facility throws away, not what it recycles. We take each facility’s reported disposal as the worst case, so “est.” outcomes are the most landfill the records allow — a ceiling, not a measurement.

UNKNOWN — the trail ends and we don’t guess. Counted as neither landfill nor diversion.

Every chart below carries these labels.

The answer

Only 18–24 of every 100 tons can even be counted diverted

The reports track landfill well: 58.3% of the city’s commercial tonnage is reported there. Recovery, far less — just 18.5% even reached a compost, recycling, or sorting facility, and reaching the door is all it takes to count as diverted. Another 23.1% stops at a transfer or processing endpoint with no reported outcome; 0.1% can’t be traced at all UNKNOWN.

At most 18 to 24 of every 100 tons count as diverted (18.48% conservative, 24.05% among classifiable). Confirmed recovery is rarer still — 1.60% reaches a compost facility, and nothing else is confirmed end to end.

A first-destination reading of the submitted reports counts 42 of every 100 tons as diverted. The records do not identify final outcomes for all of that tonnage: even re-typing every contested facility in the most diversion-friendly way reaches only 32 under this audit’s outcome-based definitions.

The city’s own numbers point the same way: in 2024, blue and green bins together were 18.55% of the 1,586,948 tons RecycLA collected (external to the audit). Let’s check on what happened to each bin.

The black bin

Black bin: the one promise the records confirm

Black-bin waste is headed for a landfill, and the records back it up: of 3,960,300 tons, 70.8% is confirmed at landfills, an estimated 20.0% more also ends there, 3.61% burned for energy, 5.20% unknown UNKNOWN.

Disposal records exist to record disposal, so this stream is easy to follow. The recycling and composting claims run through the same records — and that’s where the trail thins.

The green bin

Green: composting and anaerobic digestion mostly can’t be proven

The green stream is thin to begin with: by the city’s own reporting, ~28% of accounts had organics service by late 2023, and ~60% still had none as of 2024–25 (external).

Of 227,976 tons of organics, the trail splits in thirds: 33.8% reached a compost facility CONFIRMED, ARRIVAL; an estimated 33.0% went to landfill EST. CEILING; and 28.5% can’t be followed at all UNKNOWN.

The accounting counts 71.3% of the green bin as diverted. But only 47 of every 100 of those tons can be traced to a compost facility.

And the window hides a cliff: verified compost arrivals nearly stopped in January 2024 — from 53.0% of the stream in 2022 and 43.8% in 2023 to 8.1% in 2024 — while the counted rate never moved. The material now stops at a sorting facility, and no record shows where it goes next.

Even “arrival” isn’t “composted”: 90.4% went to one site a hundred miles north, and compost plants landfill their own residuals and contamination. Another 10,878 tons goes to “the orex,” an organics press whose recovered output and destination the reports never show.

The blue bin

Blue: counted as diverted before anything is recycled

“Diverted” is not “recycled.” A blue-bin load counts as diverted the moment it reaches a sorting facility — and no one verifies recycling past that door.

Apply each matched facility’s own reported disposal share to the 489,226 tons they received, and an estimated 93.6% goes onward to disposal EST. CEILING. End to end under that same reading: 73.4% at landfill, 24.4% that can’t be checked UNKNOWN, and 0.29% that could be real commodities.

Puente Hills MRF

Let’s look at one sorting facility (MRF) that demonstrates the accounting gap

Same building, a different story in every ledger. RecycLA’s reports: 78,646 tons arrive, all counted as diverted. CalRecycle’s records: 97.4% of its reported outbound is disposal EST. CEILING. LA County’s independent count: 92.98% disposal — with just 1.28% leaving as recovered commodities. Only the first ledger calls it diversion.

Where the trail goes cold

Much material can’t be traced at all

941,811 tons stop at transfer stations. At the facilities we could match, 98.7% of the reported disposal stream is landfill-bound EST. CEILING. Another 5,079 tons can’t be tied to any known facility type UNKNOWN. We count neither as diverted nor landfilled.

And the blind spot is widening: tonnage that dead-ends at a transfer or processing endpoint climbed every year — 22.2% (2022), 22.8% (2023), 24.3% (2024). For one hauler it jumped from 5.6% to 78.6%.

It also has a geography: South LA is the least traceable zone in the program — 99.7% of its 539,081 tons dead-ends at a transfer station.

The ledger

Let’s add the three bins back together

All 4,828,942 tons: 58.3% reported to landfill, plus a facility-level disposal scenario of 27.7% more EST. CEILING — as much as 86.0% under that upper-bound reading. Confirmed compost arrivals: 1.60%. Burned for energy: 3.02% — counted as diverted under every definition here. Trail ends: 8.85% UNKNOWN.

Of the three destinations the colors promise, only landfill holds up. Of all green-bin tonnage, we estimate only ~34% makes it to a verified compost or anaerobic-digestion facility. The picture for the blue bin is even worse: only 62.8% can be traced to a material recovery facility, and still less than that to a verified recycling sale.

The reporting gap

Why the records can’t show the full path

The state’s database records what facilities throw away, not what they recycle, and no load is tracked end to end. So disposal shows up; composting and recycling mostly don’t.

And the first stop is often the hauler’s own gate: at least 48.8% of all tonnage first lands at a facility owned by the company that filed the report. When the hauler owns the first stop, an outside, independent check matters more.

These are also the records the city enforces: the city levied $12.2 million in penalties against RecycLA haulers — against a baseline that appears in no released dataset. No one outside the sanitation agency can check the math.

Where the records stop.

What could change this → the reporting changes that would let anyone compute this answer directly from public data.