Container Recycling Institute - Beverage Container Data Archive

Datadory ships the Container Recycling Institute - Beverage Container Data Archive collection as one maintained package for Metal, Glass & Plastic Containers work. CRI is the nonprofit research organization that has tracked American beverage container behavior longer than anyone else, and its archive is the longest continuous record in the category: containers sold, recycled and wasted annually from 1970 through 2023, split across aluminum cans, PET and HDPE plastic bottles, glass bottles and cartons/pouches.

Two findings in the archive anchor nearly every packaging-sustainability conversation in the US. First, roughly half of America's beverage containers are never captured for recycling - national rates have hovered near 50% or below for decades even as container sales climbed. Second, the deposit-system gap is enormous: CRI's own comparison puts average recycling at about 64% in the ten bottle-bill states versus about 26% everywhere else. Those two numbers, and the fifty-year series behind them, are why this archive gets cited in legislative testimony, ESG reports and packaging strategy decks alike.

The archive also extends beyond the charts into CRI's companion publishing: state-by-state profiles of all ten US deposit programs dating to Oregon's 1971 statute, and the Bottle Bill Resource Guide, which carries the text of every state deposit law alongside program-level redemption counts. Datadory delivers all of it as structured records your team can query.

Get a sample of this dataset - tell us which materials or states you need and we will cut the extract accordingly.

Series records from the collection, flat exactly as they arrive:

# Chart-card record -- the national headline series
title            US beverage bottles and cans sold, recycled and wasted
year_span        1970-2023
materials        all
variables        sold, recycled, wasted
units            billion units

# Chart-card record -- the aluminum loss ledger
title            Aluminum cans wasted in the US since 1970
year_span        1970-2023
materials        aluminum
variables        wasted (landfilled + littered)
units            billion units

# Chart-card record -- the deposit-system split
title            PET plastic bottle recycling rate, deposit vs non-deposit states
year_span        1990-2023
materials        PET
variables        recycling rate
units            %

# Comparison row -- the most-quoted number in US packaging policy
metric                average beverage container recycling rate
deposit_states        64%
non_deposit_states    26%

# Chart-card record -- the glass series
title            Glass bottles recycled and wasted in the US
year_span        1980-2023
materials        glass
variables        recycled, wasted
units            billion units

Read the comparison row twice. A 38-point spread between deposit and non-deposit states is not a rounding difference or a measurement artifact - it is the cleanest natural experiment in American waste policy, and it is why the 64%/26% pair shows up in almost every container-deposit debate. The chart-card records around it supply the trend lines underneath the argument: five decades of aluminum cans lost to landfill, the PET recycling-rate series split by deposit status since 1990, and the glass series running from 1980 forward.

Every row arrives typed and documented - title, year span, materials, variables, units, geographic scope, value - so the series joins cleanly onto production volumes, state populations or your own packaging data.

Eight core fields carry the collection. Five describe each archived series - title, year_span, materials, variables and units - and they matter more than they first appear, because they are what turn a pile of chart cards into a queryable inventory: filter by materials = aluminum, narrow to variables = wasted, and you have the loss-ledger extract without reading a single image. preview preserves the pointer back to the original rendering for provenance checks, while geo_scope distinguishes national totals from deposit aggregates and the ten individual bottle-bill states, and value holds the measured figure itself in whatever unit system the series uses - billion containers, percent, pounds per capita or tons CO2 equivalent.

The complete dictionary follows; series-specific nuances fold into the request step rather than cluttering every pull.

Three chips summarize the footprint:

• Geography: United States national totals as the backbone, with deposit-state versus non-deposit-state breakdowns wherever the underlying series supports the split, and the ten individual bottle-bill states - California, Connecticut, Hawaii, Iowa, Maine, Massachusetts, Michigan, New York, Oregon and Vermont - tracked on companion pages down to program-level redemption counts.
• Temporal: 1970 through 2023 on the headline sold/recycled/wasted series - fifty-plus unbroken years, which is precisely what makes trend claims testable. Some material-specific series begin later: the PET deposit-split rate comparisons run from 1990, the glass series from 1980. Statutory coverage dates to Oregon's 1971 bottle bill.
• Granularity: annual national and state figures by material type and metric. This is a decade-scale series, not a daily feed - its value is longitudinal, and treating it that way is what the request step optimizes for.

Set against the wider Datadory catalog - where the average quality score across all datasets sits at 7.81 - this collection scores 5/10: authoritative on trends and benchmarks, bounded by annual cadence and by figures that originate as rendered charts rather than machine-native tables, which we flag explicitly rather than paper over.

API, files, or your warehouse. Daily, weekly, or hourly.

You choose the channel and the cadence; the field dictionary above travels unchanged across all three. Bulk files suit loading the full 1970-2023 history into a data lake or BI tool in one pass, structured API responses suit dashboards that track benchmark figures against your own recovery performance, and warehouse delivery puts the series one join away from your sales, plant or portfolio tables. When new annual observations land, they arrive through the same channel as everything else - changing cadence is a settings conversation, not a re-integration project.

• Deposit-policy analysis. The 64%-versus-26% benchmark and the state-by-state program profiles quantify exactly what a container-deposit system changes, which makes the collection the factual spine of legislative testimony, ballot-measure campaigns and regulatory comment letters.
• Sustainability and ESG reporting. Brands setting recycled-content or collection targets need the baseline the rest of the industry quotes - national recycling rates near 50%, five decades of per-material trends, and the wasted-container series that turns 'packaging waste' from a slogan into a number.
• Material-recovery investment. Aluminum and PET wasting series sized the feedstock available to reclaimers and MRF operators; a billion cans a year lost to landfill is a business-plan input, not just a statistic.
• Packaging market sizing. Fifty years of containers-sold series by material give analysts a demand curve for glass, aluminum and PET packaging that predates most commercial data sources by decades.
• Competitive benchmarking. Comparing your portfolio's recovery footprint against the deposit-state and non-deposit benchmarks tells you whether a sustainability claim is leadership or merely the regional norm.
• Academic and journalistic citation. The archive's longevity makes it the reference series for anything historical about American container waste, from dissertation baselines to investigative timelines.

Market researchers get a fifty-year by-material demand and recovery record for sizing and trend decks without transcribing numbers out of chart images. Journalists and academics get the citable long-run series behind packaging-waste accountability reporting, delivered on a cadence that keeps pieces current. Investors and quants get recovery-rate benchmarks and wasting volumes to overlay against reclaimer capacity, resin spreads and packaging-company revenue estimates. Data scientists get typed, documented series ready to join onto population, commodity or production data without building a chart-parsing pipeline of their own.

Three things worth knowing upfront. First, this is a family of related series, not one wide table - sold/recycled/wasted counts, rate series, per-capita measures and GHG impact estimates share materials and years but not a common key, so samples arrive cut per series rather than merged. Second, the source presents much of the archive as rendered charts rather than native tables, so where a figure exists only inside an image we flag the row's provenance explicitly - you will always know which values are tabulated and which were recovered from the rendering. Third, the cadence is annual and longitudinal: if you need sub-year resolution you want a transactional source instead, and we will say so during the sample step rather than oversell the fit.

Notes that pair well with this page:

• The metal, glass and plastic containers data hub collects every dataset in the industry, this one included.
• Best-of ranking: metal, glass and plastic containers datasets puts this collection in context against the alternatives.
• Supply-side counterpart: Glass Packaging Institute facts, recycling and shipment data covers the production and shipment side of glass containers while this archive covers the recovery side.
• Metal-can view: Can Manufacturers Institute metal can industry statistics tracks the aluminum and steel can shipments whose afterlife this archive measures.
• European benchmark: Eurostat packaging waste statistics gives the EU recovery series to set the US numbers against.