How much carbon dioxide do major music streaming platforms emit annually per listener?

1. Why the headline numbers disagree: different boundaries, different math

Studies that attempt per-listener or per-stream carbon figures pick different system boundaries: some count only data‑center and network energy, some also include end‑user device electricity, and others expand to Scope 3 emissions such as tours, physical media and merchandise — which can dominate industry totals ^{[4] [3]}. Because streaming growth has greatly lowered plastics from CDs/vinyl but increased cloud storage and transmission, comparing “streaming vs. physical” without clarifying included emissions produces contradictory headlines ^[3].

2. Representative per-hour and per-stream figures reported in the literature

Several pieces of recent reporting and secondary analyses use a working value for audio streaming that is far lower than video: one reconstruction cites an energy or carbon rate around 0.055 kg CO2 per hour (0.055 kg/hour) used to estimate the footprint of high-profile artists’ cumulative streams ^[1]. Conversely, European averages for video streaming are cited at about 55 g CO2e per hour — roughly 50× the audio figure in that reporting — underscoring the importance of stream type (audio vs video) when converting listening time to emissions ^[5].

3. Platform-level totals: Spotify as a case study

Analysts attempting platform totals come to different conclusions depending on inclusion choices. Greenly’s extrapolation places Spotify’s total emissions at about 112,000 tonnes CO2e in 2021 and an estimated 187,040 tonnes CO2e in 2025, noting that Spotify’s own reporting stopped including user-device electricity from 2023 — a decision that materially changes totals ^[2]. Spotify itself reports commitments and plans to reach net zero by dates ranging in company materials, but metrics and scopes vary between company disclosures and independent reconstructions ^{[6] [2]}.

4. Scaling from per-hour to per-listener per-year: why it’s tricky

To produce “per listener per year” you need at minimum: (a) a robust per-hour emissions rate for the actual content type (audio vs video), (b) accurate average listening hours per user annually, and (c) clarity about what emissions categories are included. Some articles provide per-hour proxies (audio ≈0.055 kg/hr reconstructions; video ≈55 g/hr in Carbon Trust/Conversation citations) but do not supply a standardized annual listening figure or a consistent scope, so multiplying one by the other yields widely divergent per‑listener estimates ^{[1] [5]}.

5. Broader context: streaming’s share of digital emissions and live-music emissions

Reporting places the internet and data centers as material contributors to global GHGs; some pieces suggest streaming (audio+video) now accounts for a few percent of global emissions, and separate research finds live concerts and tours are a major source within the music industry’s total footprint — meaning that even if per-listener streaming emissions are modest, the full sectoral impact is larger once touring and events are counted ^{[7] [8] [9]}.

6. Competing viewpoints and limitations in available reporting

Some outlets and companies emphasize improvements from moving to renewable energy for cloud services and company pledges to decarbonize ^{[10] [6]}, while independent analysts warn that user-device electricity and rising consumption levels can offset efficiency gains ^{[11] [2]}. Available sources do not present a single peer‑reviewed, cross‑platform per‑listener annual number; many figures are reconstructions or extrapolations and differ by included scopes ^{[1] [2] [3]}.

7. Practical takeaway for readers and researchers

If you want a defensible per-listener annual figure, pick and state your boundaries: whether you count only data centres and networks, include device use, or add Scope 3 like tours; choose audio vs video rates; and use a transparent average listening hours/year. Current reporting gives plausible audio streaming rates (~0.055 kg CO2/hour in some reconstructions) and platform totals for Spotify (112,000 tCO2e in 2021 → est. 187,040 tCO2e in 2025), but does not deliver a universally accepted per-listener annual CO2 figure across major platforms ^{[1] [2] [3]}.

If you want, I can: (A) show worked examples converting a chosen per-hour rate into per-listener annual CO2 for several average-listening scenarios, using only the numbers in these sources; or (B) compile a short list of what a rigorous study would need to publish to make cross-platform per-listener comparisons meaningful.