11 Points Closed: Streaming Video CPI Versus Cable and Satellite TV Since 2019

Households did not swap one bill for another overnight. They stacked apps on top of broadband while traditional multichannel TV held a large and rising price path. The Bureau of Labor Statistics (BLS) does not publish a single “Netflix CPI,” but it does publish a usable pair: cable, satellite, and live streaming television service (CUUR0000SERA02) on one side, and purchase, subscription, and rental of video (CUUR0000SERA04) — the closest published proxy for third-party video subscriptions and rentals — on the other. Rebased to 2019 calendar-year averages = 100, July 2026 prints cable at 127.6 (+27.6%) and video purchase/subscription/rental at 114.2 (+14.2%). The stream−cable gap is −13.4 percentage points. That gap bottomed at −24.4 pp in March 2023 and has since closed 11 points — a catch-up story, not a completed convergence.

The dashboard above is built for that race: dual indexes, gap area, year-over-year pace, annual ladder, quarterly scatter, and a latest-level bundle bar that also brings in internet services and all-items CPI-U.

What each BLS series actually prices

Cable, satellite, and live streaming television service covers residential TV subscriptions: basic and expanded cable, satellite, live TV delivered on internet-connected devices when it is part of a TV-provider product, premium movie channels, and related equipment and fees. BLS’s residential telecommunications factsheet is explicit about what is out: standalone, third-party pre-recorded video-on-demand streaming services (the familiar app stacks) are excluded from that cable index.

Purchase, subscription, and rental of video (SERA04) is the recreation-group series that picks up purchases, subscriptions, and rentals of video content outside that cable/satellite live-TV frame. It is imperfect as a pure “streaming-only” meter — physical and transactional rental still exist in the item concept historically — but it is the published index desks can rebase and compare without inventing a private price panel. A sibling entry-level series, subscription and rental of video and video games (SS62055), runs hotter (+46.1% since 2019 as of July 2026) because games ride along; this brief treats SERA04 as the cleaner video-content comparator and shows SS62055 as a sensitivity check in the bundle panel.

Internet services (CUUR0000SEEE03) matter because streaming without broadband is a theoretical product. Internet CPI is up 14.3% since 2019 on the same rebase — almost identical to SERA04’s 14.2% — while all-items CPI-U is up 30.6%. Cable TV service has roughly matched the all-items climb; the video subscription/rental proxy has not.

The 2019–2023 divergence

Through the pandemic and the early cord-cutting years, the two video paths diverged hard. Cable’s 2019-average-based index climbed steadily: 104.3 in 2020, 108.6 in 2021, 112.9 in 2022, 118.8 in 2023. Video purchase/subscription/rental fell in the first three of those years on the same scale — annual averages of 93.0, 89.2, and 91.9 — before a modest 2023 rebound to 96.4. Competitive promotions, expanding library competition, and a still-maturing subscription stack kept measured video-content prices soft even as traditional TV packages kept ratcheting.

The gap (stream index − cable index) opened from near zero in 2019 to an annual average of −22.4 pp in 2023. The monthly trough was −24.4 pp in March 2023. That is the arithmetic behind “streaming still looked cheap relative to cable on BLS meters” even as households complained about stacking too many apps: the cable meter rose; the video subscription/rental meter lagged.

Calendar-year averages on the 2019 = 100 scale:

The catch-up since the 2023 trough

From the March 2023 trough, the story flips from “streaming soft” to “streaming accelerating.” Video purchase/subscription/rental crossed back above its 2019 average in 2024 (annual 101.9) and kept climbing: 106.3 in 2025, 113.2 for the January–July 2026 average, 114.2 in the July 2026 print. Cable kept rising too — to 127.6 in July — but more slowly on a year-over-year basis.

July 2026 year-over-year rates make the catch-up visible without a long chart: video sub/rental +7.8%, cable +2.8%, internet +3.5%, all-items +3.4%. Peak cable YoY in this window was 6.6% (September 2023). Peak video sub/rental YoY was 10.0% (December 2025). The gap closed from −24.4 pp to −13.4 pp — 11 points of repair — while remaining clearly negative. Streaming has not “caught” cable on the cumulative 2019 base; it has stopped falling further behind at the same pace.

Internet is the silent co-bill

Any household that cancelled a $80–$120 multichannel package and kept three $12–$18 apps still pays a broadband bill. On the CPI meters, internet services have behaved more like the video subscription/rental proxy than like cable TV: +14.3% since 2019 versus cable’s +27.6%. That does not mean broadband is cheap in dollars — CPI indexes are not average ticket prices — but it does mean measured internet price change has not been the explosive leg of the entertainment stack since 2019. The expensive leg, on BLS’s published paths, remains traditional multichannel TV service.

When the dashboard overlay adds internet and all-items, the visual is consistent: cable tracks near the all-items climb; video sub/rental and internet form a lower pair; the gap ribbon between cable and video is the briefing object.

Why “closing the gap” is not the same as “streaming got cheap”

Three distinct claims get tangled in consumer commentary. First, levels: many households still pay less per month for a few apps than for a full cable tier — that is a bill-composition fact, not what these indexes measure. Second, cumulative price change since 2019: cable’s index rose nearly twice as much as the video subscription/rental proxy (27.6% vs 14.2%). Third, recent acceleration: since the 2023 trough, video subscription/rental inflation has outrun cable on a YoY basis, closing 11 pp of the gap without erasing it.

If a desk only quotes the third claim (“streaming prices are soaring”), it misses that cable is still higher on the 2019 = 100 ladder. If it only quotes the second (“streaming is still the bargain path”), it misses the post-2023 catch-up that is already in the monthly prints. The honest one-liner for July 2026 is: streaming’s published proxy remains ~13 pp behind cable on a 2019 base, but that deficit is 11 pp narrower than the March 2023 trough.

Caveats that should travel with every gap print

Several caveats belong next to the −13.4 pp headline. First, SERA04 is not a pure SVOD index — it is purchase, subscription, and rental of video; transactional and physical residue can still move the series. Second, cable’s item includes live-streaming TV products sold by TV providers, so “cable versus streaming” is not a clean brick-versus-app split; it is multichannel/live-TV service versus the broader video purchase/subscription/rental item. Third, BLS excludes standalone third-party SVOD from the cable index by design — which is why SERA04 is needed as the comparator, imperfect as it is. Fourth, indexes are not average household bills; mix shifts, promotions, and quality adjustments (including telecommunications hedonics BLS documents for bundled plans) can move indexes without matching any one provider’s list price. Fifth, October 2025 is missing from the published extract used here and is dropped from monthly panels; annual 2025 averages use the eleven available months. Sixth, SS62055 (video and video games) runs much hotter and should not be substituted silently for SERA04 when the claim is about video content alone.

None of those caveats restore a zero gap or erase the trough-to-latest closure. On the published NSA CPI-U series, rebased to 2019, video purchase/subscription/rental still sits below cable — and closer than it was in early 2023.

What the comparison is for

The stream-versus-cable CPI gap is a briefing tool for media, broadband, and household-budget desks — not a verdict on which product households “should” buy. It answers a narrow question: since 2019, how far has the published video subscription/rental price path closed on cable and satellite television service? As of July 2026, cable is +27.6%, video sub/rental is +14.2%, the gap is −13.4 pp, and 11 points of the −24.4 pp trough have been closed. Use the dashboard’s window and overlay toggles to hold the 2020–23 divergence, the post-2023 catch-up, or the full 2019–2026 race — and keep SERA04, SS62055, and the cable live-streaming item labeled as the distinct BLS concepts they are.