Skip to content

United States · CPI-U · January 2016–September 2026

Recreational-book CPI lagged the U.S. consumer basket across observed five-year windows

The recreational-book index rose 0.47% between September 2021 and September 2026, against 22.83% for all-items CPI. Its relative index fell 18.21%. The full rolling sequence contains 68 available comparisons, all negative, and one missing endpoint; this is not an affordability or causal finding.

Published
Share
Share this page
EmailCite & reuse this study
Key findings as text
Reusable imagesRecreational-book CPI lagged the U.S. consumer basket across observed five-year windows — 68 of 68: Available five-year comparisons with lower book-index growth. 69 scheduled endpoints; October 2025 unavailable. Overlapping windows are not independent tests.Download sharing image (JPG) ↓

Key findings

68 of 68

Available five-year comparisons with lower book-index growth

69 scheduled endpoints; October 2025 unavailable. Overlapping windows are not independent tests.

−18.21%

Latest five-year relative change

September 2021–September 2026; ratio of growth factors, not difference in percentage changes.

+0.47%

Book CPI change in the latest five-year window

All-items CPI rose 22.83%; difference −22.36 percentage points.

−31.50%

Relative change over the full monthly span

January 2016–September 2026; not annual averages or a completed 2026 year.

A relative-price question, not an affordability claim

The U.S. recreational-book price index increased more slowly than the all-items consumer price index in every computable five-year comparison examined here. Across monthly endpoints from January 2021 through September 2026, 68 comparisons can be calculated from observed endpoints. All 68 show a decline in the book index relative to the overall index. October 2025 is the single unavailable endpoint, not an exception with an estimated result. This consistency describes a sequence of overlapping periods; it does not represent 68 independent statistical tests.

The latest comparison illustrates the distinction between a small increase in one index and a much larger increase in another. Between September 2021 and September 2026, recreational-book CPI rose 0.47%, while all-items CPI rose 22.83%. The book index relative to the all-items index therefore fell 18.21%. Books need not show an absolute index decline for their relative index to decline. Conversely, this result does not establish that a household could afford more books, because neither household income nor its other spending commitments is measured here.

The analysis uses two published Bureau of Labor Statistics series for all urban consumers, the U.S. city average, without seasonal adjustment. Its purpose is deliberately narrow: to compare changes in a classified book-price index with changes in the broad consumer basket. It does not estimate a typical book's dollar price, explain publishers' pricing decisions, or measure a reader's personal cost of living. Keeping those questions separate makes the observed relative-price pattern useful without asking it to support claims outside the data.

What the two official series measure

The recreational-book series is CUUR0000SERG02, with a published reference base of December 1997 equal to 100. The all-items series is CUUR0000SA0, with a reference base of 1982–84 equal to 100. Their identities, adjustment status, geographic coverage and latest periods were checked against the official BLS series catalog. The analysis retrieves levels from the BLS recreation and all-items bulk files, but its public evidence files contain only the selected observations and calculations rather than those entire bulk datasets.

An index's reference base is a scaling convention, not a monetary unit. September 2026's recreational-book value of 99.067 is not a price of $99.067. Nor does comparing that number directly with the all-items value of 336.940 show that the consumer basket is several times as expensive as a book. The two series have different bases and different underlying classifications. Meaningful comparisons use each series' change from its own earlier value; the original scale then cancels out.

The selected series share population, geography, frequency and adjustment status. Both are monthly CPI-U measures and both extend through September 2026 in the retrieved catalog and source data. Annual-average records, identified as M13, are excluded. The analytical panel begins in January 2016 and contains 129 calendar-month positions through September 2026. That period is neither a comparison of annual averages nor a completed 2026 calendar year. It is an explicitly dated monthly series with one missing month.

How a five-year relative change is calculated

For each endpoint, the method looks back exactly 60 calendar months. The book growth factor is the endpoint book index divided by the book index five years earlier. The equivalent factor is calculated for all items. Dividing the book growth factor by the all-items growth factor, subtracting one and multiplying by 100 gives the relative percentage change. This ratio-of-growth-factors calculation is not a ratio of the two percentage growth rates, which would answer a different and sometimes unstable question.

For September 2021 to September 2026, the book calculation uses 99.067 divided by 98.605. The all-items calculation uses 336.940 divided by 274.310. These produce growth factors of approximately 1.004685 and 1.228318. Their ratio is approximately 0.817936, corresponding to a relative change of −18.21%. The result says that the book index's movement was lower than the overall index's movement by that multiplicative measure. It does not say that the book index itself fell 18.21% during those five years.

A second useful measure is the difference between the two cumulative percentage changes. Here, 0.47% minus 22.83% is approximately −22.36 percentage points. That percentage-point gap and the relative percentage decline are both valid, but they are not interchangeable. The evidence tables retain both measures with distinct labels. Calculations use the published levels without rounding intermediate growth factors; figures in the narrative are rounded to two decimal places for readability rather than presented as estimates of greater source precision.

The complete sequence, rather than a selected pair

The first available five-year window within the chosen panel runs from January 2016 to January 2021. Recreational-book CPI fell 1.99%, while all-items CPI rose 10.41%. The resulting relative change was −11.23%. Advancing both endpoints by one calendar month produces the next comparison, and that procedure continues through September 2026. The accompanying window table includes all 69 scheduled endpoints, including the unavailable October 2025 result, so the reader can inspect the sequence rather than just two selected examples.

All 68 computable relative changes are negative, but their sizes differ. The least negative observed result is −8.96% for May 2016 to May 2021. The most negative is −24.26% for April 2021 to April 2026. These are descriptive bounds within this defined set of windows, not claims about all historical five-year periods. Neither extreme was used to select the analysis period, and neither should be interpreted as evidence of a sudden structural turning point without a separate research design.

Moving windows overlap heavily. Adjacent comparisons share almost their entire underlying period, so a persistent relationship between the two series can appear repeatedly. The count of 68 negative results is a transparent summary of the observed sequence, not a sample size for an independent-events probability calculation. It does not justify a p-value, a significance claim or a prediction that the next five-year window will have the same sign. The complete downloadable sequence makes this dependence visible rather than hiding it behind the count.

What the longer monthly span adds

From January 2016 through September 2026, recreational-book CPI fell 2.58%, while all-items CPI rose 42.22%. Comparing their growth factors gives a relative decline of 31.50%. This longer-span result supplies context for the five-year windows: a broadly different cumulative movement in the two indexes can coexist with month-to-month variation and with individual five-year periods in which the book index rises slightly. The finding is about relative index movement, not uninterrupted annual or monthly book-price declines.

For an intuitive common display, both series can be independently rebased to January 2016 equal to 100. That operation divides every observation by its own series' January 2016 value and multiplies by 100. It creates comparable measures of cumulative change, without pretending that the underlying goods or original reference periods are the same. The monthly evidence file includes those rebased values alongside the original published levels, allowing the transformation to be checked directly.

The January-to-September endpoints deserve particular care. Unlike the five-year comparisons, which pair the same calendar month, the full-span calculation starts in January and ends in September. Because the inputs are not seasonally adjusted, that comparison may retain endpoint seasonality. It is still a valid change between two published observations, but it is not a seasonally neutral annual-average estimate. No partial-year 2026 movement is annualized, and no value is projected for October, November or December 2026.

Why October 2025 remains a gap

Both selected series record October 2025 as unavailable, with the BLS footnote identifying the lapse in appropriations. The absence is preserved in the monthly panel and in the five-year window ending that month. It is not replaced with zero, carried forward by this analysis, or filled with an interpolated midpoint. An observed value of zero would mean something entirely different from an unavailable price index, and treating the two alike would corrupt both the rebased series and percentage-change calculations.

The implementation uses calendar identifiers rather than counting rows after missing observations are removed. This matters because a lag of 60 surviving rows would no longer reliably represent 60 calendar months. It could compare the wrong starting month once the missing observation is passed. Keeping the complete calendar panel ensures that every computable window retains the intended five-year spacing, and that the unavailable comparison appears explicitly instead of silently disappearing from the schedule.

A window can contain the missing month internally while still having two observed endpoints. September 2021 and September 2026, for example, are both published, so their endpoint change can be calculated without reconstructing October 2025. Such windows are flagged in the evidence file. That arithmetic does not erase the underlying disruption: the BLS shutdown documentation describes imputation and subsequent index production, and later published levels inherit those procedures. The analysis uses those official levels, not an alternative estimate of what uninterrupted collection would have produced.

Recreational books is a defined category

The official classification is broader than printed novels but narrower than every product or service connected with books. BLS descriptions include individually purchased retail books and named digital or audio formats within the relevant entry-level item. They also identify exclusions, including textbooks, reference sets and subscriptions to eBook services. The aggregate index cannot therefore be treated as a census of every book format, every retail channel or every subscription payment made by readers.

The classification also distinguishes book-club purchases from purchases at other retail outlets. BLS states that the book-club entry-level item no longer entered new samples beginning in 2016. This is material context for a panel starting that year. It does not require inventing a replacement series or discarding the published index, but it does rule out describing the entire period as an unchanged collection of identical products and channels. The series is the continuing official index under its documented sampling arrangements.

No separate paperback, hardback, eBook or audiobook result can be recovered merely by looking at the aggregate. Likewise, the analysis cannot identify prices of newly released books, a particular genre, a bestseller list or books purchased by a particular income group. A reader's purchases may differ substantially from the classification's sampled mix. The carefully supported conclusion concerns the published recreational-book CPI category, not a universal experience shared by every reader or every book buyer.

Continuing series do not mean frozen methods

The catalog provides continuing series with no replacement identifier or base-period splice required inside the analytical window. That continuity makes a consistent extraction possible, but it should not be mistaken for proof that the sampling frame, expenditure weights or collection process never changed. CPI is maintained over time, and the official history documents developments affecting how its observations represent consumer spending. The analysis follows the published series instead of attempting to reverse those developments.

Among the relevant changes, BLS records a geographic sample revision in 2018 and changes to the establishment frame during 2019–2020. Annual spending-weight updates began with January 2023 data. These are reasons to describe the comparator as the evolving all-items CPI basket rather than one permanently fixed shopping list. The index remains suitable for describing its own published movement, while stronger assertions about constant-product price trajectories would require additional data and different methods.

BLS also announced collection reductions in 2025, noting that item-specific or subnational indexes could become more volatile. Combined with the later appropriations lapse, this is an important qualification for recent observations. The study does not quantify how much any particular procedural development changed the book result, because it has no counterfactual series with which to do so. Acknowledging those limitations is different from assigning them a causal explanation for the observed divergence.

What a lower relative index cannot establish

All-items CPI includes recreational books rather than excluding them. The comparison is therefore between one classified component and a broad index that contains that component, not between books and a perfectly separate collection of everything else. Its interpretation remains straightforward as a component-versus-overall comparison. The study does not construct an excluding-books index, and it does not claim that the difference is a clean decomposition of every other category's contribution to inflation.

Affordability requires information not present in these series. A household facing lower relative book-price growth might still have less money available for books if its income or unavoidable expenditure changed adversely. Another household could experience the opposite. The indexes alone cannot rank those household situations or establish improved purchasing power. Terms such as cheaper relative to the CPI basket must refer to the index comparison, not be allowed to drift into a statement about personal financial ease.

The same boundary applies to explanations. Retail competition, changing formats, production costs, promotional practices and consumer choices may be relevant topics, but these two time series do not isolate their effects. There is no treatment group, identified intervention or causal model here. The analysis also says nothing directly about publisher revenues, unit sales, reading frequency or the quality of purchased books. Those outcomes should not be inferred from the direction of a relative price index.

Evidence, precision and responsible reuse

The public evidence package contains the full selected monthly panel, the complete scheduled five-year window sequence, and the successive-window summary. The monthly file has 129 calendar positions, including the missing October 2025 values. The rolling file has 69 scheduled endpoints and 68 numerical results. Original levels are retained so that every percentage can be recalculated, while rebased values make the cumulative comparison interpretable without confusing unequal original index bases.

An independent reviewer checked the series identities and recomputed the headline results from the official records using calendar-month matching and a separate arithmetic implementation. A further independent calculation checks all available relative changes using a cross-product formulation. These checks address extraction and numerical consistency; they are not an independent survey or a guarantee against limitations in the source measurement. The downloadable reproduction code provides an auditable path from the selected observations to the reported results.

The numbers support a bounded conclusion: throughout the computable five-year comparisons ending January 2021 through September 2026, the recreational-book index rose less, or fell more, than the all-items index. They do not support treating every format as identical, treating overlapping windows as independent evidence, or treating a relative-price decline as an affordability gain. BLS public-domain data may be reused with attribution; the calculations, relabeling and explanation here are separate analytical work and imply no BLS endorsement.

Book prices and the overall consumer basket

United States · CPI-U · Not seasonally adjusted. January 2016–September 2026

Full chart and source
Book prices and the overall consumer basket. United States · CPI-U · Not seasonally adjusted. October 2025 missing · no interpolation. Recreational-book category, not all formats; ebook subscriptions excluded. Book-club sampling changed in 2016; sampling and weights evolved. The 2025 collection interruption and imputation procedures affect interpretation of later indexes. Price-index comparison only: no affordability or causal conclusion. Index · January 2016 = 100 Source: U.S. Bureau of Labor Statistics · U.S. city average Recreational books: 100 → 97.42; All items: 100 → 142.22.

Scroll horizontally to inspect the chart. ↔

Chart and values

Scroll horizontally to see all columns.

Complete monthly evidence, January 2016–September 2026
MonthBook CPI, December 1997=100All-items CPI, 1982–84=100Book CPI, January 2016=100All-items CPI, January 2016=100
2016-01101.693236.916100.00100.00
2016-02102.714237.111101.00100.08
2016-0397.719238.13296.09100.51
2016-0497.654239.26196.03100.99
2016-0596.914240.22995.30101.40
2016-0696.729241.01895.12101.73
2016-0795.721240.62894.13101.57
2016-08100.823240.84999.14101.66
2016-0999.166241.42897.52101.90
2016-1098.293241.72996.66102.03
2016-1198.990241.35397.34101.87
2016-1299.481241.43297.82101.91
2017-0199.549242.83997.89102.50
2017-0297.043243.60395.43102.82
2017-0397.238243.80195.62102.91
2017-0496.152244.52494.55103.21
2017-0594.934244.73393.35103.30
2017-0693.140244.95591.59103.39
2017-0794.444244.78692.87103.32
2017-0894.719245.51993.14103.63
2017-0995.444246.81993.86104.18
2017-1095.682246.66394.09104.11
2017-1195.458246.66993.87104.12
2017-1296.292246.52494.69104.06
2018-0196.525247.86794.92104.62
2018-0296.135248.99194.53105.10
2018-0395.254249.55493.67105.33
2018-0497.061250.54695.45105.75
2018-0597.198251.58895.58106.19
2018-0696.420251.98994.81106.36
2018-0796.725252.00695.11106.37
2018-0896.713252.14695.10106.43
2018-0995.247252.43993.66106.55
2018-1095.310252.88593.72106.74
2018-1194.957252.03893.38106.38
2018-1295.586251.23393.99106.04
2019-0197.419251.71295.80106.25
2019-0296.293252.77694.69106.69
2019-0395.684254.20294.09107.30
2019-0496.138255.54894.54107.86
2019-0594.951256.09293.37108.09
2019-0693.414256.14391.86108.12
2019-0794.601256.57193.03108.30
2019-0895.032256.55893.45108.29
2019-0993.871256.75992.31108.38
2019-1096.465257.34694.86108.62
2019-1196.211257.20894.61108.57
2019-1295.611256.97494.02108.47
2020-0195.563257.97193.97108.89
2020-0298.419258.67896.78109.19
2020-0397.037258.11595.42108.95
2020-0495.811256.38994.22108.22
2020-0597.351256.39495.73108.22
2020-0699.258257.79797.61108.81
2020-0798.370259.10196.73109.36
2020-0898.916259.91897.27109.71
2020-0997.760260.28096.13109.86
2020-1099.908260.38898.24109.91
2020-1198.890260.22997.24109.84
2020-12100.111260.47498.44109.94
2021-0199.671261.58298.01110.41
2021-0298.427263.01496.79111.02
2021-0397.039264.87795.42111.80
2021-0498.641267.05497.00112.72
2021-0598.869269.19597.22113.62
2021-0697.778271.69696.15114.68
2021-0798.472273.00396.83115.23
2021-0898.455273.56796.82115.47
2021-0998.605274.31096.96115.78
2021-1097.840276.58996.21116.75
2021-1199.106277.94897.46117.32
2021-1297.559278.80295.93117.68
2022-0198.459281.14896.82118.67
2022-0299.608283.71697.95119.75
2022-0398.116287.50496.48121.35
2022-0499.109289.10997.46122.03
2022-0598.688292.29697.05123.38
2022-06100.273296.31198.60125.07
2022-0798.559296.27696.92125.06
2022-0899.786296.17198.12125.01
2022-09100.043296.80898.38125.28
2022-1097.832298.01296.20125.79
2022-1199.127297.71197.48125.66
2022-1298.988296.79797.34125.28
2023-0198.118299.17096.48126.28
2023-0298.558300.84096.92126.98
2023-0398.692301.83697.05127.40
2023-04100.091303.36398.42128.05
2023-0599.602304.12797.94128.37
2023-0697.165305.10995.55128.78
2023-0798.131305.69196.50129.03
2023-0897.974307.02696.34129.59
2023-09100.170307.78998.50129.91
2023-1099.639307.67197.98129.87
2023-1199.374307.05197.72129.60
2023-1297.912306.74696.28129.47
2024-0199.218308.41797.57130.18
2024-0295.536310.32693.95130.99
2024-0399.858312.33298.20131.83
2024-0498.626313.54896.98132.35
2024-0597.492314.06995.87132.57
2024-06100.234314.17598.57132.61
2024-0799.194314.54097.54132.76
2024-0899.385314.79697.73132.87
2024-09101.521315.30199.83133.09
2024-1098.658315.66497.02133.24
2024-1194.679315.49393.10133.17
2024-1297.433315.60595.81133.21
2025-01100.312317.67198.64134.09
2025-0299.108319.08297.46134.68
2025-0398.982319.79997.33134.98
2025-0499.644320.79597.99135.40
2025-0595.605321.46594.01135.69
2025-0699.282322.56197.63136.15
2025-0797.433323.04895.81136.36
2025-08100.747323.97699.07136.75
2025-0999.769324.80098.11137.10
2025-10UnavailableUnavailableUnavailableUnavailable
2025-11101.385324.12299.70136.81
2025-1298.221324.05496.59136.78
2026-0199.931325.25298.27137.29
2026-0296.404326.78594.80137.93
2026-0395.972330.21394.37139.38
2026-0493.168333.02091.62140.56
2026-0593.841335.12392.28141.45
2026-0691.816333.95290.29140.96
2026-0796.293333.91894.69140.94
2026-0898.520334.98096.88141.39
2026-0999.067336.94097.42142.22

Original levels have different bases and are not directly comparable. Rebased indexes independently set January 2016=100. October 2025 remains unavailable.

October 2025 missing · no interpolation Recreational-book category, not all formats; ebook subscriptions excluded. Book-club sampling changed in 2016; sampling and weights evolved. The 2025 collection interruption and imputation procedures affect interpretation of later indexes. Price-index comparison only: no affordability or causal conclusion. Index · January 2016 = 100 Source: U.S. Bureau of Labor Statistics · U.S. city average

Download chart (SVG) ↓ · Download chart (PNG) ↓

Books lagged the basket in all 68 observed windows

United States · CPI-U · Not seasonally adjusted. 69 scheduled windows · 68 valid · overlapping, not independent tests

Full chart and source
Books lagged the basket in all 68 observed windows. United States · CPI-U · Not seasonally adjusted. October 2025 missing · no interpolation. Ratio of growth factors: [(books end / start) / (all-items end / start) − 1] × 100 September 2021–September 2026: books +0.47%; all items +22.83%. −18.21% is a relative change, not a nominal book-price fall or a percentage-point gap. Price-index comparison only: no affordability or causal conclusion. 69 scheduled windows · 68 valid · overlapping, not independent tests Recreational-book category, not all formats; ebook subscriptions excluded. Book-club sampling changed in 2016; sampling and weights evolved. The 2025 collection interruption and imputation procedures affect interpretation of later indexes. Source: U.S. Bureau of Labor Statistics · U.S. city average

Scroll horizontally to inspect the chart. ↔

Chart and values

Scroll horizontally to see all columns.

Complete rolling five-year evidence
StartEndBook change (%)All-items change (%)Difference (percentage points)Relative change (%)Missing month within window
2016-012021-01−1.9910.41−12.40−11.23None
2016-022021-02−4.1710.92−15.10−13.61None
2016-032021-03−0.7011.23−11.93−10.72None
2016-042021-041.0111.62−10.61−9.50None
2016-052021-052.0212.06−10.04−8.96None
2016-062021-061.0812.73−11.64−10.33None
2016-072021-072.8713.45−10.58−9.33None
2016-082021-08−2.3513.58−15.93−14.03None
2016-092021-09−0.5713.62−14.19−12.49None
2016-102021-10−0.4614.42−14.88−13.01None
2016-112021-110.1215.16−15.05−13.06None
2016-122021-12−1.9315.48−17.41−15.08None
2017-012022-01−1.0915.78−16.87−14.57None
2017-022022-022.6416.47−13.82−11.87None
2017-032022-030.9017.93−17.02−14.44None
2017-042022-043.0818.23−15.16−12.82None
2017-052022-053.9519.43−15.48−12.96None
2017-062022-067.6620.97−13.31−11.00None
2017-072022-074.3621.03−16.68−13.78None
2017-082022-085.3520.63−15.28−12.67None
2017-092022-094.8220.25−15.43−12.84None
2017-102022-102.2520.82−18.57−15.37None
2017-112022-113.8420.69−16.85−13.96None
2017-122022-122.8020.39−17.59−14.61None
2018-012023-011.6520.70−19.05−15.78None
2018-022023-022.5220.82−18.30−15.15None
2018-032023-033.6120.95−17.34−14.34None
2018-042023-043.1221.08−17.96−14.83None
2018-052023-052.4720.88−18.41−15.23None
2018-062023-060.7721.08−20.31−16.77None
2018-072023-071.4521.30−19.85−16.36None
2018-082023-081.3021.77−20.46−16.80None
2018-092023-095.1721.93−16.76−13.74None
2018-102023-104.5421.66−17.12−14.07None
2018-112023-114.6521.83−17.18−14.10None
2018-122023-122.4322.10−19.66−16.10None
2019-012024-011.8522.53−20.68−16.88None
2019-022024-02−0.7922.77−23.55−19.19None
2019-032024-034.3622.87−18.51−15.06None
2019-042024-042.5922.70−20.11−16.39None
2019-052024-052.6822.64−19.96−16.28None
2019-062024-067.3022.66−15.36−12.52None
2019-072024-074.8622.59−17.74−14.47None
2019-082024-084.5822.70−18.12−14.77None
2019-092024-098.1522.80−14.65−11.93None
2019-102024-102.2722.66−20.39−16.62None
2019-112024-11−1.5922.66−24.25−19.77None
2019-122024-121.9122.82−20.91−17.03None
2020-012025-014.9723.14−18.17−14.76None
2020-022025-020.7023.35−22.65−18.36None
2020-032025-032.0023.90−21.89−17.67None
2020-042025-044.0025.12−21.12−16.88None
2020-052025-05−1.7925.38−27.17−21.67None
2020-062025-060.0225.12−25.10−20.06None
2020-072025-07−0.9524.68−25.63−20.56None
2020-082025-081.8524.65−22.79−18.29None
2020-092025-092.0624.79−22.73−18.22None
2020-102025-10UnavailableUnavailableUnavailableUnavailable2025-10
2020-112025-112.5224.55−22.03−17.692025-10
2020-122025-12−1.8924.41−26.30−21.142025-10
2021-012026-010.2624.34−24.08−19.372025-10
2021-022026-02−2.0624.25−26.30−21.172025-10
2021-032026-03−1.1024.67−25.77−20.672025-10
2021-042026-04−5.5524.70−30.25−24.262025-10
2021-052026-05−5.0924.49−29.58−23.762025-10
2021-062026-06−6.1022.91−29.01−23.602025-10
2021-072026-07−2.2122.31−24.53−20.052025-10
2021-082026-080.0722.45−22.38−18.282025-10
2021-092026-090.4722.83−22.36−18.212025-10

69 scheduled endpoints; 68 observed calculations. Endpoints are exactly 60 calendar months apart; adjacent windows overlap and are not independent tests. A missing interior month is flagged, not interpolated.

October 2025 missing · no interpolation Ratio of growth factors: [(books end / start) / (all-items end / start) − 1] × 100 September 2021–September 2026: books +0.47%; all items +22.83%. −18.21% is a relative change, not a nominal book-price fall or a percentage-point gap. Price-index comparison only: no affordability or causal conclusion. 69 scheduled windows · 68 valid · overlapping, not independent tests Recreational-book category, not all formats; ebook subscriptions excluded. Book-club sampling changed in 2016; sampling and weights evolved. The 2025 collection interruption and imputation procedures affect interpretation of later indexes. Source: U.S. Bureau of Labor Statistics · U.S. city average

Download chart (SVG) ↓ · Download chart (PNG) ↓

Data & downloads

Scroll horizontally to see all columns.

Complete monthly evidence, January 2016–September 2026
MonthBook CPI, December 1997=100All-items CPI, 1982–84=100Book CPI, January 2016=100All-items CPI, January 2016=100
2016-01101.693236.916100.00100.00
2016-02102.714237.111101.00100.08
2016-0397.719238.13296.09100.51
2016-0497.654239.26196.03100.99
2016-0596.914240.22995.30101.40
2016-0696.729241.01895.12101.73
2016-0795.721240.62894.13101.57
2016-08100.823240.84999.14101.66
2016-0999.166241.42897.52101.90
2016-1098.293241.72996.66102.03
2016-1198.990241.35397.34101.87
2016-1299.481241.43297.82101.91
2017-0199.549242.83997.89102.50
2017-0297.043243.60395.43102.82
2017-0397.238243.80195.62102.91
2017-0496.152244.52494.55103.21
2017-0594.934244.73393.35103.30
2017-0693.140244.95591.59103.39
2017-0794.444244.78692.87103.32
2017-0894.719245.51993.14103.63
2017-0995.444246.81993.86104.18
2017-1095.682246.66394.09104.11
2017-1195.458246.66993.87104.12
2017-1296.292246.52494.69104.06
2018-0196.525247.86794.92104.62
2018-0296.135248.99194.53105.10
2018-0395.254249.55493.67105.33
2018-0497.061250.54695.45105.75
2018-0597.198251.58895.58106.19
2018-0696.420251.98994.81106.36
2018-0796.725252.00695.11106.37
2018-0896.713252.14695.10106.43
2018-0995.247252.43993.66106.55
2018-1095.310252.88593.72106.74
2018-1194.957252.03893.38106.38
2018-1295.586251.23393.99106.04
2019-0197.419251.71295.80106.25
2019-0296.293252.77694.69106.69
2019-0395.684254.20294.09107.30
2019-0496.138255.54894.54107.86
2019-0594.951256.09293.37108.09
2019-0693.414256.14391.86108.12
2019-0794.601256.57193.03108.30
2019-0895.032256.55893.45108.29
2019-0993.871256.75992.31108.38
2019-1096.465257.34694.86108.62
2019-1196.211257.20894.61108.57
2019-1295.611256.97494.02108.47
2020-0195.563257.97193.97108.89
2020-0298.419258.67896.78109.19
2020-0397.037258.11595.42108.95
2020-0495.811256.38994.22108.22
2020-0597.351256.39495.73108.22
2020-0699.258257.79797.61108.81
2020-0798.370259.10196.73109.36
2020-0898.916259.91897.27109.71
2020-0997.760260.28096.13109.86
2020-1099.908260.38898.24109.91
2020-1198.890260.22997.24109.84
2020-12100.111260.47498.44109.94
2021-0199.671261.58298.01110.41
2021-0298.427263.01496.79111.02
2021-0397.039264.87795.42111.80
2021-0498.641267.05497.00112.72
2021-0598.869269.19597.22113.62
2021-0697.778271.69696.15114.68
2021-0798.472273.00396.83115.23
2021-0898.455273.56796.82115.47
2021-0998.605274.31096.96115.78
2021-1097.840276.58996.21116.75
2021-1199.106277.94897.46117.32
2021-1297.559278.80295.93117.68
2022-0198.459281.14896.82118.67
2022-0299.608283.71697.95119.75
2022-0398.116287.50496.48121.35
2022-0499.109289.10997.46122.03
2022-0598.688292.29697.05123.38
2022-06100.273296.31198.60125.07
2022-0798.559296.27696.92125.06
2022-0899.786296.17198.12125.01
2022-09100.043296.80898.38125.28
2022-1097.832298.01296.20125.79
2022-1199.127297.71197.48125.66
2022-1298.988296.79797.34125.28
2023-0198.118299.17096.48126.28
2023-0298.558300.84096.92126.98
2023-0398.692301.83697.05127.40
2023-04100.091303.36398.42128.05
2023-0599.602304.12797.94128.37
2023-0697.165305.10995.55128.78
2023-0798.131305.69196.50129.03
2023-0897.974307.02696.34129.59
2023-09100.170307.78998.50129.91
2023-1099.639307.67197.98129.87
2023-1199.374307.05197.72129.60
2023-1297.912306.74696.28129.47
2024-0199.218308.41797.57130.18
2024-0295.536310.32693.95130.99
2024-0399.858312.33298.20131.83
2024-0498.626313.54896.98132.35
2024-0597.492314.06995.87132.57
2024-06100.234314.17598.57132.61
2024-0799.194314.54097.54132.76
2024-0899.385314.79697.73132.87
2024-09101.521315.30199.83133.09
2024-1098.658315.66497.02133.24
2024-1194.679315.49393.10133.17
2024-1297.433315.60595.81133.21
2025-01100.312317.67198.64134.09
2025-0299.108319.08297.46134.68
2025-0398.982319.79997.33134.98
2025-0499.644320.79597.99135.40
2025-0595.605321.46594.01135.69
2025-0699.282322.56197.63136.15
2025-0797.433323.04895.81136.36
2025-08100.747323.97699.07136.75
2025-0999.769324.80098.11137.10
2025-10UnavailableUnavailableUnavailableUnavailable
2025-11101.385324.12299.70136.81
2025-1298.221324.05496.59136.78
2026-0199.931325.25298.27137.29
2026-0296.404326.78594.80137.93
2026-0395.972330.21394.37139.38
2026-0493.168333.02091.62140.56
2026-0593.841335.12392.28141.45
2026-0691.816333.95290.29140.96
2026-0796.293333.91894.69140.94
2026-0898.520334.98096.88141.39
2026-0999.067336.94097.42142.22

Original levels have different bases and are not directly comparable. Rebased indexes independently set January 2016=100. October 2025 remains unavailable.

Scroll horizontally to see all columns.

Complete rolling five-year evidence
StartEndBook change (%)All-items change (%)Difference (percentage points)Relative change (%)Missing month within window
2016-012021-01−1.9910.41−12.40−11.23None
2016-022021-02−4.1710.92−15.10−13.61None
2016-032021-03−0.7011.23−11.93−10.72None
2016-042021-041.0111.62−10.61−9.50None
2016-052021-052.0212.06−10.04−8.96None
2016-062021-061.0812.73−11.64−10.33None
2016-072021-072.8713.45−10.58−9.33None
2016-082021-08−2.3513.58−15.93−14.03None
2016-092021-09−0.5713.62−14.19−12.49None
2016-102021-10−0.4614.42−14.88−13.01None
2016-112021-110.1215.16−15.05−13.06None
2016-122021-12−1.9315.48−17.41−15.08None
2017-012022-01−1.0915.78−16.87−14.57None
2017-022022-022.6416.47−13.82−11.87None
2017-032022-030.9017.93−17.02−14.44None
2017-042022-043.0818.23−15.16−12.82None
2017-052022-053.9519.43−15.48−12.96None
2017-062022-067.6620.97−13.31−11.00None
2017-072022-074.3621.03−16.68−13.78None
2017-082022-085.3520.63−15.28−12.67None
2017-092022-094.8220.25−15.43−12.84None
2017-102022-102.2520.82−18.57−15.37None
2017-112022-113.8420.69−16.85−13.96None
2017-122022-122.8020.39−17.59−14.61None
2018-012023-011.6520.70−19.05−15.78None
2018-022023-022.5220.82−18.30−15.15None
2018-032023-033.6120.95−17.34−14.34None
2018-042023-043.1221.08−17.96−14.83None
2018-052023-052.4720.88−18.41−15.23None
2018-062023-060.7721.08−20.31−16.77None
2018-072023-071.4521.30−19.85−16.36None
2018-082023-081.3021.77−20.46−16.80None
2018-092023-095.1721.93−16.76−13.74None
2018-102023-104.5421.66−17.12−14.07None
2018-112023-114.6521.83−17.18−14.10None
2018-122023-122.4322.10−19.66−16.10None
2019-012024-011.8522.53−20.68−16.88None
2019-022024-02−0.7922.77−23.55−19.19None
2019-032024-034.3622.87−18.51−15.06None
2019-042024-042.5922.70−20.11−16.39None
2019-052024-052.6822.64−19.96−16.28None
2019-062024-067.3022.66−15.36−12.52None
2019-072024-074.8622.59−17.74−14.47None
2019-082024-084.5822.70−18.12−14.77None
2019-092024-098.1522.80−14.65−11.93None
2019-102024-102.2722.66−20.39−16.62None
2019-112024-11−1.5922.66−24.25−19.77None
2019-122024-121.9122.82−20.91−17.03None
2020-012025-014.9723.14−18.17−14.76None
2020-022025-020.7023.35−22.65−18.36None
2020-032025-032.0023.90−21.89−17.67None
2020-042025-044.0025.12−21.12−16.88None
2020-052025-05−1.7925.38−27.17−21.67None
2020-062025-060.0225.12−25.10−20.06None
2020-072025-07−0.9524.68−25.63−20.56None
2020-082025-081.8524.65−22.79−18.29None
2020-092025-092.0624.79−22.73−18.22None
2020-102025-10UnavailableUnavailableUnavailableUnavailable2025-10
2020-112025-112.5224.55−22.03−17.692025-10
2020-122025-12−1.8924.41−26.30−21.142025-10
2021-012026-010.2624.34−24.08−19.372025-10
2021-022026-02−2.0624.25−26.30−21.172025-10
2021-032026-03−1.1024.67−25.77−20.672025-10
2021-042026-04−5.5524.70−30.25−24.262025-10
2021-052026-05−5.0924.49−29.58−23.762025-10
2021-062026-06−6.1022.91−29.01−23.602025-10
2021-072026-07−2.2122.31−24.53−20.052025-10
2021-082026-080.0722.45−22.38−18.282025-10
2021-092026-090.4722.83−22.36−18.212025-10

69 scheduled endpoints; 68 observed calculations. Endpoints are exactly 60 calendar months apart; adjacent windows overlap and are not independent tests. A missing interior month is flagged, not interpolated.

Evidence & method

How we calculated it

  1. Use official BLS monthly NSA CPI-U levels for recreational books and all items, U.S. city average; exclude M13 annual averages.
  2. Predefined estimand: 100×[(B_end/B_start)/(A_end/A_start)−1], endpoints exactly 60 calendar months apart.
  3. Preserve all 129 calendar positions and October 2025 missingness; 68 of 69 scheduled rolling comparisons are computable.
  4. Independent source/math review and separate implementation check all available results; no significance test applied.

What the data cannot tell us

  • Index categories and sampling/weight methods evolve; not fixed-product prices or an all-format census.
  • October 2025 collection disruption and subsequent imputation limitations remain; 2025 collection reductions may increase item volatility.
  • Overlapping windows are dependent. All-items CPI includes books.
  • No household income, affordability, dollar book-price, sales or causal inference. Full-span mismatched calendar endpoints may retain seasonality.

Original sources

  1. BLS CPI series catalog

    BLS CPI series catalog: Series identities, population, adjustment status, reference bases and ending periods.

    Accessed October 30, 2026

  2. BLS all-items monthly levels

    BLS all-items monthly levels: CUUR0000SA0; monthly NSA CPI-U U.S. city average.

    Accessed October 30, 2026

  3. BLS recreation monthly levels

    BLS recreation monthly levels: CUUR0000SERG02; monthly NSA recreational books.

    Accessed October 30, 2026

  4. BLS entry-level item definitions

    BLS entry-level item definitions: RG021/RG022 category coverage, exclusions and book-club sampling change.

    Accessed October 30, 2026

  5. BLS 2025 shutdown CPI FAQ

    BLS 2025 shutdown CPI FAQ: Missing October 2025, imputation and later index production.

    Accessed October 30, 2026

  6. BLS chronology of CPI changes

    BLS chronology of CPI changes: 2018 geography and 2019–2020 establishment-frame changes.

    Accessed October 30, 2026

  7. BLS January 2023 weight update

    BLS January 2023 weight update: Annual spending-weight updates begin January 2023.

    Accessed October 30, 2026

  8. BLS 2025 collection reductions

    BLS 2025 collection reductions: Potential item-specific volatility from sample collection reductions.

    Accessed October 30, 2026

  9. BLS linking and copyright policy

    BLS linking and copyright policy: Public-domain data with source credit; no logo rights or endorsement implied.

    Accessed October 30, 2026

Cite & reuse this study

Alisson Bento. Recreational-book CPI lagged the U.S. consumer basket across observed five-year windows. Original secondary analysis of U.S. Bureau of Labor Statistics CPI-U monthly data, January 2016–September 2026. Sources accessed October 30, 2026. https://alissonbento.com/research/us-recreational-book-cpi-relative-five-year-windows/

Source data: U.S. Bureau of Labor Statistics, public domain; source credit requested. Original selection, calculations and explanation: Alisson Bento. No BLS endorsement or rights to its emblem are implied.

Corrections or editorial questions

Pass it on.

Recreational-book CPI lagged the U.S. consumer basket across observed five-year windows | Alisson Bentoalissonbento.com

WhatsAppFacebookXTelegramRedditThreadsEmailSMS