Real Commodity Prices (1960–2025): The Things We Grow Got Cheaper, the Ones We Burn Got Dearer

Across 66 years of World Bank commodity data, only two of eight price groups — energy and precious metals — cost more in real terms in 2025 than they did in 1960. Everything the world grows is cheaper, and the food we grow has fallen the most.

An open, inflation-adjusted dataset of World Bank commodity price indices, 1960–2025, deflated by US consumer prices and rebased to 1960 = 100 — eight category indices, from energy and metals to food and grains.

Eco3min Research · Last updated June 2026 · Frequency: Annual · 66 observations (1960–2025) · Licence: CC BY 4.0

A recurring belief returns with every commodity boom: that the world is running out of raw materials, and that real prices must therefore climb across the long run. The World Bank’s commodity price record, deflated by US consumer prices, tells a more divided story. Since 1960, the broad non-energy basket has lost roughly 43% of its real value, the things the world grows have fallen by half or more, and only two groups — energy and precious metals — have risen. This page documents that divergence with a reproducible, openly licensed dataset and explains what does and does not survive a change of inflation measure.

66 observations · Annual · 1960–2025 · CC BY 4.0
Methodology · Cite this dataset

The 65-Year Divergence

The hero chart above plots eight World Bank commodity indices in real terms — nominal dollar indices divided by US consumer prices, then rebased so that each equals 100 in 1960. On a single log scale, the fan opens wide. Two lines, energy and precious metals, climb far above the 1960 line and stay there. The rest — industrial metals, the non-energy basket, agriculture, food, grains — drift below it and end the period worth a fraction of their 1960 real value.

The bar chart below collapses that 65-year journey into one number per group: the real change from 1960 to 2025. It is the cleanest summary of the page’s claim.

What got cheaper, what got dearer

Real (inflation-adjusted) price change by commodity group, 1960→2025. Only two of eight rose.

Sources: World Bank Commodity Price Data (“Pink Sheet”); US BLS (CPI-U). Real annual-average index, 1960 vs 2025.

The Scarcity Reading

The intuition is old and powerful. A finite planet holds finite stocks of oil, copper, arable land and fresh water; a growing world economy draws those stocks down; therefore the real cost of raw materials would rise over time as the easy deposits and the best acreage are used up. Each commodity boom revives the argument. The 1970s brought talk of running out of oil and metals; the 2000s “supercycle,” powered by Chinese industrialisation, brought it back; the 2022 spike, with grain and gas prices surging after the invasion of Ukraine, revived it once more.

On a short enough window, the data appears to cooperate. From its 2001 low to the 2011 peak, the real non-energy basket more than doubled (39 to 85). An investor or commentator who began watching in the early 2000s saw a decade of rising real commodity prices and a coherent scarcity story to explain it. The question this dataset asks is what that same story looks like when the window is opened to 65 years rather than ten.

What the Data Shows

Stretched across the full record, the scarcity reading holds for energy and precious metals and fails for almost everything else. Three findings carry the page.

First, the basket you would expect to confirm scarcity is the one that fell most. Food, grains and broader agriculture are the commodities the Malthusian fear is really about — the things a growing population must keep eating. In real terms they are down 52%, 60% and 51% respectively since 1960. Industrial metals, the inputs of every factory and grid, are down about 19%. The only two groups that rose are energy (+288%) and precious metals (+627%) — and gold is less a consumed raw material than a monetary asset, while energy sits at the intersection of depletion and geopolitics rather than ordinary industrial demand.

Second, the headline index hides this. The all-commodity index is up 20% in real terms, which sounds like mild confirmation of rising prices. But energy carries by far the largest weight in that index, so the 20% is essentially energy bleeding through. Remove energy and the non-energy basket is down 43%. The single number most people cite for “commodities” is, in real terms, an energy index wearing a broad label.

Third, what looks like trend is mostly cycle. The grown commodities do not decline smoothly; they spike violently and revert. Food’s real index reached 198 in 1974, fell to about 39 by 2000, and the headline-grabbing 2022 spike reached only 67 — barely a third of the 1974 level and below where food traded through much of the 1960s. The mechanism is not mysterious: high prices pull forward investment and innovation — new acreage, higher-yield seed, better extraction — and supply catches up. The Green Revolution did to grain prices what shale later did to US gas.

A concrete way to see the deflation: between 1960 and 2025, US consumer prices rose about 10.9-fold. The nominal price of food rose 5.2-fold and grains 4.3-fold over the same span — far less than consumer prices, which is exactly why they fell by half in real terms. Energy’s nominal price rose 42-fold and gold’s 79-fold, outrunning inflation by a wide margin. Same dollars, opposite directions.

Where the Bears Are Right

The abundance reading has real limits, and a careful scarcity argument survives several of them.

The exceptions are not trivial. Energy is the master input of the entire economy, and a near-tripling of its real price over 65 years is precisely the kind of structural scarcity the bears describe. Precious metals rose even more. Pointing out that everything else got cheaper does not refute scarcity in the two categories that matter most for inflation and for safe-haven demand.

The result depends on the ruler. We deflate by US CPI because it is the most widely understood measure of “real.” The World Bank instead deflates its own real series by an index of manufactured-goods prices (the MUV index), the standard in the terms-of-trade literature. Under that ruler the picture shifts: food is down about 14% rather than 52%, the non-energy basket is roughly flat (+3%) rather than −43%, and industrial metals actually rise 46% rather than falling. The direction that is robust across both deflators is narrow but real: energy and precious metals up, food and agriculture down. The magnitudes, and the sign on industrial metals, are deflator-dependent — and we report both rather than choosing the flattering one. A complementary angle: our reference page on commodities as macroeconomic regime signals.

A finite endpoint is not a trend. Real prices sit near cyclical lows for grown commodities and well off their highs for energy. A scarcity proponent can reasonably argue that the next supercycle, or a sustained shift in energy and food systems, could reverse decades of decline. Nothing in a backward-looking record forecasts the next 65 years. What the record establishes is narrower: across this period, productivity has repeatedly overwhelmed depletion for the commodities the world grows and builds with, and the broad “running out of everything” claim is not what the deflated data shows.

The Food Anomaly

Food deserves its own panel because it is the cleanest case and the most counterintuitive. The thing a growing population most obviously needs more of has become, in real terms, roughly half as expensive as it was in 1960 — and every spike along the way has fully reversed.

The real price of food, 1960–2025

Each spike reverted; the trend is down. Real food prices have roughly halved.

Source: World Bank Food Price Index deflated by US CPI, rebased 1960 = 100. Annual averages.

The 1974 spike — driven by the oil shock, poor harvests and the Soviet grain purchases — took the real index to 198, nearly double its 1960 level. It never returned. Each later episode peaked lower: 74 in 2008, 77 in 2011, 67 in 2022. By 2025 the index sits at 48. The pattern is the signature of a market where each price shock funds the supply response that ends it, around a real trend that bends down.

Mean-Reversion by Real-Price Band

If real prices revert, then unusually high starting levels would tend to be followed by weak forward returns, and unusually low levels by stronger ones. Sorting the 66 years by the real level of the non-energy basket into thirds, and measuring the subsequent change, gives a directional but in-sample-limited confirmation.

The signal is real but is easily overstated. With 66 annual observations and overlapping ten-year windows, the effective sample is small, and the middle band’s forward returns are not cleanly ordered between the two extremes. The honest reading is directional: buying the non-energy basket when it was historically cheap in real terms has, in this record, tended to pay over a decade; buying it when expensive has not. That is consistent with mean-reversion around a declining real trend, not with a tradable rule.

Levels to Watch

These reference points describe where current values sit against the historical record. They are descriptive markers for reading future releases, not signals or targets.

Decade Table

Decade averages of the real index (1960 = 100) show the cycle from a longer vantage and make the divergence legible without the year-to-year noise.

*2020s = average of 2020–2025. Real index, 1960 = 100, deflated by US CPI.

Methodology

The dataset combines two primary, openly available sources and applies one transformation.

Commodity prices. Nominal annual commodity price indices (2010 = 100) come from the World Bank Commodity Price Data, known as the “Pink Sheet,” covering 1960 to 2025. We use eight category indices: the all-commodity headline, energy, the non-energy aggregate, agriculture, food, grains, metals & minerals, and precious metals. The World Bank constructs these from underlying spot prices using fixed value-of-production weights; full definitions are in its documentation.

Deflator. Nominal indices are divided by the US Consumer Price Index for All Urban Consumers (CPI-U), expressed on the same 2010 = 100 base. Annual CPI values for 1960–2024 come from the World Bank’s series for the United States; the 2025 value is the average of the available 2025 monthly CPI-U readings from the US Bureau of Labor Statistics, chained to the 2024 base. Each real series is then rebased so that 1960 = 100.

Choice of deflator, and robustness. CPI answers “expensive relative to what a US consumer pays.” The World Bank’s own real series instead uses an index of manufactured-goods export prices (the MUV G-5 index), the convention in the terms-of-trade literature, which answers “expensive relative to manufactured goods.” We report CPI-deflated figures as primary and cross-check against the MUV-deflated World Bank series. The robust, deflator-independent findings are that energy and precious metals rose and that food and agriculture fell. The magnitudes differ between deflators, and for industrial metals the sign differs: −19% under CPI, +46% under MUV. We disclose this rather than select the deflator that sharpens the headline.

Forward returns. The mean-reversion table sorts years by the real level of the non-energy basket into terciles and measures the subsequent 5- and 10-year real change. Windows overlap and the sample is small; the table is illustrative of direction, not a statistical test.

Download the Dataset

The full dataset — nominal indices, CPI-deflated real indices rebased to 1960 = 100, forward returns and real-price bands for all eight categories — is released under a Creative Commons Attribution 4.0 licence. Attribute to “Eco3min Research” with a link to this page.

Limitations

• Annual frequency. The dataset is annual; intra-year peaks (for example the monthly highs of 2008 or 2022) exceed the annual averages shown here.
• Deflator dependence. As detailed above, magnitudes and the sign on industrial metals depend on whether CPI or MUV is used. Real-price statements are only as fixed as the chosen deflator.
• Index construction. The World Bank’s fixed production weights and series revisions (noted in its documentation) shape the aggregates; category indices are not investable baskets and ignore roll yield, storage and financing.
• 1960 start. Beginning in 1960 excludes the high real prices of the early twentieth century. Longer reconstructions (for example academic series back to 1850) find production-weighted real prices roughly flat to modestly higher since 1900 — a reminder that the start date shapes the trend.
• Backward-looking. The record documents what happened; it is not a forecast and contains no view on future prices.

Frequently Asked Questions

Related research

• US CPI inflation history — the deflator behind this dataset
• The WTI–Brent spread, 1987–2026 — inside the energy index
• Eco3min commodities datasets — oil, gold, copper, wheat and more