↵ select ↓ ↑ navigate esc close

marginal revolution read out: USG assets are worth less than Arnold Kling had thought

marginal revolution read out ·

Listen: https://serve.buildtall.systems/56beeb80f557cbc88ce89af3cb17cbc49ac544c1ea91a657af42c27536485b4c.mp3

Source: https://marginalrevolution.com/marginalrevolution/2026/11/usg-assets-are-worth-less-than-arnold-kling-had-thought.html?utm_source=rss&utm_medium=rss&utm_campaign=usg-assets-are-worth-less-than-arnold-kling-had-thought


Today's marginal revolution read out episode covers the post USG assets are worth less than Arnold Kling had thought. The post draws on Arnold Kling's update to a 1984 valuation paper and reaches a sobering conclusion about federal asset values.

The first linked item is the Marginal Revolution post itself, USG assets are worth less than Arnold Kling had thought. It highlights the headline numbers from Kling's work: federal oil and gas rights are worth about $625 billion on the paper's own method, down 75 percent in real terms since 1981, while federal land has moved in the other direction, to roughly $1.1 trillion to $1.6 trillion. Land, not minerals, is now the larger asset. The post also points to the full piece, which the author describes as the full post.

The second item, the full post, is Arnold Kling's essay U.S. Government Assets: Less than I thought on his Substack. Kling begins by saying he was under the vague impression that in a pinch the government could sell off land and other assets, but compared to the tens of trillions of liabilities, the assets do not amount to much. He presents Claude's update of a 1984 paper by Boskin, Robinson, O'Reilly and Kumar that valued federal mineral rights and land. The updated central estimate puts federal oil and gas rights at $625 billion, down 75 percent in real terms since 1981, while federal land now comes in between $1.08 trillion and $1.56 trillion depending on the benchmark. On the paper's original comparison, oil and gas rights were 103 percent of privately held federal debt in 1981; today they are less than 2 percent of debt held by the public. Land has fallen too, though less dramatically, from 22 percent of debt to roughly 3 to 5 percent.

Kling's update explains three main drivers of the decline in mineral value. First, the Hotelling assumption failed: the 1984 paper assumed real oil prices would rise at the rate of interest, so no discounting was needed, but real oil prices instead fell about 33 percent over 45 years in his central case. Second, royalties and bonuses shrank: the offshore royalty rate fell from 16.67 percent to 12.5 percent, and bonus bids dropped from 27 percent of the 1981 value to about 10 percent. Third, economically recoverable offshore gas collapsed, from 167 trillion cubic feet to about 31 trillion, even though oil volumes are similar to the 1981 assessments. The author says he prefers a discounted figure of $321 billion for oil and gas rights, because the assumption that justified not discounting did not hold, and he notes the full range runs from about $195 billion to $875 billion.

The composition of mineral value has shifted. Onshore resources are now 41 percent of the total, up from 18 percent in 1981, largely because shale brought large federal volumes in New Mexico. Oil and liquids are about 84 percent of royalty value, against about two-thirds in 1981. Kling also gives market checks that point toward the lower end. Fiscal 2025 oil and gas revenue of $13.3 billion is 2.1 percent of the $625 billion value, but 4.1 percent of the discounted $321 billion, which is closer to a normal asset yield. Early Gulf lease sales under the 2025 law have drawn relatively small bids, while one onshore sale in May 2026 raised more than $4 billion.

On land, Kling extends the original series using farm real estate and house price indexes and also constructs a new benchmark from state pasture values and acreage. The central new estimate is about $1.55 trillion, with a lower bound near $1.15 trillion if federal land is judged worse than farmland. He cautions that pasture value is a generous comparison for desert and mountain land, and the urban figure is the weakest number because it relies on house prices rather than land itself. Much of this land also could not readily be sold, since national parks, wilderness and refuges are valued here as if they were grazing land, which says nothing about their public value.

The post places these assets in the federal balance sheet. Reported federal liabilities are $47.8 trillion against $6.1 trillion of reported assets. Federal land is unreported, as are oil and gas rights except for a small supplementary proved-reserves figure. Gold at market prices is now about $1 trillion, more than the mineral rights, reversing the 1984 paper's point that mineral rights were twice the value of federal gold. The official accounts have caught up only partly: they now report the royalty value of proved reserves, but undiscovered resources, future bonuses and land values remain outside the balance sheet.

Kling closes with limitations and a research agenda. The softest input is his own assumption about onshore economic volumes, since the USGS publishes only technically recoverable figures, not economic ones. Access is handled crudely, bonus history is incomplete, some volumes are dated, and costs and externalities are excluded. He says a price-supply curve for federal onshore resources, like the one for offshore, would be the most valuable new data, along with a parcel-level land benchmark and an annual bonus and royalty series.

The throughline across both linked items is that the federal government's saleable assets are far smaller relative to its liabilities than they were four decades ago. Land has held up better than minerals, but neither is an answer to a debt that now roughly equals GDP.

  1. USG assets are worth less than Arnold Kling had thought
  2. the full post