Segment 1 — The Cold Open

Segment 2 — The Good Intention

The Conservation Reserve Program was written into the 1985 Food Security Act after years of visible soil loss across the Great Plains and Midwest. Lawmakers and USDA officials wanted to reduce water and wind erosion, cut federal grain surpluses, and give farmers a reliable rental payment in exchange for retiring highly erodible acres for ten-year periods. At the time the policy looked straightforward: pay landowners to take marginal cropland out of annual tillage, and the environmental gains would follow directly. JunJie Wu later documented how widely this expectation was shared among program designers. The contracts were voluntary, the payments were set by competitive bids, and early enrollment targets were met quickly because many operators already knew which fields were losing topsoil fastest. No one at the drafting table treated the rest of each farm as a fixed quantity; the assumption was that retired acres would stay retired.

The arithmetic seemed simple enough on paper. If an operator idled 100 acres that had been losing eight tons of soil per acre each year to sheet erosion, the gross saving would be 800 tons annually. The same calculation applied to wind erosion on the Northern Plains, where blowing dust from summer-fallowed fields could remove several inches of topsoil in a single dry season. Because the program used a bid system, the agency could select the most erosive parcels first, and the rental rate was calibrated to the local cash-rent market so participation would be attractive without overpaying. The designers therefore expected the measured reductions on contracted acres to translate almost one-for-one into watershed-level improvements in sediment load and downstream water quality.

Segment 3 — The Implementation

By 1990 more than 30 million acres had been accepted into CRP contracts across the central United States. The Department of Agriculture published annual reports showing large measured drops in sheet-and-rill erosion on the enrolled parcels. Proponents pointed to those numbers as proof the rental payments were working. At the same time, a few agricultural economists began asking whether the acres that remained in production were changing their own use patterns. Shinsuke Uchida would later use farm-level census records to test exactly that question. The early success stories focused on the contracted fields; the data that could reveal substitution across the rest of the same farms were not yet assembled.

Early implementation reports therefore recorded only the acres inside contract boundaries. When a county reported 50,000 acres enrolled, the erosion-reduction tables subtracted the historic erosion rates for those 50,000 acres and stopped there. The possibility that the same operator might now plant an additional 10,000 acres elsewhere on the same ownership was outside the accounting frame. Satellite imagery from the period later showed that some of the newly cultivated parcels had been hay meadows or woodlots whose conversion required clearing, drainage tiling, or breaking sod—operations that themselves generated short-term erosion spikes before the new cropland stabilized.

Segment 4 — The Unintended Consequences

JunJie Wu’s 2000 study in the American Journal of Agricultural Economics examined county-level data across the central United States and found that for every 100 acres enrolled in CRP, roughly 20 acres of non-cropland were converted to crops. The conversions occurred on land that had previously been pasture, hay, or woodland. Because those new acres were often more erosive than the average retired parcel, the net reduction in water erosion was about 9 percent smaller than the gross figure implied, and the net reduction in wind erosion was about 14 percent smaller. Shinsuke Uchida’s 2011 analysis of Agricultural Census longitudinal records reached a similar conclusion at the individual-farm level: partial participants converted an average of 25 percent of their remaining non-cropland to cropping after enrollment. Uchida’s 2014 working paper refined the estimate to 14 percent using fixed-effects methods that controlled for time-invariant farm characteristics. The mechanism was simple. A farmer who received CRP rental income on part of the operation could use that income to cover the costs of clearing or draining another part. The program rule paid only for the acres inside the contract boundary; it placed no restriction on the acres outside it. The result was slippage that appeared in satellite imagery as well as in census tabulations, documented by David Fleming’s 2010 study across more than a thousand Northern Plains and Corn Belt counties. The second-order effect was that total cultivated area declined less than program administrators expected, so some of the sediment and nutrient reductions that had been forecast never materialized at the watershed scale.

One objection sometimes raised is that the converted acres were marginal and would have been cropped eventually anyway. Yet the longitudinal farm records show the timing of conversion clustered immediately after CRP enrollment on the same ownership, and the magnitude of conversion scaled with the share of non-cropland the farm held at the moment of enrollment. Farms that entered the program with large reserves of convertible pasture or woodland showed the largest subsequent increases in cropped area. The cash-flow effect is straightforward: the annual CRP rental check arrived regardless of commodity prices, lowering the risk of investing in new drainage or clearing equipment on the remaining acres. Because the contract did not require the operator to keep the rest of the farm in its pre-enrollment state, the payment effectively subsidized the expansion of the cultivated footprint rather than simply shrinking it.

Segment 5 — The Aftermath

Later CRP rules added “sodbuster” and “swampbuster” compliance requirements that tied benefits to the treatment of non-enrolled land, yet the core rental contract still covered only the acres bid into the program. Researchers continued to publish slippage estimates, and some auction designs began to incorporate points that penalized farms with large reserves of convertible non-cropland. No large-scale reversal of existing contracts occurred. The program itself was reauthorized in subsequent farm bills with tighter environmental criteria, but the basic tension between paying for retired acres and leaving the rest of the farm unconstrained remained.

Segment 6 — The Lesson

When a payment rule rewards the retirement of one parcel while leaving every other parcel on the same ownership free to change use, the measured gain on the paid parcel will overstate the net change across the whole farm, as Wu’s and Uchida’s data showed. A second principle is that slippage becomes larger when participants hold more convertible non-cropland at the moment they enroll, because that capacity is what the payment frees them to exercise. A third is that adding compliance conditions on the rest of the farm after enrollment began was an attempt to close the gap the original contracts had left open. How might a new land-retirement program today decide which acres outside the contract boundary it is willing to monitor and constrain?

💬 Reply to this email — Patrick reads every one.

Share: X · LinkedIn · WhatsApp

Forwarded this email? Subscribe here — it's free.