The U.S. Department of Agriculture (USDA) recently released its 2025 state-level data on farmland values. These estimates come from a survey in which producers are asked to report how much their land would sell for in a market transaction. As I describe in a recent OSU Extension article, the 2025 estimates indicate a small 0.4% real (inflation-adjusted) decline in farm real estate values overall, which measure the value of land and farm-related buildings combined. This amounts to the lowest annual rate of growth since 2013. Although it might be tempting to read it as evidence that Oregon’s farmland market may be stabilizing, it’s important to emphasize that the USDA’s values capture producer perceptions, or self-assessments, rather than observable market outcomes.
A while back I wrote a post that compared the self-reported USDA survey estimates to observed farmland prices. With newer data now available, this seems like a good opportunity to revisit that analysis and dig a bit deeper into why survey estimates differ from market prices. (Side note: Over the coming month or so, I’m expecting to receive another update that will allow our sales database to reflect sales through most of 2025.)
In general, market prices run considerably higher than survey estimates in a given year. This sometimes leads people to dismiss the USDA estimates as inaccurately capturing land market conditions. The estimates, however, come from surveys of producers, who, if anyone, should be well-informed about current market conditions. And it turns out that producers are actually pretty good at estimating the value of their land once we consider the population the USDA survey aims to capture.
Figure 1 plots the per-acre farm real estate (land and buildings) value from USDA for 2000-2025 against farmland prices over roughly the same period. The price trend, shown in light grey, represents a rolling 12-month average, updated monthly from January 2000 to July 2024 (the most recent month of sales data available). The smoothed black dashed line uses the same price data but makes it easier to distinguish the overall trend from month-to-month noise. This initial price trend uses 18,577 farmland sales at least 10 acres in size.
Across the 2000-2024 period, average market prices are $3,700 higher than the corresponding USDA survey estimate. Notably, this gap has widened in recent years, with the average difference being about $4,400 since 2019.
Although there are many small-acreage sales of farmland, involving 10 acres or less, these tend to inflate the overall price level because of the “small parcel premium” I’ve discussed before. When we raise the minimum sale area to 20 acres (Figure 2), we lose about 23% of the sales but the average gap between prices and survey estimates narrows to $2,800. Raising the threshold to 40 acres, which removes another 27% of the original 10+ acre sales sample, narrows the gap further to $1,700.
With a 100-acre threshold (Figure 4), we’re working with 23% of the initial sales sample, and the farmland price trend closely matches the USDA trend, at least in terms of its scale. On average, market prices are now just $284 greater than the USDA estimate, with several years where the USDA values actually exceed market prices. With the exception of the year or so, where the sales data are still incomplete, both measures of land value paint a consistent picture. I’m reluctant to read too much into the 2024 price spike until we have another year of complete sales data.
So why do the USDA values better reflect market prices for larger acreages? The reason is that the survey asks producers to report the market value of all land in their operation, not the market value of an individual 10-, 20-, or 40-acre parcel. According to the most recent USDA Census data from 2022, the average farm size in Oregon is 430 acres, though this varies widely from under 100 acres in some Willamette Valley counties to over 3,000 acres in parts of eastern Oregon. Excluding Clackamas County (38 acres), no other Oregon county has an average farm size under 40 acres. Put differently, although half of all sales involve parcels with a total area between 10 and 40 acres, the total amount of farmland on farms of that size represents a tiny fraction of Oregon’s farmland base, which is what the survey aims to capture.
Thus, the USDA survey estimates are best thought of as capturing, on average, what whole farms would be worth in a single market transaction. Farmland markets are thin to begin with, and there are very few actual sales that involve the types of acreage the USDA survey is designed to capture. In this sense, the survey estimates are hypothetical on two fronts: (1) they capture producer perceptions, as opposed to actual market prices, and (2) they represent sales that occur rarely in practice. It is worth keeping these two points in mind the next time you see what appear to be low land value estimates coming from the USDA survey.
Notes: Farmland price data come from a database of agricultural property transactions I developed using CoreLogic’s proprietary nationwide property transactions database. The 1999-2024 agricultural property sales used in this analysis are: (1) exclusively made up of 10 or fewer agricultural parcels (per CoreLogic’s land use codes), (2) between 10 and 4,000 total acres in size, (3) priced between $100 and $75,000/acre, (4) outside urban growth boundaries, (5) have at least 25% of the parcel area zoned exclusive farm use, farm-forest, marginal farmland, or non-public, and (6) have at least 50% of the parcel area in non-irrigated land capability classes 1-6.