Buy Farmland Where Your Harvest Justifies the Cost of Land

We mapped all the farmland in the US by its ratio of land cost to expected agricultural value.

While great strides in technology have made it possible to grow food indoors and in otherwise occupied spaces, the vast majority of farming requires land—and a lot of it. In the US, the agricultural sector is going through a time of succession with an aging farming population looking to sell or lease their land to younger generations. Whether due to this demand or other market forces, farmland has been rising meteorically in price. In real terms, farmland in the US has doubled in value in the last 20 years—from an average of $1,500 per acre to $3,000. We examined publicly available data from the USDA to examine how farmland is distributed and to map out where the cost of farmland is best justified by its expected agricultural output.

Where do farms already exist?

Figure 1a: Map of the United States divided by county and colored by the percentage of each county's total area that is made up of farmland. The counties are colored on a spectrum from Black (representing a county with 0% farmland), to dark Red (50% farmland), to bright Yellow (Approaching 100% farmland).

Figure 1b: Map of the United States divided by county and colored by the percentage of each county's market value of agricultural goods per acre (differentiated into deciles). The counties are colored on a spectrum from Black (representing a county in the bottom decile of Agricultural Value Per Acre), to dark Red (Agricultural Value near the Median), to bright Yellow (Agricultural Value in the Top decile of all counties in the US).

The distribution of farmland in the United States has been driven by complex factors such as soil quality, population density, policy, and the availability of irrigation technology. The map we see today is the result; the vast majority of intensively farmed counties are in the middle of the country, with smaller regions of intensity on the East and West Coasts. When examining the distribution of farms by agricultural value per acre, we see a more even distribution of farming across the country, with gaps only in the most mountainous and arid regions.

While farmland value is certainly driven by its agronomic attributes, such as soil health and climate, the price of farmland is often related to factors that have nothing to do with agriculture. For example, a plot of farmland near a city has more value because it could be turned into a high-value residential neighborhood. Thus, a hypothetical new farmer with no preference about where they want to start farming should locate a parcel of land whose price is not inflated by non-agricultural influences. That’s not to say being near an urban area has no agricultural value. It is well-documented that proximity to an urban area can help a farmer’s bottom line through reduced transport costs and access to a premium-priced market. We’re going to look for those places using a ratio of the annual value of a county’s agricultural produce to its raw land value.

A Ratio of Land Cost to Annual Agricultural Revenue

The National Agricultural Statistics Service ( NASS) routinely gathers farmland statistics, and its most robust data comes from the bidecadal Census of Agriculture (AKA Ag Census). We retrieved the following data from the 2017 Ag Census the following data:

Name	Description	Unit
Value of Agricultural Land and Buildings	Current expected market value of farmland and buildings	Dollars per Acre in County ($/Acre)
Total Acres of Agricultural Land	All acres of land used for agricultural purposes in the county.	Land area per County (Acres)
Revenue from Agricultural Operations	Gross market value of produce before taxes and production expenses of all agricultural products sold	Dollars per County ($)
Net Income of Agricultural Operations	Net Income of farm after operating expenses	Dollars per County ($)

We further constructed the below variables from that data:

Name	Description	Unit
Percentage of County Acreage used for Agriculture	County Acreage in Farms divided by County Total Acreage	%
Revenue from Agricultural Operations per Acre	Revenue from Agricultural Operations per county divided by the number of acres of agricultural land in that county.	Dollars per Acre per Year ($/Acre-Year)
Net Income from Agricultural Operations per Acre	Net Income from Agricultural Operations per county divided by the number of acres of agricultural land in that county.	Dollars per Acre per Year ($/Acre-Year)

And finally we arrive at the ratios of interest:

Name	Description	Unit
Land Value to Annual Revenue	Value of Agricultural Land and Buildings divided by Annual Revenue from Agriculture	Unitless. Meaningful only for comparison across counties.
Land Value to Annual Net Income	Value of Agricultural Land and Buildings divided by Annual Net Income from Agriculture	Unitless. Meaningful only for comparison across counties.

Through mapping these ratios, we can identify counties where the value of the land’s expected agricultural output is high relative to its cost—in essence, where you can get the biggest bang for your buck if all you care about is profitable farming.

Mapping the Ratio

Figure 2a: Map of the United States divided by county and colored by the ratio of the cost of an acre of farmland to its expected revenue from agricultural production. The bright yellow and orange counties would be a good place for aspirational farmers to look for land, as they have a low ratio of land cost to agricultural revenue, whereas the dark purple and black counties have a high cost of land relative to their average annual agricultural value.

Figure 2b: Map of the United States divided by county and colored by the ratio of the cost of an acre of farmland to its expected net income from agricultural production. As above, the bright yellow and orange counties would be a good place for aspirational farmers to look for land, as they have a low ratio of land cost to agricultural revenue, whereas the dark purple and black counties have a high cost of land relative to their average annual agricultural value.

Considered on its own, Figure 2 is a way to visualize a cost-benefit analysis of buying land in one of these counties for agricultural production. Table 1 below shows the top 5 counties in the US with the lowest ratio—where the cost of land is very cheap relative to average annual agricultural revenue:

County	Ratio of Land Cost to Rev.	Land Cost ($/Acre)	Total Farmland (Acres)	Annual Rev. ($/Acre)
Kenedy, TX	0.09	$737	2,452	$8,036
Richmond, NC	0.29	$4,365	12,435	$15,213
Haskell, KS	0.48	$1,527	363,751	$3,187
Clark, NV	0.52	$3,109	2,130	$5,939
Beaver, UT	0.55	$2,226	63,597	$4,057

Table 1: Top 5 counties with the lowest ratio of land cost to annual revenue.

Exploring Figure 2 above shows a pretty similar pattern to Figure 1, but it’s where these two maps differ that are the most interesting. If a county is in the upper deciles of agriculture revenue per acre (brightly colored in Figure 1), but its land cost to revenue ratio is still high (darkly colored in Figure 2) that could indicate the land value is being driven up by non-agricultural factors. Conversely, if a county is in the middle or lower deciles of agriculture revenue per acre (darkly colored in Figure 1), but its land cost to revenue ratio is low (brightly Colored in Figure 2) then it may be a place to find land that is not attractive to, and therefore not priced for, non-agricultural uses.

We can find the biggest of these divergences by ranking each county by these two measures, and finding the counties with the biggest differences across the two rankings. For example, Napa County (the biggest agricultural county in Table 3 below), is ranked 147th of all counties for Annual Revenue per Acre, but falls to 2495th when land cost is considered. This is a tumble from the Top 5% to the bottom 20%. In Table 2, we see the hidden gems—counties that are not attractive when looking at revenue per acre, but become much more attractive relative to their cost of land.

County	Rank Change	Annual Rev. ($/Acre)	Ratio of Land Cost to Rev.	Annual Rev. Rank	Ratio of Land Cost to Rev. Rank
San Juan, UT	2,213	$107	3.43	2,549	336
Lea, NM	2,131	$99	4.15	2,589	458
Chaves, NM	2,020	$174	2.71	2,249	229
Apache, AZ	2,007	$43	6.35	2,881	874
Luna, NM	2,001	$138	3.74	2,392	391

County	Rank Change	Annual Rev. ($/Acre)	Ratio of Land Cost to Rev.	Annual Rev. Rank	Ratio of Land Cost to Rev. Rank
Philadelphia, PA	-2,555	$1,287	45.60	331	2,886
Bristol, RI	-2,420	$1,875	24.05	183	2,603
Napa, CA	-2,348	$2,241	19.70	147	2,495
Passaic, NJ	-2,283	$1,512	21.13	252	2,535
Nantucket, MA	-2,156	$2,536	15.43	120	2,276

Top 5 and bottom 5 counties that look relatively better/worse when we consider land cost.

With the below tool, you can explore these comparisons visually:

State Comparison Tool

Figure 3: Select a state with the drop-down menu and compare agricultural revenue to our ratio of land cost to agricultural revenue. (This tool may take a few seconds to load)

Limitations

NASS surveys only active farmland, so this map will not capture the potential of land that was not reported as part of a farm in 2017. Similarly, the revenue and net income measures reflect the type of agriculture that existed at the time of the survey. It may be that a farmer could take land in one of these counties and make more revenue by growing different crops or applying different farming techniques.

Furthermore, farmland has an incredible breadth of variation even on the acre-level of granularity, so the county-level average ratios presented here will not accurately represent every parcel within the county and they won’t express the other local issues of an area. For example, a number of New Mexico counties appeared on our Top 5 list (Table 2) above; our further research showed that those counties’ farmers and ranchers are not worried about land costs, but they are worried about where their water is coming from season to season. In the opposite direction, New Jersey farmland is extremely expensive, but the state has a very attractive tax structure for agricultural businesses that has allowed farming to stay there. An aspiring farmer should take this map as an invitation to look at counties with low land-cost-to-revenue ratios but applying due diligence to any particular parcel will remain essential.

Conclusions

Choosing where to buy land to start a farm is rarely a decision driven only by a profit motive, but farmland seekers should consider whether they are paying for non-agricultural drivers of land value. The tools and maps presented here have attempted to provide a preliminary filter to determine which regions are relatively better priced for their productive potential.