Topmost menu

Who Benefits from Agricultural Subsidies?

Who benefits from farm subsidies? If you ask a farmer, you’ll get the answer, “I do.” Ask an economist, however, and you will get an entirely different answer. Here are quotes from some prominent economists that convey the conventional wisdom.

Andrew Schmitz and Richard Just:

A large share of the benefits of an agricultural subsidy goes to the landowner.

Hal Varian:

The benefits from the subsidy program still falls on those who initially own the land, since it is ultimately the value of what the land can earn—either from growing crops or farming the government—that determines its market value.

Robert Reich:

The lavish farm subsidies contained in the new farm bill won’t make the nation more secure. They will only stimulate even more production [and] inflate land values …

In other words, while the farmer claims to benefit from farm subsidies, economists attribute the benefits to land owners. You might think these two points of view can be reconciled if the farmer owns the farmland being subsidized, in which case the farmer benefits from the increased asset value. In fact 45 percent of subsidized farmland is rented. And it’s not rented from other farmers. Less than 10 percent of land owned by farmers is rented to other farmers. So if economists are right, and land owners primarily reap the benefits of farm subsidies, only about 60 percent of subsidy dollars benefit farmers. The rest get passed through to landlords—who do not farm—through higher rental rates and land values.

If these two viewpoints can’t be reconciled, which one is correct? Fortunately, this is a question with an objective answer. If we have the right data, we can measure the relationship between subsidies, rental rates, and farmers’ net returns. If we have the right tools, we can say with some confidence how subsidies affect rental rates and net returns. The important thing is not to confuse the relationship with the effect. Subsidies and rental rates are positively related. But that’s because more productive land commands a higher rental rate and receives higher subsidies, not (necessarily) because subsidies cause the rental rate to be higher. This is illustrated in the following two figures.


Figure 1: The Geographic Distribution of County-Average Corn Yield



Figure 2: The Geographic Distribution of County-Average Direct Payments

Figure 1 illustrates the geographic distribution of county-average corn yield in the U.S., and figure 2 shows the geographic distribution of direct payments per acre. Clearly these are highly correlated. But a lot of that correlation comes from the fact that, on a qualifying acre, direct payments are calculated as the subsidy rate (set by Congress) times the “program yield.” For corn, the program yield is the Olympic average of the actual yield in 1981–1985. Since soil productivity changes very slowly, the yield in the early eighties is still highly correlated with the yield today. So it’s possible that the correlation observable in the figures is entirely due to farmland productivity. Because more productive land receive higher subsidies, it is crucial to disentangle the subsidy effect from the productivity effect.

The Approach

Over the past 10 years numerous studies have concluded that landlords receive little of the subsidy dollar; the estimates range from about 6 percent in the European Union to 30 percent for Kansas wheat farms in the long run. These studies separate the subsidy effect from the productivity effect by observing the same farms over multiple years. This has the advantage that, since land productivity changes very slowly, it’s possible to essentially “subtract out” the productivity effect, leaving only the subsidy effect.

While this approach has its strengths—subtracting out the productivity effect—it also has its weaknesses—it isn’t possible to distinguish between subsidies paid to rented land and subsidies paid to owned land; only total subsidies are observed. However, subsides and land quality on rented land may differ from owned land. Since most farms possess both rented and owned acreage, farm-level measures of subsidies and rental rates may bias the estimated relationship. Field-level data are necessary to accurately estimate the effect subsidies have on farmland rental rates.

Unfortunately, no nationally representative data in the U.S. observes the same field annually. The only annual, field-level, nationally representative data available in the U.S. come from the Agricultural Resource Management Survey (ARMS). The ARMS sampling procedure, however, is explicitly designed to prevent farms and fields from being surveyed in multiple years (Perry et al., 1993). Yet it targets about 5,000 fields and 30,000 farms each year. Michael Roberts, at the University of Hawaii, and I have teamed up to examine the possible bias from farm-level data. We added questions to the 2006 and 2007 ARMS survey that elicited information on rent paid and subsidy payments received on a randomly chosen field on the surveyed farm. These data are unique because they link subsidies to the specific cash-rented parcels being subsidized.

Empirical Results

Table 1: Incidence Estimates for Three Crops at Three Levels of Aggregation

Farm Level Field-Level Rent Field Level
Farm-Average Subsidy
Soybeans 0.836*** 0.511*** 0.137*
(0.190) (0.124) (0.081)
Rice 0.514** 0.168** 0.142***
(0.248) (0.073) (0.041)
Cotton 0.383 0.170 0.249**
(0.318) (0.143) (0.115)

Table 1 reports the estimated change in rental rates due to an extra dollar of subsidy per acre (standard errors are in parenthesis)—what economists call the “incidence.” The table reports the results for three crops at three levels of aggregation. The same farms are used in all three columns within each crop. The first column, labelled “Farm Level,” reports the estimated effect when we use farm-average rental rates and farm-average subsidies—the approach one would have to take if only farm-level data were available. The second column reports the estimated effect using the field-level rental rate and the farm-average subsidy as a proxy for the field-level subsidy. The third column reports the result when field-level rental rates and subsidies are used. Comparing the three columns, it’s clear that when you can’t observe the same farm over time, the farm-level data substantially overestimates the true relationship. The field-level incidence estimate for both soybean fields and rice fields is about 0.14—the field’s rental rate increases by fourteen cents when the field receives an extra subsidy dollar. The farm-level estimate is 0.836 for soybeans and 0.514 for rice, a substantial overestimate. For cotton fields, the farm-level estimate is closer to the field-level estimate but still 50 percent bigger.

The second column reveals that the farm-average per-acre subsidy is a bad proxy for the soybean fields, but a fairly good proxy for rice and cotton fields. We are examining the reasons for this distinction in on-going research; it might suggest that for rice and cotton farms, rented and owned fields have similar characteristics. The estimate for soybean fields, however, indicates substantial differences between rented and owned soybean fields.


Based on field-level estimates, it appears that landlords do receive some of the subsidy benefits through higher rental rates, but they receive far less than economists would have expected. We find that using farm-level data from the most widely used farm-level data set, ARMS, overestimates the incidence considerably. Using a new, field-level data set that, for the first time, precisely links subsidies to land parcels, we show that farm-level data lead to considerably biased estimates: Where farm-level estimates suggest an incidence of 39–84 cents of the marginal subsidy dollar, field-level estimates from the same farms indicate that landlords capture just 14–25 cents.


Hendricks, Nathan P, Joseph P Janzen, and Kevin C Dhuyvetter. 2012. “Subsidy Incidence and Inertia in Farmland Rental Markets: Estimates from a Dynamic Panel.” Journal of Agricultural and Resource Economics 37 (3): 361–378.

Michalek, Jerzy, Pavel Ciaian, and d’Artis Kancs. 2014. “Capitalization of the Single Payment Scheme into Land Value: Generalized Propensity Score Evidence from the European Union.” Land Economics 90:260–289.

Perry, Charles, Jameson Burt, and William Iwig. 1993. “Redrawing the 1993 Farm Cost and Returns Survey List Sample to Reduce Overlap with Three Other 1993 Surveys and the 1992 FCRS,” 632–637.

Reich, Robert. 2001. Subsidies Aren’t a Wartime Necessity, The Wall Street Journa.l October.

Varian, Hal R. 2010. Intermediate Microeconomics: A Modern Approach. 8th ed. WW Norton New York.