Survey Results
Adoption rate of site-specific crop management technologies among US corn growers
By: Stan G. Daberkow and William D. McBride

Corn producers are the largest users of cropland and agri-chemicals in U.S. agriculture and represent a major market for precision agriculture technologies. Based on a USDA survey of corn producing farms in 16 states, about 9% utilized some aspect of precision agriculture for corn production in 1996, representing nearly 1/5th of 1996 harvested corn acreage. An analysis of the data indicated that farmers were more likely to adopt precision technologies if they farmed a large number of corn acres, earned a sizable farm income, and had high expected corn yields. The probability of adoption was also higher for farm operators using a computerized record system, who were less than 50 years of age, and who relied on crop consultants for precision agriculture information.
    U.S. corn producers, by virtue of their significant use of the nation's cropland and extensive agri-chemical, seed, and tillage applications, represent a potentially dominant market for precision agriculture technologies. Much of the U.S. corn is grown on or near environmentally sensitive lands which requires more intensive management. For example, in 1995 about 20% of the corn was grown on land designated as HEL and additional corn was produced near wetlands, shallow aquifers, rivers, streams, lakes, karst areas, etc. (USDA, ERS, 1997). Corn production accounts for nearly one-third of the cropland on which conservation tillage is practiced (USDA, ERS, 1997).

Data and Methods
Data collected in the 1996 Agricultural Resource Management Study (ARMS) were used to examine the use of precision agriculture technologies in corn production. One version of the 1996 ARMS collected detailed information about corn production practices and costs, and farm financial information. Sampling and data collection for the corn version of the ARMS involved a 3 phase process (Kott and Fetter, 1997). Phase 1 involved screening a sample of producers to determine whether or not each farm produced corn for grain. For phase 2, production practice and cost information was collected on a randomly selected corn field for a sample of corn growers. All respondents to the phase 2 interview were questioned about farm financial, management, and demographic information in phase 3.
    Respondents to all three phases of the 1996 ARMS for corn included 950 farms in 16 States. The States and their regional designation are: North Central-IL, IA, IN, OH, MI, MN, MO, WI; Southeast-KY, NC, SC; Plains-NE, KS, SD, TX; and Northeast-PA. The target population of this sample is farms planting any corn with the intention of harvesting grain. Each sampled farm represents a number of similar farms in the population as indicated by it's expansion factor. The expansion factor, or survey weight, is determined from the selection probability of each farm and expands the ARMS sample to represent the population of corn farms1.

Measuring Adoption
A farm operator is classified as an adopter of precision farming for corn production if, 1) any corn acres were soil grid sampled and mapped, or 2) any corn acres were fertilized or limed with variable rate technology (VRT), or 3) any corn acres were harvested using a combine equipped with a yield monitor. Because precision agriculture includes a relatively new set of technologies, this definition is intentionally broad to include the spectrum of farms experimenting with one or more components of precision agriculture.
    Using our broad definition of adoption, about 31,000 corn farms (approximately 9% of all corn farms) in the 16 States, with a range of 23,000 to 40,000 farms2, used one or more precision agriculture technology for corn production during the 1996 season. Among the specific technologies, 7% of farms used grid samples/maps, 4% applied fertilizer or lime with variable rate technology, and 6% used a yield monitor during corn harvest. Only 4% of farms used the yield monitor information to develop yield maps.
    The farm adoption estimate of about 9%, with no more than 7% for any individual technology, suggests that the diffusion of these technologies is very early in the typical adoption process (Rogers, 1983). However, these early adopters controlled a disproportional large share of the corn acreage. The 9% of farms using any precision agriculture technology farmed 19% of corn acreage, indicating that adoption has primarily occurred on larger farms.
    Corn producers employed the different precision agriculture technologies on different shares of their corn acreage. Among the adopters of precision agriculture technologies, soil grid sampling/mapping was the most widely used technology with 70% of farms sampling/mapping 64% of their corn acres. About 60% of adopters reported sampling in 2.5 acre grids, with 43% indicating a sampling frequency of once every 4 years. VRT was the least frequently used technology-36% of the adopters used VRT on about 55% of their corn acreage. Yield monitors were used by just over half the adopters (54%) on nearly all of their corn acres (94%).

Comparing Adopters and Non-Adopters
The comprehensive nature of the ARMS provided data on a variety of operator, farm structural and financial, and corn enterprise characteristics. Given the sufficiently large number of respondents indicating the use of one or more precision agriculture technologies, a number of traits of adopters and non-adopters could be compared using a difference of means test.

Operator Characteristics
The personal characteristics of farm operators who have adopted some form of precision agriculture technology differ in a variety of ways from non-adopters. While the average age of operator of the two groups was nearly the same, a significantly larger share of adopters (nearly 70%) were less than age 50 years of age compared to less than half of the non-adopters. A smaller share of adopters had only a high school or less education compared to the non-adopters, whereas a larger share of adopters had completed college. Over 90% of the adopters gave farming as their major occupation as opposed to only 75% of non-adopters-indicating more management time available for crop production decision-making. Likewise, adopters had gained more experience with computers in the farm business than non-adopters. The sources of information about precision agriculture differed between the two groups. Both groups relied heavily on farm suppliers and dealers, but adopters were more likely to seek out crop consultants and rely less on the Extension Service. In general, adopters were younger, more educated, less likely to work off the farm, more likely to seek out crop consultants, and more extensive computer users relative to non-adopters.
    Based on several measures of risk preference, adopters appeared to be less risk averse than non-adopters. Adopters had a significantly higher debt/asset ratio, had less crop and income diversity, and owned a smaller share of the land they farmed relative to non-adopters. A high debt/asset ratio indicates a willingness to accept greater financial risk. While adopters had a larger share of their gross cash income from crops, non-adopters relied more heavily on both crops and livestock. Diversification is one strategy to reduced both production and market risks. Nearly 85 % of the acres harvested by adopters consisted of two crops: corn and soybeans, whereas non-adopters were much more diversified. Even though adopters owned a smaller share of their corn acres (thus risking the loss of cropland in future years), they reduced their production and/or financial risk by share renting significantly more of their cropland than did non-adopters.

Farm Characteristics
By nearly any standard farm size or financial measure, those farms which have begun to utilize one or more precision technologies are bigger and more profitable than other corn farms. Acres operated, acres harvested, asset values, return on equity, and net income measures were between 1.5 to over 3 times larger for adopting farms relative to non-adopters. Adopting farms reported much higher normal and actual yields which likely reflects higher inherent soil productivity on these farms. The distribution of farms by sales class confirms the correlation of size with adoption of precision agriculture technologies. While over half of the adopting farms have sales of $250K or more, less than 20% of non-adopting farms were of that size. Nevertheless, about 18% of the adopting farms had less than $100,000 in gross sales in 1996. With net cash and farm income over $90,000 in 1996, adopting farms had the financial ability to experiment with this new technology. Even though adopting farms are much larger than non-adopters, the average net worth per farm is not statistically different between the two groups. This may reflect the more risk averse nature of non-adopters as indicated by their relatively modest use of debt.
    The location and farm type of precision technology adopters may reflect the availability of vendors as well as demand for such services. The vast majority (70%) of the corn farms (adopting and non-adopting) are located in the north central states with over 1/3 of all corn farms located in Indiana, Illinois, and Iowa. However, over half of the adopting farms were in the three central cornbelt states and over 2/3 of the non-adopters were located in the other states. The adopting farms are overwhelmingly specialized in cash grain production rather than in livestock production relative to non-adopters.
    Precision agriculture technologies offer a way to manage sub-field variability of soils, pests, landscape, and microclimates by spatially adjusting input use to maximize profits and potentially reducing environmental risks. Hence, the level and rate of adoption of precision agriculture technologies has implications for farm income as well as for the land and water resources associated with production agriculture. Furthermore, the characteristics of the farms and farm operators who have begun to adopt these spatial technologies offer insights to policy-makers concerned with improving farm income and reducing environmental risk.

2 Because estimated totals are from a sample survey, and not a complete census, interval ranges are included to indicate the extent of sampling error. A 95% confidence interval around the estimate is assumed.

About the authors:
Stan G. Daberkow and William D. McBride are economists with the Resource Economics Division of U.S. Department of Agriculture's Economical Research Service. The views expressed here are those of the authors and do not necessarily reflect positions of the U.S. Department of Agriculture.

Back