2020 Data Overview
The Agricultural Trade Multiplier (ATM) model prepared by USDA’s Economic Research Service (ERS) provides estimates of the economic activity and jobs supported by U.S. agricultural trade, with detail for 124 product groups. In addition to crops and livestock products, U.S. agricultural exports require output and labor from the firms that process, transport, and market the agricultural output that is exported, as well as from the firms that manufacture, transport, and market the inputs used to produce that output. Thus, U.S. agricultural exports supported economic activity in both the farm and nonfarm sectors. For calendar year 2020, ERS’s ATM model indicates that the $150.1 billion in U.S. agricultural exports supported 1,133,200 jobs and generated $154.3 billion in additional economic activity. In the context of this model, the word “jobs” refers to full-time civilian jobs, as measured in full-time equivalents (FTEs). An FTE is a conversion of the number of hours worked to an equivalent number of full-time positions; for its ATM model, ERS uses a ratio of 2,080 hours per year per FTE.
ERS’s ATM model relies upon the definition of agricultural products used by the World Trade Organization (WTO), except that the model also treats biodiesel as an agricultural product. USDA adopted the WTO’s definition of agricultural products as its standard definition to report agricultural trade at the start of calendar year 2021 (see USDA, Foreign Agricultural Service, “Updated ‘Agricultural Products’ Definition for Trade Reporting”). In the WTO definition, agricultural products are those products listed in Chapters 1-24 of the Harmonized Commodity Description and Coding System (HS), less fish and fish products, plus a handful of products in other chapters, such as cotton, essential oils, and hides and skins. Using the WTO definition, U.S. agricultural exports totaled $149.7 billion in 2020. In the ATM model, we also include biodiesel as an agricultural product, given the use of soybean oil and other agricultural products as feedstocks for U.S. biodiesel production, even though biodiesel is not classified by the WTO as an agricultural product. The inclusion of biodiesel adds about $388 million to U.S. agricultural exports in 2020, raising the total value of U.S. agricultural exports in the ATM model to $150.1 billion.
A Diversity of Export Products, Concentrated in Certain Product Groups
The United States exports various agricultural goods, but this trade is concentrated in a few product groups. In 2020, 10 product groups in the ATM model accounted for nearly half (49.8 percent) of total U.S. agricultural exports (including biodiesel) in terms of value: soybeans; corn; bovine meat; swine meat; wheat; cotton; soybean meal; almonds, fresh or dried, shelled; chicken cuts and edible offal; and distillers’ dried grains with solubles (DDGS) (see table 1). The labor required to produce, process, market, and transport agricultural exports varies by product group. So does the additional economic activity generated by the exports. For instance, each $1 billion of soybean exports in 2020 supported about 5,275 jobs and $720 million in additional economic activity (output), while each $1 billion of bovine meat exports supported about 11,329 jobs and $1.63 billion in additional output.
Impacts of Agricultural Trade in 2020
ERS’s ATM model for calendar year 2020 differs from the one used for calendar year 2019 in two key respects: the source of the model’s input-output data and its definition. While both models use data from the U.S. Department of Commerce’s Bureau of Economic Analysis (BEA), the model for 2020 uses the supply and use tables, while the model for 2019 relied upon a make-use approach.
The ATM model for 2020 uses the supply-use framework, which conforms to the international economic accounting standards. The supply table provides production information from domestic industries and imports. The use table identifies the usage of outputs in intermediate and final consumptions as well as exports. In contrast, the ATM model for 2019 used a make-use framework. The make-use approach also specifies the production and consumption of commodities by industry. This information creates the framework for the estimation process—with adjustments, additions, and other transformations introduced to translate the make-use tables into the supply-use framework. Switching to the supply-use framework enables the model to better incorporate the total domestic supply of goods and services from both domestic and foreign producers and how this supply is used across the U.S. economy.
The supply-use tables used in the model for 2020 rely on data before redefinitions, while the make-use tables used in the model for 2019 contain data provided after redefinitions. Data before redefinitions represent the industries’ outputs as reported to the BEA, while data provided after redefinitions represent the industries’ outputs where the output value of the secondary commodities has been attributed to the industry that produces those commodities as primary commodities. Because of differences in the definition of industries, the data before and after redefinitions are not directly comparable—even though the commodity totals match, the industry sizes are different.
Table 2 presents the results of the ATM model for calendar years 2019 and 2020 showing how much of U.S. economic activity was supported by agricultural trade. Due to rounding, figures in the text may not match those in the table, and individual numbers in the table do not necessarily sum to the totals shown. In 2020, the $150.1 billion of U.S. agricultural exports (line 6) produced an additional $154.3 billion (line 7) in economic activity for a total of $304.4 billion of economic output (line 1). The $150.1 billion in exports is freight on board (f.o.b.), meaning that the total export value includes trade and transport margins added to the production or farm cost to bring the product to the export port for shipping.
On average, each dollar of U.S. agricultural exports supported $2.03 of business activity, including the exports themselves (line 15). This output multiplier equals the ratio of the total economic output supported by U.S. agricultural exports ($304.4 billion) to the total value of these exports ($150.1 billion). By subtracting the number one from the output multiplier, we can calculate the additional economic activity supported on average by each dollar of U.S. agricultural exports—i.e., $1.03 (line 16). This new value equals the ratio of the additional economic activity supported by U.S. agricultural exports ($154.3 million) to the total value of these exports ($150.1 billion).
Agricultural exports in 2020 supported 1,133,200 full-time civilian jobs—439,500 jobs in the farm sector and 693,700 jobs in the nonfarm sector—although the latter two numbers do not sum precisely to the total due to rounding (lines 17–19). In the farm sector, these jobs included labor provided by farm operators, their family members, hired farmworkers, and contract workers. In the nonfarm sector, farmers’ purchases of fuel, fertilizer, and other inputs to produce commodities for export spurred economic activity in manufacturing, trade, and transportation.
The $154.3 billion of supporting or indirect activity generated by agricultural exports in 2020 includes activities both in the farm sector—i.e., crop and livestock production— and in nonfarm sectors related to production agriculture—e.g., food processing, other manufacturing sectors, and activities required to move exports to their final destinations. Examples of these nonfarm activities include computer and financial services, warehousing and distribution, packaging, and additional processing. The results emphasize the importance of agricultural exports to the services, transportation, and trade sectors, which account for $68.0 billion (line 13) of the total $154.3 billion in indirect activity. Supporting activity in the farm sector is estimated to be $32.2 billion (line10). The farm sector's $82.5 billion of output associated with agricultural exports (line 2) is the sum of exported raw agricultural products ($50.3 billion; not shown in the table) and the value of supporting activity in the farm sector ($32.2 billion). Overall, the farm sector’s share of the income supported by agricultural exports is 23.6 percent (line 37) whereas the share corresponding to the services, trade, and transportation sectors is 48.5 percent (line 38).
In its Bulk-Intermediate-Consumer Oriented (BICO) classification scheme, USDA divides agricultural trade into three broad categories of products: bulk, intermediate, and consumer-oriented. The last two categories make up nonbulk products—such as soybean meal, distillers’ dried grains with solubles (DDGS), and ethanol (intermediate products) and meat, fruit, and vegetables (consumer-oriented products). In the ATM model, bulk exports are defined as soybeans, other oilseeds, corn, rice (including milled rice), sorghum, wheat, other grains, pulses, cocoa beans, coffee (not roasted, not decaffeinated), tobacco, cotton, sugar beets and sugarcane, and seaweed and other algae. These product groups do not precisely match those in the BICO classification scheme because a few products classified as bulk in BICO are included in nonbulk product groups in the ATM model. Thus, the trade values for bulk and nonbulk products in the multiplier model differ somewhat from the ones reported in U.S. agricultural trade statistics. For the latest U.S. agricultural trade statistics, as well as a complete listing of the products in each BICO category, see USDA’s Global Agricultural Trade Statistics (GATS) database.
In 2020, U.S. bulk agricultural exports (as represented in the ATM model) equaled about $53.3 billion, (line 6). These exports supported an additional $46.5 billion of supporting business activity (line 9). Thus, on average, each dollar of bulk exports generated $0.87 of additional economic output (line 16). Nonbulk exports (all agricultural exports other than bulk exports; estimated at $96.9 billion in the model) stimulated an additional $107.7 billion of supporting business activity. Thus, each dollar of nonbulk exports generated $1.11 of additional output, on average.
Bulk exports have a proportionately larger effect on the nonfarm economy than nonbulk exports. For bulk exports, 90.1 percent ($41.9 billion) of the additional business activity took place in the nonfarm economy—including $25.7 billion in the services, trade, and transportation sectors. In comparison, 74.5 percent ($80.2 billion) of the additional business activity resulting from nonbulk exports occurred in the nonfarm economy, including $42.4 billion in services, trade, and transportation. By dividing the total economic activity supported by agricultural exports (line 1) by the economic activity in the farm sector supported by these exports (line 2), one can calculate the farm sector’s share of this activity. This calculation shows that the farm sector’s share of the total economic activity supported by agricultural exports is larger for bulk exports (46.4 percent) than for nonbulk exports (17.7 percent). This outcome makes sense because the processing of these bulk exports and the marketing and retailing of the resulting final products primarily take place abroad.
Accounting for the Impacts of Agricultural Imports
For the multiplier estimates, imports are interpreted as a subtraction from economic activity that would have occurred in the United States in the absence of those imports. This simplifying assumption is made due to data limitations that prevent us from accurately assessing the impact of imports and exports in the same way. It is not now possible to measure the total economic activity associated with imports because there are no end-use data on imports. When imports enter the United States, their value is recorded, but they are no longer tracked after that.
The end use of a product determines its multiplier effects. Imports can be put into inventory (an almost negligible multiplier) or used in a highly processed product (a very large multiplier). Thus, without end-use data, the indirect or supporting impacts of actual agricultural imports cannot be measured in terms of output, employment, or value added—or as a multiplier. Only the value of imports, as measured upon entry into the United States, can be discerned (direct effects).
To illustrate this point, consider cocoa. Almost all the cocoa consumed in the United States is imported. If reliable statistics on consumer demand for cocoa and its many uses were available, the supporting activity required to deliver imported cocoa could be measured. This would be only part of imported cocoa's contribution to the U.S. economy. Cocoa is used as an input for various products (e.g., cocoa powder, cocoa butter, and different types of chocolate). Moreover, some of these products are intermediate inputs to produce other products, such as chocolate bars, powdered drink mixes, candy, pastries, medicinal products, skincare products, ointments, and suppositories. To fully measure the economic activity supported by cocoa imports, one would also need to distinguish the effect of imported cocoa products from domestically produced cocoa products. The data to make this distinction do not exist.
Impacts of Agricultural Imports on U.S. Output
U.S. agricultural imports equaled about $146.8 billion in 2020 (line 7). ERS’s ATM model incorporates the assumption that certain agricultural products, primarily grown or manufactured in tropical areas, are noncompetitive with U.S. production and do not substitute for domestic production, as the United States lacks the extensive tropical climatic zones needed to produce these items in large quantities for commercial purposes. The imported product groups categorized as noncompetitive are bananas, plantains, pineapples, cocoa beans, coffee, palm oil, and tequila. All other imported agricultural products are treated as being potentially competitive with U.S. production. Within the total of $146.8 billion of U.S. agricultural imports, $11.9 billion are considered noncompetitive imports (not shown in table 2), and $134.9 billion are considered competitive imports (line 8).
If competitive imports had been produced domestically, the effect on domestic output would have been about $288.7 billion (line 24). Like exports, moving imported products to consumers supports jobs in the data processing, financial, legal, management, administrative, marketing, and transportation sectors. The import multiplier of 2.14 may be defined as the ratio of the economic value assumed to be displaced by competitive agricultural imports, including the imports themselves, ($288.7 billion) to the value of these imports ($134.9 billion). By subtracting one from this import multiplier, one gets 1.14, which equals the ratio of the additional economic value that would have been created in the U.S. economy if the competitive agricultural imports had been produced domestically instead to the value of those imports. Simply, each dollar spent on competitive agricultural imports in 2020 would have required another $1.14 in supporting goods and services had those imported items been produced domestically.
Overall Impacts of Agricultural Trade
U.S. agricultural trade (exports and imports) had a positive effect on all sectors of the economy in 2020. In the farm sector, $82.5 billion of output was either directly or indirectly associated with agricultural exports (line 2). This number more than offsets the estimated $73.8 billion of domestic economic output associated with competitive agricultural imports, had those imports been produced domestically instead (line 25). Thus, the farm sector’s net gain from agricultural trade was $8.7 billion (line 30). For all sectors of the economy, the net economic activity gained from U.S. agricultural trade (after the theoretical loss from competitive agricultural imports) was $15.7 billion (line 29). This number is larger than the net economic activity obtained by the farm sector from agricultural trade by the farm sector primarily because of the large amount of economic activity in the services, trade, and transportation sectors supported by U.S. agricultural exports. Considering both exports and competitive imports, trade in bulk agricultural products was associated with a net gain in economic activity of $79.6 billion; trade in nonbulk agricultural products was associated with a net loss of $63.9 billion (line 29). The net loss associated with trade in nonbulk agricultural products is because many agricultural imports embody food processing activities conducted abroad.
Comparison of Agricultural Trade Multipliers for 2019 and 2020
ERS’s Agricultural Trade Multipliers vary from one year to the next (table 2). Between 2019 and 2020, the output multiplier—the total economic activity supported by U.S. agricultural exports, including the exports themselves—decreased from $2.14 to $2.03 (5.1 percent) per dollar of such exports, and the jobs multiplier—the number of full-time equivalents (FTEs) supported per billion dollars of agricultural exports—fell from 7,784 to 7,750 (0.4 percent).
One possible explanation of the year-to-year decreases in the output and jobs multipliers is that bulk products accounted for a higher share of U.S. agricultural exports in 2020 than in 2019 (35.5 percent versus 30.9 percent). Since bulk exports by definition have received limited processing and have not been used yet to make a nonbulk product, their output and jobs multipliers are smaller than the ones for nonbulk exports. However, the output and jobs multipliers for bulk exports also declined between 2019 and 2020, as did the output multiplier (but not the jobs multiplier) for nonbulk exports.
A second possible explanation lies in the relative changes in export values and export quantities within specific categories of agricultural exports. Using export quantity data from USDA’s Global Agricultural Trade System (GATS), we can calculate the approximate total mass of U.S. agricultural exports. These calculations reveal that between 2019 and 2020, the value of bulk agricultural exports (measured in dollars) increased at a faster rate than their approximate mass (measured in metric tons): 21.7 percent versus 18.3 percent. Thus, each dollar of bulk exports corresponded to a smaller quantity of product in 2020 than in 2019, supporting less economic activity and employment according to the ATM estimates. In contrast, the value of nonbulk agricultural exports declined slightly—by 0.9 percent—while their approximate mass increased by 1.8 percent—indicating that each dollar of nonbulk exports corresponded to a larger quantity of product in 2020 than in 2019. This outcome corresponds to the increase in the output multiplier, but not the decrease in the jobs multiplier, for nonbulk exports.
 Young, J., T. Howells III, E. Strassner, and D. Wasshausen, 2015, “BEA Briefing: Supply Use Tables for the United States,” Survey of Current Business 95:9: 1-8.