Show Summary Details

Page of

PRINTED FROM the OXFORD RESEARCH ENCYCLOPEDIA,  ENVIRONMENTAL SCIENCE ( (c) Oxford University Press USA, 2020. All Rights Reserved. Personal use only; commercial use is strictly prohibited (for details see Privacy Policy and Legal Notice).

date: 21 February 2020

CAFOs: Farm Animals and Industrialized Livestock Production

Summary and Keywords

Industrialized livestock production can be characterized by five key attributes: confinement feeding of animals, separation of feed and livestock production, specialization, large size, and close vertical linkages with buyers. Industrialized livestock operations—popularly known as CAFOs, for Concentrated Animal Feeding Operations—have spread rapidly in developed and developing countries; by the early 21st century, they accounted for three quarters of poultry production and over half of global pork production, and held a growing foothold in dairy production.

Industrialized systems have created significant improvements in agricultural productivity, leading to greater output of meat and dairy products for given commitments of land, feed, labor, housing, and equipment. They have also been effective at developing, applying, and disseminating research leading to persistent improvements in animal genetics, breeding, feed formulations, and biosecurity. The reduced prices associated with productivity improvements support increased meat and dairy product consumption in low and middle income countries, while reducing the resources used for such consumption in higher income countries.

The high-stocking densities associated with confined feeding also exacerbate several social costs associated with livestock production. Animals in high-density environments may be exposed to diseases, subject to attacks from other animals, and unable to engage in natural behaviors, raising concerns about higher levels of fear, pain, stress, and boredom. Such animal welfare concerns have realized greater salience in recent years.

By consolidating large numbers of animals in a location, industrial systems also concentrate animal wastes, often in levels that exceed the capacity of local cropland to absorb the nutrients in manure. While the productivity improvements associated with industrial systems reduce the resource demands of agriculture, excessive localized concentrations of manure can lean to environmental damage through contamination of ground and surface water and through volatilization of nitrogen nutrients into airborne pollutants.

Finally, animals in industrialized systems are often provided with antibiotics in their feed or water, in order to treat and prevent disease, but also to realize improved feed absorption (“a production purpose”). Bacteria are developing resistance to many important antibiotic drugs; the extensive use of such drugs in human and animal medicine has contributed to the spread of antibiotic resistance, with consequent health risks to humans.

The social costs associated with industrialized production have led to a range of regulatory interventions, primarily in North America and Europe, as well as private sector attempts to alter the incentives that producers face through the development of labels and through associated adjustments within supply chains.

Keywords: CAFOs, factory farms, agricultural productivity, industrialized agriculture, industrial livestock production, animal agriculture, animal welfare, manure regulation


Much of modern livestock production occurs in industrialized farms. The animals are not raised in pastures but are instead confined in pens, barns, or houses, where they consume feeds specifically formulated for their species, age, and life cycle. Much or all of the feed for the animals is purchased rather than grown onsite. The farms are specialized, focusing on a single species of livestock and often on a single stage of production. The livestock enterprises on these farms are often quite large, with much larger herds and flocks than was common as recently as the 1980s, and the farms often maintain close relationships, through contracts or ownership, with processors, meatpackers, or independent integrators who serve as intermediaries.

These attributes—confinement feeding, separation of feed and livestock production, specialization, large size, and close vertical linkages with buyers—characterize a large and growing share of livestock production around the world. Industrial systems are particularly important in poultry and swine: by the early 2000s, industrial systems accounted for 55% of global pork production and 72% of global poultry meat production, and they have grown since (Table 1). They are quite widespread in developed countries but are important and spreading rapidly in developing countries. In particular, a rapidly growing middle class, particularly in Asian countries, has led to growing demand for meat and dairy products. Total meat consumption in developing countries, which nearly tripled between 1980 and 2002, is projected to grow by another 150% between 2002 and 2050, compared to a 25% increase in developed countries (Thornton, 2010). Much of that consumption growth is occurring in pork and poultry products, and in urbanized areas, and industrialized production systems are being introduced in developing countries to meet that demand growth.

Table 1. Industrial Livestock Systems in Pork and Poultry

Country Grouping

% of Production in Industrialized Systems


Poultry Meat

Developed Countries



Developing Countries






Source: Steinfeld et al. (2006).

Industrialized livestock operations consolidate manure as well as animals; they therefore can generate environmental risks and are subject to regulation of their manure management practices. In the United States the entities that are regulated are called Concentrated Animal Feeding Operations, or CAFOs. The term is now a widely used all-purpose signifier of large industrialized livestock operations, but not all industrialized operations are CAFOs, and some CAFOs are rather small. Here, the term “industrialized” refers to the broad set of practices and farms, and the term “CAFO” is used to refer to specific regulatory initiatives.

The estimates in Table 1 are drawn from statistics generated by the United Nations Food and Agriculture Organization (FAO), and in FAO analyses the term “industrialized production” is used interchangeably with “landless production,” where no more than 10% of feed is raised on the farm (Steinfeld, Wassenarr, & Jutzi, 2006). This is a restrictive definition: even in the United States, where industrialized systems are most widely used, large confined feeding dairy farms combine homegrown with purchased feed, while many poultry and pork producers raise crops, even if the crops are not used directly to feed the animals onsite. This analysis relies on a less restrictive definition, allowing for wider use of homegrown feed in industrialized systems.

How Industrial Livestock Production Is Organized

Because the United States relies most heavily on industrialized production systems, and experience there influences the design of such systems around the world, it is useful to start by detailing the key features of US livestock production.

For most of the country’s history, US farms kept multiple species of livestock; even as late as 1960, more than 70% of farms had cattle, nearly 60% had chickens, and nearly half had hogs—the same incidence as milk cows (see Figure 1). That changed sharply in the last half century, as hog, dairy, and poultry production consolidated onto a relatively small number of farms that specialized in a single species and a single animal production stage. By 2010, only 3% of US farms had hogs, another 3% had milk cows, and 10% had poultry of any kind.

CAFOs: Farm Animals and Industrialized Livestock Production

Figure 1. U.S. livestock operations became more specialized throughout the 20th century.

Source: U.S. Department of Agriculture and U.S. Bureau of the Census, Census of Agriculture, with linear interpolation between census.

Industrial systems are most developed in poultry. Almost all production of broilers—young chickens raised for meat—is carried out by independent growers under contract with firms called integrators (MacDonald, 2014). Those firms—leading examples include Tyson Foods, Pilgrim’s Pride, and Perdue Farms—provide growers with feed, chicks, and veterinary services, and then transport the mature birds from growers’ farms to their own processing plants. The integrators mix feed rations—tailored to broiler ages—in their own mills from purchased corn, soybean meal, and other ingredients. They also contract with other “broiler breeder” farms to raise young hens and to produce eggs for the integrators’ hatcheries, the source of chicks for contract growers’ farms. Growers, who contract exclusively with a single integrator, provide housing, labor, and utilities, and they bear responsibility for disposal of broiler litter. The birds are raised in houses, which are now fully enclosed and feature automated feeding and climate control systems. Some growers also raise crops, using poultry litter as fertilizer, but many do not.

As in broiler production, turkey and egg production occurs on specialized farms, while other specialized operations raise replacement layers for egg operations, or operate turkey hatcheries and egg producers for those hatcheries. Also like broilers, production occurs indoors in specially designed houses. The farms rarely grow their own feed but instead rely heavily on feed produced elsewhere. However, in contrast to broilers, full vertical integration is more common in turkeys and eggs, with the farms often owned and operated by the firms who also process and distribute turkey and egg products. Specialized firms operate primary breeding facilities focused on genetic improvements in poultry, and primary breeders provide integrators with parent stock for their own breeding flocks.

Through the 1980s, most hog production occurred on farms that birthed (farrowed) hogs, raised them to market weight with feed largely grown on the farm, and then sold the hogs to packing plants for slaughter and processing. While such independent “farrow-to-finish” operations still exist, most production now occurs within networks of more specialized farms. Pigs are farrowed on sow farms, where they are kept until weaning (18 to 30 days); typically, they then are transported to nursery farms for 35 to 50 days where they receive specifically designed feeds and veterinary attention before being sent to finishing farms, where they will be raised to market weight (60 to 75 days) with rations designed for that stage (some farms combine nursery and finishing stages). In each stage, the animals are confined to specially designed houses with controlled access. While many finishing operations also raise field crops and use hog manure for crop nutrients, the crops are usually sold rather than fed directly to the hogs on the farm.

Integrators organize production through contracts with specialized nursery and finishing farms; they provide those farms with young animals, with technical guidance and veterinary services, and with feed either purchased from mills or formulated in mills owned by the integrators. Some integrators own their own processing facilities, but in contrast to broiler production many instead market their hogs to meatpackers under other contract arrangements.

Beef production is different. Like pork, it is arranged into separate stages, but most of the animals move between stages through market transactions, with integration across stages still a rare phenomenon. Calves are born on “cow-calf” operations, where they are raised outdoors on pasture or rangeland until weaned from their mothers; at that point, they may continue to be raised on the same farm or ranch, or they may be sold to “stocker” operations, where the animals are still grazed outdoors. Replacement cows and bulls for cow-calf herds may be retained on site, or they may be purchased from other operations that specialize in the production of high-quality seed stock. Cow-calf and stocker farms and ranches are located all over the United States; in the most recent census of agriculture (2012), one third of U.S. farms and ranches (about 728,000) had beef cow herds (in contrast, 28,000 farms had sows for breeding pigs, and 64,000 farms had milk cows).

The industrial stage of beef production in the United States occurs next, in feedlots; while there are cattle feeders across the country, most feeding occurs in the Great Plains, particularly in the States of Texas, Colorado, Kansas, and Nebraska, near slaughter plants. Cattle are shipped to feedlots when they weigh between 500 and 900 pounds, with an average near 700 pounds. There they are confined in pens with similar cattle, fed rations of grain, protein, and forages, and raised until they reach market weight (between 1,150 and 1,400 pounds) and are sold to a meatpacker. About one in six cattle are fed in small “farmer-feedlots,” with one-time capacities of less than 1,000 head, and where the feeder combines homegrown and purchased feeds. However, most are fed in a few hundred large specialized feedlots, with capacities of up to 100,000 head and that rely on purchased feeds, often mixed in onsite mills.

U.S. dairy farms vary widely in size (up to 30,000 milk cows) and rely on specialized dairy breeds (over 90% of U.S. milk cows are Holsteins, and about 5% are Jerseys). Many smaller farms still graze their cows on pasture; however, most are confined within large “free-stall” dairy barns or lots. Replacement heifers for dairy herds are birthed on site but are often shipped to specialized heifer ranches to be raised. Most dairy farms raise at least some of their own feed, although almost all also purchase some.

Unlike hogs, poultry, and cattle-feeding, there are few large public corporations involved in milk production as integrators or producers, and the vast majority of dairy farms are independent family-owned businesses; however, they typically operate within a network of firms and cooperatives that provide testing, artificial insemination, procurement, and consulting services. Dairy farms usually market their milk through farmer-owned cooperatives that process the milk directly or sell it to independent firms for processing.

Some U.S. farms raise “free-range” broiler chickens, turkeys, egg layers, and hogs in open fields, pastures, and woods; some raise fully “grass-fed” beef cattle who do not go through feedlots; and a significant share of milk cows are fed on pasture. These are outside the industrial livestock system: the animals are not confined; much of their feed is grazed or homegrown; and they are raised on fairly small operations with a diversified mix of crops and livestock. However, they constitute small shares of national production. In the aggregate, only 31% of U.S. milk production occurs on farms that graze their animals, and less than 1% of hog and broiler production (Table 2). Relatedly, almost all broiler production, and close to 80% of hog production, is carried out on farms that grow none of their own feed. Few dairy farms raise no feed, but most combine purchased with homegrown.

Table 2. Feeding Choices in U.S. Livestock Sectors


Raise No Homegrown Feed

Do Not Graze Animals

Percent of Farms or Production

Dairy Farms



Milk Production



Hog Farms



Hog Production



Broiler Farms



Broiler Production



Notes: Surveys cover commercial operations only: at least 1,000 broilers, 25 hogs, or 10 milk cows on site at any time during the year. Production is measured by weight of milk produced, weight gain of hogs on finishing operations, and liveweight delivered from broiler operations.

Sources: U.S. Department of Agriculture, Agricultural Resource Management Survey, 2006 (broiler version), 2010 (dairy version), 2015 (hog version).

Livestock production has been shifting to larger farms over time. Farm sizes are highly skewed in U.S. statistics: Many small operations raise a few animals for home consumption, local sales, or show. But most animals are on a relatively small number of large farms. Because of this skewness, simple average measures of herd or flock size are not very informative. Instead, we use the midpoint size: the herd or flock size at which half of all animals are on larger operations and half are on smaller.

Midpoint herd and flock sizes have increased sharply in recent decades as production shifted to larger operations (Table 3). For example, the midpoint dairy herd in 1987 was 80 cows—half of cows were on herds no larger than 80, and half were in herds that were no smaller. As cows and production shifted to larger operations, that midpoint rose sharply over time, reaching 900 cows—more than a tenfold increase—by 2012. Hogs and egg layers also showed striking growth over 25 years. The midpoint for fed cattle stabilized after 1997, and those for broilers and turkeys after 2007, but each underwent major reorganizations and shifts to larger operations during the 1960s and 1970s (Martin, 1979; Rogers, 1979).

Table 3. Consolidation in U.S. Livestock Sectors






Sales Midpoint: Number of Head Sold or Removed in Year






Fed Cattle





Hogs and Pigs










Inventory Midpoint: Number of Head in Herd/Flock

Beef Cows





Egg Layers





Milk Cows





Note: The midpoint is a median that defines the farm size at which half of animals are on larger farms, and half are on smaller farms.

Source: U.S. Department of Agriculture, Census of Agriculture for the years noted

The exception is in beef cow herds, where the midpoint grew only modestly, from 89 cows in 1987 to 110 in 2012. There has been little reorganization of the cow-calf sector and no consolidation of its land base in pasture and rangeland, as land actually shifted away from the largest ranches. There has been little movement toward closer integration of cow-calf operations with cattle feeding and meatpacking through contracts or joint ownership. The cow-calf sector remains the least-industrialized U.S. livestock sector.

Industrialized Livestock Systems in Other Countries

Industrial production systems, encompassing confinement feeding on specialized farms, with purchased feed formulations targeted at animal life cycles, and networks of farms with close commercial ties between growers and processors, are typical of Western European swine operations in major producing areas in Belgium, Denmark, Germany, the Netherlands, and Spain. They are less common in Eastern European swine regions in Poland and Romania, although industries in that region have been adopting industrial processes in recent years. Similarly, most European poultry production follows industrialized processes, concentrated in France, Germany, Italy, and the United Kingdom, while many small-scale producers still operate in Eastern Europe.

Little specialized cattle feeding occurs in Europe, where most beef comes from farms with a primary focus on dairy production. In turn, while some large-scale dairy operations have begun to appear in Europe, and production has shifted away from small herds, most milk still comes from farms that are far smaller than those that dominate U.S. production. The European cattle sector remains much less industrialized than poultry or swine.

Industrialized systems are growing rapidly in Asia, spurred by the emergence of large integrators that control feed and processing operations, while also contracting with independent farmers or operating their own large-scale farms. According to the summaries published in the U.S. trade publication National Hog Farmer, there were seven large Asian firms, five of them in China, that each owned at least 100,000 sows in 2017, and collectively owned 2.68 million sows. Given typical production statistics for large scale operations, the sows owned by these firms could collectively produce nearly 60 million hogs in a year. By contrast, 11 firms in the United States own at least 100,000 sows, according to the annual list provided by the trade publication Successful Farmer, and of course there are many other smaller-scale integrators, with 10,000 or more sows, in North America and in Asia. A typical farm could house 1,000 sows in a barn, and they in turn would produce 22,000–25,000 hogs annually, raised in two to four finishing barns, so a 100,000 sow integrator would require an extensive network of sow, nursery, and finishing facilities, either owned or operated under contract.

While there are still many small-scale village-based Asian poultry producers, integrated poultry producers—with networks of feed mills, processing plants, breeding farms, hatcheries, and contract growers—now account for most Asian production. They tend to contract with farmers—who usually operate on a much smaller scale than in the United States—and provide them with feed and chicks, who are raised in enclosed facilities following company-provided guidelines.

Large industrialized dairy farms, often owned or financed by global corporations, have also begun to appear in Asia. For example, the largest dairy business in China, China Modern Dairy, operates 26 farms with a combined total of 230,000 cows, while the New Zealand cooperative Fonterra has 35,000 cows in farms at two Chinese “dairy hubs.” In Vietnam, TH Milk operates farms with a total of 45,000 cows.

Research, Information, and the Spread of Industrialized Production

The expansion of industrialized production systems, in the United States and around the world, relies on a set of related developments in animal genetics and breeding, nutrition and feed formulations, housing, and management practices. In breeding, new selection methods have allowed for greater accuracy in the prediction of the breeding value of animals, spurring more rapid dissemination of high-quality breeding stock. The use of artificial insemination, embryo transfer, sexing, and the identification of genetic markers has accelerated the rate of multiplication of the best genotypes, leading to steady improvements in milk yields (de Haan, Gerber, & Opio, 2010).

Industrial production has benefited from decades of research on feed formulations, in which rations that combine amino acids and minerals with feed grains and oil crops are optimized to the animal’s stage of development for faster and more efficient growth and maintenance. Developments in animal housing have focused on continuing incremental improvements in ventilation, climate controls, feed provision, manure handling, and building security to limit the spread of animal diseases and foster more rapid and efficient growth. The development and spread of various management practices have led to improved biosecurity on facilities—thereby contributing to the control of disease—while also leading to improvements in feed absorption and production efficiency.

Each of these developments requires investment and research leading to innovations, but they also require significant investments in standard operating procedures, training, and management protocols to consistently apply the innovations to breeding and production environments. Some of these tasks may be carried out by specialized firms—such as animal genetics companies that sell genetic material and breeding animals to producers—or they may be carried out by large integrated operations who encompass research, feed production, breeding, livestock production, and processing stages. Once firms develop these knowledge assets, they can be replicated on many different production units and environments, which facilitates the emergence of large integrators with multiple production sites.

Why Industrialize? The Economics of Industrial Livestock Production

Industrialized systems offer substantial improvements in productivity and concomitant reductions in resource use and production costs. Productivity improvements arise from better feed conversion, reduced animal mortality, greater rates of multiplication from breeding animals, and reduced requirements of labor and capital for a given amount of meat and dairy production.

In developing countries, lower agricultural production costs resulting from productivity improvements can lead to lower food prices and improved real incomes for food consumers. In the United States and in other developed countries, agriculture accounts for small shares of total food expenditures (processing, retailing, and transportation account for the bulk of food costs). In turn, food accounts for small shares of total consumer spending, so reduced agricultural production costs associated with productivity growth have quite modest effects on food prices and consumer incomes. Instead, the larger impact of agricultural productivity growth arises from freeing land, labor, energy, and capital resources for use elsewhere in the economy.

In 1955, when industrialized broiler production was just starting to expand, the average U.S. bird weighed 3.1 pounds and took 73 days to raise while consuming 8.84 pounds of feed, which is a feed conversion ratio of 2.85. That was a substantial improvement from 1945, when average feed conversion for broilers stood at 4.6 (Rogers, 1979; Reimund, Martin, & Moore, 1981). By 2011, after years of innovations in broiler genetics, feed formulations, housing, and production practices, it took 38 days to produce a comparable bird, and feed conversion had improved to 1.74 pounds of feed per pound of weight gain (MacDonald, 2014).

By that time, growers were producing birds in a range of sizes, from 3 to 9 pounds and with continuing improvements in feed conversion, production cycles, and mortality. Reductions in production cycles also saved capital, since more birds could be produced within a given space over the course of a year. More intense utilization of housing, greater reliance on automated equipment, and improvements in broiler genetics also supported large increases in labor productivity: liveweight broiler production per hour of labor grew at an annual rate of 5.5% between 1955 and 2011, from 78 pounds per labor hour to 1,573.

U.S. hog production underwent a dramatic transformation toward greater industrialization in a relatively short period of time in the 1990s and 2000s, featuring much larger operations, greater specialization by stage of production, reduced reliance on homegrown feeds, and tighter vertical coordination by integrators. McBride and Key (2013) identified the striking impacts of that transformation of industry productivity between 1992 and 2009, with a focus on finishing operations. There, average feed conversion ratios (pounds of feed per pound of liveweight gain) fell from 3.83 to 2.07, while labor productivity (liveweight pounds gained per hour of labor) rose from 112 to 833. Overall growth in total factor productivity (accounting for labor, capital, and feed inputs) averaged 4.9% per year between 1992 and 2009, compared to annual productivity growth of 1.0% per year in the U.S. nonfarm private business sector, according to the U.S. Bureau of Labor Statistics.

Economies of scale in confined feeding systems favor larger operations. McBride and Key (2013) found evidence of substantial unexploited scale economies in hog feeding in 1992, when most production occurred on farms producing fewer than 1,500 hogs per year; by 2009, available scale economies appeared to be fully exploited as almost all production has shifted to farms producing at least 5,000 hogs per year. In dairy production, Mosheim and Lovell (2009) found evidence of large gains from scale economies up to herds of 500 head—and continuing modest scale economies up to 3,000 head. That study examined U.S. herds in 2000, and after that production continued to shift rapidly to much larger herds, supporting their finding of scale economies that had not been fully exploited (MacDonald, Cessna, & Mosheim, 2016).

Most U.S. broiler production is carried out on relatively small farms; nevertheless, production has steadily shifted toward larger operations, with most broilers raised on farms with at least five broiler houses, compared to two or three smaller houses in earlier decades (MacDonald, 2014). MacDonald and Wang (2011) found evidence of modest scale economies in broiler production, with additional productivity improvements associated with the use of automated ventilation and climate control systems.

Most industrial production occurs in tightly coordinated systems, and complementarities among system components can contribute to the realization of production economies. For example, broiler production occurs in regional complexes of feed mills, broiler breeder farms, hatcheries, contract grow-out farms and processing plants (90% of U.S. broilers are produced within 60 miles of a processing plant, according to MacDonald, 2014). Poultry processing is subject to economies of scale, and also realizes economies from slaughtering uniform birds. Integrators manage the production network to assure a steady flow of birds, of common sizes during a given production period, so as to minimize system-wide costs (MacDonald, 2014; Ollinger, MacDonald, & Madison, 2000). Similarly, hog processing is subject to economies of scale and of uniformity, and integrators manage the system to realize steady flows of uniform animals though processing plants. Even in beef production, where there is little ownership integration between cattle feeding and meatpacking, most feedlots work with a limited number of packing plants. Packers maintain the steady flows of cattle needed to realize economies of scale in packing through the extensive use of marketing contracts with cattle feeders (MacDonald, Ollinger, Nelson, & Handy, 2000).

In turn, meat processors sell products through four major channels: supermarkets, food service buyers such as restaurant chains and institutions, wholesalers who distribute to smaller domestic buyers, and export channels. Each channel features large flows of meat products, with contractual commitments to provide assurance to processors against unexpected fluctuations in demand. With large-volume buyers of uniform products, processors have the assurance needed to invest in large-scale processing facilities, while integrators and contract growers have the assurance needed to invest in capital-intensive production facilities. The close integration of production—through vertical integration as well as contractual relations that bind different links of the system in high-volume and predictable—facilitates productivity-improving investments.

Social Costs of Industrial Livestock Production

Livestock production generates social costs, not reflected in the costs borne by producers or the prices paid by consumers but nonetheless imposed on society, and the high stocking densities that are typical of industrialized systems can exacerbate them. In particular, the social costs include potential harm to animals’ welfare, environmental damages from excessive concentrations of manure, and risks to human health from the overuse of antibiotic drugs.

Animal Welfare in Industrialized Production

Animal welfare concerns have emerged as an important issue in recent years, particularly with regard to agriculture. Contemporary concerns were initiated in the mid-1960s, with the publication of a book (Harrison, 1964) and follow-up government report in the United Kingdom. Those publications spurred the development of a framework for evaluating animal welfare, encompassing “five freedoms” for animals, which have been adopted by a wide range of animal welfare organizations as well as producer groups. They include:

  • freedom from hunger and thirst.

  • freedom from discomfort.

  • freedom from pain, injury, or disease.

  • freedom to express normal behavior.

  • freedom from fear and distress.

The original publication (Harrison, 1964) was primarily concerned with industrialized livestock production. Animal welfare harms are not unique to industrialized production, but the high density of animals in confined feeding environments, and the practices adopted to manage them, have led to several specific concerns (Pew Commission on Industrial Animal Production, 2008). For example, animals may be physically altered without pain relief, even when such alteration causes pain: this is to prevent them from injuring one another in close environments. Hogs may have their tails docked to prevent tail biting by other hogs; dairy cows may have their tails docked and their horns removed; and laying hens and broilers may have their spurs and beaks clipped.

During pregnancy and after parturition, sows have generally been physically constrained in metal enclosures called gestation crates to prevent fighting among sows and to protect piglets who can be crushed when sows roll over. Standard crates measure 2 meters by 60 centimeters, small enough to limit the movements of a 350-pound sow; they have no bedding material but are instead floored with slatted plastic or metal to allow wastes to be gathered below the crates. Critics argue that crates sharply limit the natural behaviors that sows would engage in the wild and impose significant health and stress risks on the animals.

Egg-laying hens are usually confined in linked batteries of metal cages, tiered in pyramids to facilitate the collection of eggs and removal of wastes. A single cage usually contains 6 to 12 birds. Critics contend that birds confined to battery cages are unable to engage in many natural activities, such as wing stretching and flapping, preening, feather ruffling, and ground scratching, and that the animals suffer from distress and boredom when they are not able to engage in these behaviors.

Cattle that are finished in feedlots are fed grains rather than forage. A grain-based diet allows them to put on weight faster, but since their digestive systems are designed to metabolize forage diets and not grain they often experience internal abscesses and discomfort.

Producers often argue that market forces provide them with strong incentives to account for the welfare of the animals on their operations. High herd or flock mortality directly costs producers, since they lose the value of those animals and their production, and animals suffering from stress or disease produce low-quality meat and lower volumes of milk or eggs. Producers therefore have every reason, they argue, to ensure that animals are healthy, calm, and stress free. However, the level of animal welfare needed to realize the maximum in profits from a farm operation does not necessarily preclude overcrowding, discomfort, pain, or stress (Lusk & Norwood, 2011). Improvements in animal welfare require expenses, and unless producers see offsetting gains in revenue from higher prices paid for meat or dairy products, they have little incentive to undertake those expenses beyond that necessary to realize target levels of production and quality.

Public and Private Regulatory Initiatives Regarding Animal Welfare

Public concerns with farm animal welfare have been expressed through three channels. Some governments have introduced legislative initiatives aimed at regulating some farm practices. In addition, producer groups have introduced guidelines for members aimed at improving animal welfare. Finally, food retailers and processors have developed guidelines and labeling aimed at identifying products produced under certain welfare standards.

Federal laws in the United States are primarily focused on slaughter and transportation to plants—not on farm practices. However, a recent government action in California concerning space requirements for egg-laying hens carries several important insights for regulation of animal welfare (Sumner, 2018). In 2008, California voters approved a ballot measure requiring greater space standards in housing for egg-laying hens in the state. Egg production costs are affected by housing design, and increased space per bird would likely lead to higher labor and capital costs: Mathews and Sumner (2015) estimated that operating costs in conventional systems, per dozen eggs produced, would be 4% lower than in enriched colony houses, and 19% lower than in aviary houses. When capital costs are included the differences are 11% and 27% for enriched colony and aviary systems, respectively.

By itself, given the cost differences, the measure likely would have led to a shift of egg production from California to other states. Out-of-state producers would then ship eggs to California retailers, thus disadvantaging in-state compared to out-of-state producers with no ultimate effect on production practices. To address this issue, the state legislature passed a law in 2010 that applied the ballot measure’s rules to eggs sold in California, even if the hens were housed elsewhere. Then, in 2013, the states issued regulations specifying the actual space requirements to apply under both laws when implemented at the beginning of 2015. Throughout this period, there was considerable uncertainty regarding the actual space standards, uncertainty that continues because of continuing litigation. Uncertainty matters because it affects capital investment in long-lived houses and equipment.

The economic impact of the regulations were substantial: retail egg prices in California rose by one third in the year of implementation (2015) and were one-fifth higher in 2016 than they had been before implementation, in 2014, shifts that were broadly consistent with projected production cost increases (Mathews & Sumner, 2015; Mullaly & Lusk, 2018). Moreover, California egg production in 2015 and 2016 fell by about one third, compared to its pre-ban levels, despite the efforts to equalize the impact on California and out-of-state producers (Mullaly & Lusk, 2018). Sumner (2018) argues that the uncertainty surrounding the California rules led to reduced investment in production facilities in California, thus leading to rising California operating costs and a widening gap in production costs between California and other states. In this case, the economic impacts arose not only from the differences in costs between systems but also from the process of implementing the rules.

Animal welfare concerns have been more widely expressed through voluntary industry guidelines, such as those offered by the United Egg Producers, the Pork Board, and the National Chicken Council in the U.S. Retail chains—both supermarkets and restaurants—have also begun to set standards for producers in their supply chains, in some cases with labels specifying certain levels of animal welfare standards and with certification from third parties. If effective, such standards and labels can be a way to channel consumer demands for improved animal welfare and in that case can provide incentives for producers to adopt welfare-improving practices.

The standards cover a mix of targets (Lagerkvist, Hansson, Hess, & Hoffman, 2011). Some specify resource-based measures such as space per animal, flooring type, feed composition, or light; these are fairly easy to measure but do not directly monitor animal welfare. Others are management-based routines, which focus on documentation of feeding routines, treatment protocols for sick animals, or the absence of objectionable practices; like resource-based measures, they can be based on existing plant information, and therefore are easy to document, but they don’t directly monitor animal welfare.

Environmental Risks From Industrialized Livestock Production

Livestock production generates animal wastes—combinations of manure, urine, and bedding material. Wastes may be stored in pits beneath animal housing, with the wastes dropping through slatted floors, or tanks and lagoons located near housing, to which wastes are moved throughout the day. In poultry grow-out farms, wastes accumulate on the dirt floors of houses, which are cleaned out and sanitized periodically after flock removal. When removed from storage, wastes may be spread on farm fields to provide crop nutrients, or they may be shipped off the farm to nearby farms or to processing uses.

Pollution arises from the runoff of manure nutrients, organic matter, and pathogens to surface water; from leaching of manure nitrogen and pathogens to ground water; from volatilization of manure gases and odors to the atmosphere; and from emissions of fine particulates (Aillery et al., 2005). Each of these sources can impose economic damage on other users of water and air resources. They can also create human health risks (Sneeringer, 2009). Pollutants can originate from production houses, storage structures, or from land where manure is applied. Manure that is applied in excess of the absorptive agronomic capacity of cropland can exacerbate the risks of nutrient runoff and leaching.

Industrial production systems affect these environmental risks in three ways. The improvements in feed conversion associated with industrial production reduce the total amount of manure created in livestock production, for any given amount of meat or dairy production, thereby reducing the amount to be disposed of (Capper, Cady, & Bauman, 2009; Capper, 2011). However, industrial operations feature larger herds and flocks and concentrate more production and more manure within a farm’s land area, thereby increasing the likelihood of runoff, leaching, volatilization, and air emissions (Figure 2). In the United States, farms with fewer than 300 animal units operate with about two animal units per acre of land receiving manure (an animal unit, used to compare across species and stages of production, is equal to 1,000 pounds liveweight onsite in inventory). Stocking densities increase with herd size, and farms with at least 1,000 animal units on site operate with densities ranging from five animal units per acre (broilers and dairy) to 12 (hogs).

CAFOs: Farm Animals and Industrialized Livestock Production

Figure 2. Larger U.S. farms have more animal units per acre of land receiving manure.

Source: Agricultural Resource Management Survey, 2009 (Hogs), 2010 (Diary), 2011 (Broilers).

Moreover, the emergence of industrialized systems in the United States has also consolidated livestock and manure production within more concentrated geographic areas, so that manure stocks may substantially exceed the absorptive capacity of all of the cropland in the region surrounding a network of industrialized farms. For example, if one sorted U.S. counties according to pig production, just 81 accounted for half of all production in 2007, compared to 251 counties in 1969. Egg-laying, fed cattle, and dairy showed similar patterns of geographic concentration (McBride, 1997; O’Donoghue et al., 2011). In each case, while industrialization reduced the amount of waste generated per animal, it also concentrated those wastes in smaller geographic areas (Sneeringer, 2009).

Using detailed U.S. farm-level records on livestock inventories, crop and pasture land, and crop mix, Gollehon et al. (2001) evaluated the degree to which the amount of manure nutrients produced exceeded the absorptive capacity of the available land base. Using data from 1997—shortly before the introduction of tighter environmental regulations for industrialized operations—they estimated that 78% of U.S. animal feeding operations had sufficient crop and pastureland to use all manure nitrogen produced on the farm (69% of animal operations could utilize all phosphorus). However, the farms that did not have sufficient land for nutrient absorption, who were typically large industrialized operations that accounted for most livestock production, produced 60% of total manure nitrogen and 70% of total manure phosphorus.

This finding does not conclusively show that manure actually is over-applied, because manure can be removed from animal feeding operations and spread on nearby crop farms—nationally, about 20% of dairy manure, 25% of hog manure, and 60% of poultry litter is removed from the farm that generates it (MacDonald & McBride, 2009). To account for removal, Gollehon et al. (2001) then expanded their analysis to consider the absorptive capacities of cropland with the same county as a livestock operation (manure is costly to transport, and thus the county is a relevant localized area). Many of those farms were located in counties with sufficient cropland—located on specialized crop farms—to absorb the excess manure nutrients produced on large livestock feeding operations. However, about 20% of all manure nitrogen, and 23% of phosphorus, were produced in counties that had insufficient crop and pastureland throughout the county to absorb the nutrients at agronomic rates, even if all excess nutrients could be removed from feeding operations and shipped to a willing user elsewhere in in the county.

In follow-up work, Gollehon, Kellogg, and Moffitt (2016) found that the number of counties with excess manure increased sharply over time, from 82 in 1997 to 205 in 2012, in line with the continued shift of production to larger animal feeding operations (Table 3). The total quantities of county-level excess nitrogen and phosphorus roughly tripled in the decade between 1997 and 2007, and then stabilized between 2007 and 2012.

The calculations used so far are rather mechanical applications of stocking densities. Actual application rates of manure nutrients to cropland reflect several choices that farmers make. Farmers can adjust manure applications on their land by altering the total amount of crop and pastureland on the farm, the share of their land that receives manure, or the amount of manure that is removed from the farm for spreading elsewhere. Farmers can alter the absorptive capacity of their cropland by altering their crop mix, toward crops with greater nutrient utilization, and by adjusting the amount of synthetic fertilizers applied to their land. They can also affect manure production by adjusting herd sizes and feed formulations, or by providing feed additives designed to promote better absorption of nutrients in livestock.

Some excess manure nutrients could be shipped further—out of the county—and some could be applied to uses other than crops, such as feedstock for electricity generation. However, long-distance transport is generally quite costly, and finding willing users of manure, even in counties with enough absorption capacity, can be a challenging task because the nutrients in manure may not match to the needs of the crop producer. In that case, the county-based estimate likely understates the amount of excess manure nutrients actually present.

Regulation of Manure for Environmental Protection

The Netherlands provides an example of the evolving nature of manure regulation, particularly as policy has also featured economic incentives to reduce excess manure loadings (Wossink, 2004; Schroder & Neeteson, 2008; Van Grinsven, Tiktak, & Rougoor, 2016). The densely populated country has a large and industrialized livestock feeding sector, often located on relatively small plots of land, among abundant surface and groundwater sources. By the mid-1980s a substantial surplus of manure production over the assimilative capacity of cropland had developed, and nitrate levels in groundwater had risen above EU standards in most agricultural land in the Netherlands, with particularly challenging problems in the sandy soils of the eastern and southern parts of the country.

In 1984 the Dutch government introduced a moratorium on the creation of new livestock farms in some regions and placed restrictions on the expansion of existing farms. This was followed by a more comprehensive approach aimed at achieving nutrient balance by 2000, with a focus on the phosphorus content in manure. The law introduced a manure bookkeeping system and, by 1994, a system of transferable manure production rights. Actual manure production on each farm was estimated from the farm’s inventory of animals, combined with standards for manure and phosphorus production for animals of a given species. The assimilative capacity of each farm was assessed through evaluation of the amount and type of land owned or under long-term lease. Farms with manure surpluses had to acquire more land, or reduce livestock herds, while farms with manure deficits could increase livestock production.

In 1994, quota rights became tradeable, subject to a set of constraints on interregional and interspecies trades (Wossink, 2004). One goal of tradeable quotas is to provide incentives for livestock production to shift from manure-surplus regions to manure-deficit regions; another is to provide incentives for individual farms to reduce manure production, since unused quota rights could be sold; finally, farms with high manure surpluses might exit production.

Nutrient balances decreased significantly after 1990, with a relatively high rate of decrease high during 1998–2005, under the MINAS mineral accounting system, and there is also evidence of improving water quality (Van Grinsven et al., 2016). Swine production expanded in manure-deficit parts of the country; many quota sellers exited the industry; and the introduction of newer feed formulations reduced the phosphorus content of feed. Each of these outcomes provides evidence of the impact of economic incentives on producer behavior.

In 2003, the European Court of Justice ruled that the Netherlands program was in conflict with the EU nitrates directive; the regulation had no specified application rates for nitrogen, since it focused on nutrient balances; the regulatory framework was reproached for implicitly allowing application rates higher than a formal threshold without possessing formal permission from the European Commission; and there were no explicit linkages between surpluses and water quality (Schroder & Neeteson, 2008). Current Dutch policy imposes a system of soil- and crop-specific application standards for nitrogen and phosphorus, including statutory equivalences for N in manure (Van Grinsven et al., 2016). Farmers with excess manure must pay to have it removed to other cropland, and the expense is high enough (15 to 25 euros per ton) to provide an economic incentive to conserve nutrients. Measures of nutrient balances, nitrates in water, and ammonia concentrations in air have continued to improve under the more recent policy initiative.

In the United States, the Environmental Protection Agency (EPA) regulates manure management practices at certain livestock operations under provisions of the 1972 Clean Water Act. Those provisions were strengthened under regulations issued by the EPA in 2003 and finalized in 2008, and are commonly known collectively as the “CAFO Rule.” The regulations require farms to seek permits if they have been designated as a concentrated animal feeding operation (CAFO), and have documented discharges of manure effluent. In turn, the permits require the adoption of practices, including a nutrient accounting system, aimed at reducing the environmental impacts of manure production, storage, and application.

The rule defines an animal feeding operation (AFO) as a facility that confines animals for at least 45 days during a year and grows no vegetation where animals are raised and manure is stored. All large AFOs—with, for example, at least 700 dairy cows, 1,000 other cattle, 2,500 swine of at least 55 pounds, or 125,000 chickens on site—are designated as CAFOs. Medium size AFOs—with, for example, 37,500 to 125,000 chickens or 200 to 699 dairy cows onsite—are designated as CAFOs if they also discharge manure nutrients via a manmade conveyance or if animals at the operation come into contact with federally regulated waters. Small AFOs may be designated as CAFOs at the discretion of the enforcement agencies in US states, depending on discharges and other factors.

Over time, confined livestock feeding has concentrated onto larger operations (Figure 3). Gollehon et al. (2016) estimate that about 3,700 US farms would have been classified as large AFOs in 1982 had the EPA rules been in effect in that year and that they accounted for 24% of all animal units in AFOs; by 2012, 12,600 farms would be classified as large AFOs, and they accounted for 62% of all animals units in AFOs.

CAFOs: Farm Animals and Industrialized Livestock Production

Figure 3. By 2012, most U.S. confined livestock feeding was in large AFO’s.

Source: Gollehon et al. (2016), derived from census of agriculture. AFO’s are animal feeding operations.

CAFO permits specify requisites for the livestock production and land application areas of farms. The production area must function so that it can contain animal wastes inclusive of precipitation from rare, large storms. With regard to the land application area, the permit requires implementation of a comprehensive nutrient management plan (CNMP) with specific guidelines for setback distances, minimizing nutrient runoff, measuring crop nutrient uptake, sampling of manure and soil, setting manure application target rates based on manure nutrient sampling and estimated crop nutrient uptakes, and treatment and transfer of manure. It should be noted that permits, and the guidelines embodied in them, apply only to the land controlled by the CAFO, and not to any crop farms that receive manure from the CAFO.

The Federal CAFO regulation sets minimum standards for regulated entities; states enforce the federal laws and may impose further regulations on their own. In addition, some state and local governments have pursued lawsuits against farms for damages under the Clean Water Act and have reached settlements under which the farms agree to take actions to further limit nutrient discharges.

CAFO regulations have led to behavioral changes among industrialized operations. Key et al. (2011), relying on surveys conducted by the U.S. Department of Agriculture, find that large hog producers (more than 1,000 animal units) had more cropland, and applied manure to more acres, in 2009 than they had in 1998, prior to issuance of the CAFO rule. Their average application intensity (animal units per acre with manure applied) fell by one third from its average level in 1998 between that year and 2009. Small and medium AFOs, less likely to be regulated under the CAFO rule, showed no significant change in acres or in application rates, which suggests that the rule drove practice changes among larger operations.

Larger hog operations increased their manure removals between 1998 and 2009, while small and medium operations did not. Larger operations increased their use of microbial phytase, an additive in finishing hog diets that aids the absorption of organic phosphorus, thus reducing use of mineral phosphorus and reducing phosphorus excretion in manure. Each of these steps would be expected if the rules were constraining large regulated operations.

Using farm-level records on broilers from the same USDA survey, MacDonald (2014) found that application rates (broilers per acre of manure applied) fell between 2006 and 2011 for growers with a CNMP, but rose for growers without one, such that growers without plans had per-acre manure application rates that were 63% higher than those with plans in 2011.

Finally, Sneeringer and Key (2011) found that CAFO regulation affected farm size choices. Specifically, recall that farms above a certain size threshold are more likely to be regulated under the CAFO rule than firms below the size threshold, and that the rule was strengthened in 2003. Sneeringer and Key examined the size distribution of new hog operations who entered the business between 2002 and 2007 and compared it to farms that entered between 1997 and 2002. They found a distinct bulge of new entrants just below the size threshold in the later period, when there had been no such bulge in the earlier period. Farms who could make a size choice leaned, after the rule’s issuance, to enter at a size just below the threshold, thus avoiding regulation.

CAFO rules and civil lawsuits aim to compel changes in farm practices through regulation and compliance. Other federal and state policies provide subsidies to induce improved manure management practices and alternative uses. For example, under the Environmental Quality Incentives Program (EQIP), a major conservation tool, the federal government shares some of the costs of practices aimed at reducing nutrient run-off on farms. Under the program, farmers propose practices that they intend to implement under a contract with USDA, along with a proposed cost-sharing formula (to be made by USDA); local committees meet under the aegis of USDA and select proposals based on the likely environmental benefits and the costs to the government.

Manure can also be used as a feedstock in power plants for energy generation, and anaerobic digesters can be used, usually on farm, to capture the methane gas from manure and use it as a feedstock for on-farm electricity generation (in digestion processes, lagoons are covered to help capture the methane, and then manure liquids are separated from solids, with the liquids sprayed on fields for nutrients and the solids processed and sold as animal bedding or garden soil supplements). Manure-based energy generation is relatively costly, and is unlikely to be carried out without subsidies, which are provided by some US states either directly as financial subsidies or indirectly through mandating utilities to originate some of their power needs from alternative energy sources (Key & Sneeringer, 2011).

Use of Antibiotic Drugs in Industrial Agriculture

Antibiotic drugs are used for the treatment of animal diseases, and they are provided in feed or water to control and prevent transmission of disease among herds and flocks. Antibiotics have also been used for “production purposes,” because they have been found to improve the efficiency with which feed is converted to weight gain and to reduce the time needed for animals to reach market weights. When used for disease treatment, control, and prevention, antibiotic drugs improve productivity growth in livestock agriculture by reducing animal mortality and morbidity; drugs administered for production uses also contribute to productivity growth by reducing the amount of feed and housing required for any given amount of production.

The impact of antibiotic provision on feed conversion (a production purpose) appears to have declined over the years (Sneeringer, MacDonald, Key, McBride, & Mathews, 2015). The decline may reflect growing pathogen resistance to drugs, making them less effective. However, it may also reflect changes in animal genetics, feeding formulations, housing, and production practices that reduce the incidence and spread of pathogens in herds and flocks, and hence can also reduce the effectiveness of antibiotics in suppressing pathogens and facilitating improved feed conversion.

While used widely, antibiotic drugs have not been used universally in U.S. production (Sneeringer et al., 2015). While almost all hog and broiler production can be said to be carried out in industrial systems, significant amounts of production appear to be done without antibiotics for disease treatment or for production purposes, and there is no clear association between use and the size of farm operations.

Antibiotics kill a wide range of pathogenic bacteria that harm humans and animals and have provided enormous health benefits since the 1940s. However, they are losing their effectiveness to treat human illnesses because bacteria are evolving resistance to them. Use of antimicrobial drugs creates selective evolutionary pressure that enables antimicrobial-resistant bacteria to multiply more rapidly than antimicrobial-susceptible bacteria and thus increases the opportunity for individuals to become infected by resistant bacteria. Resistance to specific antibiotic drugs can spread among bacteria, jump from one type of bacteria to another, and move across regions, elements that create considerable scientific uncertainty in tracking resistance back to specific locations or practices.

Growing resistance follows in part from the widespread use of antibiotics in both human and animal medicine. While the scientific uncertainty surrounding the spread of resistance precludes accurate estimation of precise shares of responsibility arising from human or animal use, there are some clear linkages from animal use to human resistance (Sneeringer et al., 2015). There is widespread concern that growing global poultry and pork production, particularly in the developing world, will lead to an acceleration of global antibiotic use, and that expanded use may lead to a more rapid spread of pathogen resistance to antibiotics, especially those that are used in both human and animal medicine.

Individual decisions to use antibiotics, whether prescribed by a doctor to treat a cold or sore throat that will get better without drugs, or administered by a farmer to promote more rapid weight gain, are generally carried out without regard to the costs that may be imposed on others through the impact on antimicrobial resistance. On the other hand, a decision to forego an antibiotic carries some risks for the person going without, while the benefits, in terms of a reduced spread of resistance, go to others. In short, antibiotic use creates uncompensated costs for others, which have little influence on decision makers’ calculations. This is a classic instance of externalities, under which private markets are likely to induce overuse of antibiotic drugs.

Regulation of Antibiotic Use in Agriculture

The United States and the European Union (EU) have each taken steps to regulate antibiotic use in agriculture in recent years but by following sharply different paths. The EU has banned the use of specific drugs for growth promotion purposes. In the early 1970s, the United Kingdom banned the use of penicillin and tetracycline for growth promotion. Other European countries followed with bans of specific drugs in the 1980s and 1990s; Sweden banned all antibiotic growth promoters in 1986, while Norway, Poland, Finland, and Switzerland banned over-the counter sales of antibiotics in feed. An EU ban on all antibiotics used for growth promotion was promulgated in 2006.

Denmark’s experience with limiting antibiotic use has been well-documented (Aarestrup, Jensen, Emborg, Jacobsen, & Wegener, 2010; Jensen and Hayes, 2014), in part because the country is a major swine producer and because Denmark has invested in a significant program of tracking and data collection: the Danish Antimicrobial Resistance Monitoring and Research Program (DANMAP). Denmark has highly efficient and industrialized pork and poultry sectors; pork is much larger, accounting for about 80% of antibiotics use and has been the focus of most analysis.

Denmark was among the first countries to ban the use of antimicrobials for growth promotion (between 1995 and 1998) and in 2000 banned the use of certain drugs used for disease prevention among nursery pigs. In 2010, Denmark began to pursue a more aggressive program. In that year, the industry agreed to a voluntary ban on the use of cephalosporin drugs (which are also used in human medicines) to treat respiratory problems in young pigs. The country also introduced a scheme to monitor antibiotic usage at individual farms and to issue warnings to farms with above average usage. Failure to reduce above-threshold usage would result in additional monitoring, paid for by the producer, and ultimately a requirement to reduce herd size.

After the bans on growth-promoting drugs, total consumption of antibiotics in Denmark fell by half, reflecting a cessation of use for growth promotion but an increase in use for disease treatment (Jensen & Hayes, 2014). Total use continued to grow during the 2000s, but in line with total Danish swine production. There did not appear to be any significant impacts on animal health from the cessation of growth-promoting antibiotics. Producers adjusted by using other non-antibiotic feed additives, shifting to greater use of vaccinations, investing in improved housing and changing certain production practices.

Later initiatives have proved more challenging (Jensen & Hayes, 2014). There has been increased incidence of animal health problems, particularly among nursery pigs, in the wake of limits on the use of antibiotics for disease prevention; and there has been an increase in drug use for disease treatment. The incidence of resistance to antibiotics does not appear to have declined in recent years, although establishing links from livestock restrictions to the incidence of resistance is fraught with great uncertainty, given the range of paths through which resistance can spread.

In the United States, there is growing consumer and retailer interest in meat products that have been raised without antibiotics, and their actions can affect antibiotic drug use irrespective of changes in federal policies. However, such products must be advertised and labeled in order to influence consumer behavior. Many consumers appear to be willing to pay price premiums for products labeled as “raised without antibiotics” (implying no antibiotics provided for production purposes) and for products labeled as “never administered antibiotics” (implying no antibiotics administered for any purpose). However, because consumers have no way of directly ascertaining the truthfulness of claims regarding antimicrobial drug use (or other on-farm production practices), firms could have strong incentives to issue misleading or inaccurate claims. Without customer assurance, markets are unlikely to effectively transmit consumer preferences to producers.

The USDA has responsibility for approving label claims, including claims regarding antibiotic usage, and also (through a different USDA agency) provides process verification services for animals raised according to certain specified production practices (including processes that forego antibiotics). If consumer preferences are to affect farm practices, USDA must define the substances that are to be classified as antibiotics in label claims and define precisely what specific claims mean in informative and useful ways for consumers.

The United States also has regulatory processes in place concerning antibiotic use. The Food and Drug Administration (FDA) has responsibility for approval of drugs used in human or animal medicine. As part of the responsibility, the agency is also charged with ensuring that antimicrobial drugs are used judiciously so as to slow the development of resistance. In pursuit of that goal, the agency introduced two important regulations (“guidances” in FDA parlance), beginning in 2017, in the marketing of medically important antimicrobial drugs used in livestock production. The agency classifies specific drugs as “medically important” based on their microbiological effects on bacteria of human health concern—that is, medically important to human medicine.

Since January, 2017, the agency no longer allows medically important drugs to carry label claims for growth promotion or production purposes; drugs can only be legally used in livestock production for disease treatment, control, or prevention (therapeutic purposes). Drugs that were previously labeled for use for multiple indications (e.g., growth promotion and disease prevention) can still be marketed but without the production purpose label claim. Hence, those drugs may still contribute to faster and more efficient weight gain, even if used for disease prevention. It therefore remains to be seen whether the FDA rescission of approval for production purposes will have a substantive impact on consumption of medically important antibiotics; if livestock operations simply shift to use for stated disease prevention purposes from use for production purposes, the effect on drug consumption will be minimal.

However, domestic sales of medically important antibiotic drugs sold for production purposes, or for combined production and therapeutic purposes, fell by 17% in 2016 (by weight), while sales of medically important drugs sold only for therapeutic purposes fell by 7% and sales of antibiotics classed as not medically important fell by 4% (U.S. Food and Drug Administration, 2017). These declines in 2016 follow on steady and substantial annual sales growth in each category extending back to 2009, when the FDA first began reporting this data. The sharper decline in medically important drugs used for production purposes likely reflects actions taken by feeding operations in advance of the 2017 ban, but it may also reflect important ongoing reductions in antibiotic use realized in poultry production over time (Sneeringer et al., 2015; MacDonald, 2014).

The second FDA regulation moves all medically important drugs used in livestock production to veterinary oversight. This is a substantial change, as 96% of such drugs (by weight) were distributed via over-the-counter channels in 2016 (U.S. Food and Drug Administration, 2017). Under the new guidance, medically important antibiotic drugs must be distributed under Rx prescription (for drugs delivered in water) or Veterinary Feed Directive (VFD, for products delivered in feed). By placing medically important drugs under veterinary oversight, and by adding programs of education and information on antibiotic resistance for veterinarians, the FDA hopes to further limit the use of medically important drugs in livestock production.


The industrialization of livestock production has led to substantial reductions in the feed, labor, capital, and energy required to produce animal products and has thereby contributed to improved real incomes in developing countries through lower food prices and expanded choices, as well as to a smaller footprint of agriculture on resource use in developed countries. However, industrialization also exacerbates several of the important social costs associated with animal agriculture. This review focuses on impacts of industrial production systems on animal welfare, air and water quality, and the resistance of human pathogens to antibiotics. Each represents a challenging scientific problem, and each has been the object of important public policy initiatives, with many unresolved research issues attending the science and the policy impacts.

We can identify desired outcomes for animal welfare, for antibiotic resistance, and for air or water quality. However, the connections between those outcomes and observed farm practices is often quite murky, in part because the science may be unsettled or difficult and in part because we lack adequate measures of practices and of outcomes.

Governments have introduced a range of public policy initiatives aimed at ameliorating the effects of livestock agriculture on the environment and on resistance to antibiotics. This review has touched on the United States and on two EU countries with innovative approaches. There are significant differences in approaches between the United States and the EU countries, and there are still wider differences when all developed countries are considered. There is still only a limited literature reviewing the effects of different regulatory initiatives on farm-level practices, on the outcomes of those practices, and on the aggregate impacts on the countries.

Countries also use a variety of economic incentives to induce changes in farm-level behavior. Some are embedded in private markets and strategies, as when retailers and processors use labels to indicate the use of certain farm practices, certified by government or independent third party agencies, in an attempt to use consumer demand acting through market forces to improve animal welfare, resource sustainability, or pathogen resistance. While there is great interest in these initiatives, there is still a great deal to be done to understand their design and impacts.

Other economic incentives have been built into public regulatory strategies, including the use of tradeable permits or quotas, subsidies for certain farm practices, or taxes on undesirable activities or products. There’s still a great deal to be learned about the proper design of incentives and their impact on farm behavior and outcomes.

Finally, industrialized production is spreading rapidly in the developing world, certainly in Asia but also in South America. There has been significant research on the attributes, finances, and production impacts of industrialized production systems in the United States and Western Europe, but much less is known about precisely how those systems have been adapted elsewhere and about their financial, productive, and environmental performance in other countries.


The views expressed herein are the author’s own and are not those of the Economic Research Service or the U.S. Department of Agriculture.

Further Reading

Gardner, B. L. (2002). American agriculture in the twentieth century: How it flourished and what it cost. Cambridge, MA: Harvard University Press.Find this resource:

Hart, J. F. (2003). The changing scale of American agriculture. Charlottesville: University of Virginia Press.Find this resource:

MacDonald, J. M. (2014, June). Technology, organization, and financial performance in U.S. broiler production. Economic Information Bulletin no. 126. Washington, DC: U.S. Department of Agriculture, Economic Research Service.Find this resource:

McBride, W., & Key, N. (2013, October). US hog production From 1992 to 2009: Technology, restructuring, and productivity growth. Economic Research Report no. 158. Washington, DC: U.S. Department of Agriculture Economic Research Service.Find this resource:

Pew Commission on Industrial Farm Animal Production. (2008, April). Putting meat on the table: Industrial farm animal production in America. A Project of The Pew Charitable Trusts and Johns Hopkins Bloomberg School of Public Health.Find this resource:

Steinfeld, H., Mooney, H. A., Schneider, F., & Neville, L. E. (2010). Livestock in a changing landscape: Drivers, consequences, and responses (Vol. 1). Washington, DC: Island.Find this resource:


Aarestrup, F. M., Jensen, V. F., Emborg, H. D., Jacobsen, E., & Wegener, H. C. (2010). Changes in the use of antimicrobials and the effects of productivity of swine farms in Denmark. American Journal of Veterinary Research, 71, 726–733.Find this resource:

Aillery, M., Gollehon, N., Johansson, R., Kaplan, J., Key, N., & Ribaudo, M. (2005, September). Managing manure to improve air and water quality. Economic Research Report no. 9. Washington, DC: U.S. Department of Agriculture, Economic Research Service.Find this resource:

Capper, J. L. (2011). The environmental impact of beef production in the United States: 1977 compared with 2007. Journal of Animal Sciences, 89, 4249–4261.Find this resource:

Capper, J. L., Cady, R. A., & Bauman, D. E. (2009). The environmental impact of dairy production: 1944 compared with 2007. Journal of Animal Sciences, 87, 2160–2167.Find this resource:

de Haan, C., Gerber, P., & Opio, C. (2010). Structural change in the livestock sector: Livestock in a changing landscape: Drivers, Consequences, and Responses (Vol. 1). Washington, DC: Island.Find this resource:

Gollehon, N., Caswell, M., Ribaudo, M., Kellogg, R., Lander, C., & Letson, D. (2001, June). Confined animal production and manure nutrients. Agriculture Information Bulletin no. 771. Washington, DC: U.S. Department of Agriculture, Economic Research Service.Find this resource:

Gollehon, N. R., Kellogg, R., & Moffitt, D. C. (2016, June). Estimates of recoverable and non-recoverable manure nutrients based on the census of agriculture—2012 results. Washington, DC: U.S. Department of Agriculture, Natural Resource Conservation Service.Find this resource:

Harrison, R. (1964). Animal machines. London: Vincent Stuart.Find this resource:

Jensen, H. H., & Hayes, D. J. (2014). Impact of Denmark’s ban on antimicrobials for growth promotion. Current Opinion in Microbiology, 19, 30–36.Find this resource:

Key, N., & Sneeringer, S. (2011, February). Climate change policy and the adoption of methane digesters on livestock operations. Economic Research Report No. (ERR-111). Washington, DC: U.S. Department of Agriculture, Economic Research Service.Find this resource:

Key, N., McBride, W. D., Ribaudo, M., & Sneeringer, S. (2011, September). Trends and developments on hog manure management: 1998–2009. Economic Information Bulletin No. (EIB-81). Washington, DC: U.S. Department of Agriculture, Economic Research Service.Find this resource:

Lagerkvist, C. J., Hansson, H., Hess, S., & Hoffman, R. (2011). Provision of farm animal welfare: Integrating productivity and non-use values. Applied Economic Perspectives and Policy, 33, 484–509.Find this resource:

Lusk, J. L., & Norwood, F. B. (2011). Animal welfare economics. Applied Economic Perspectives and Policy, 33, 463–483.Find this resource:

MacDonald, J. M. (2014). Technology, organization, and financial performance in U.S. broiler production. Economic Information Bulletin no. 126. Washington, DC: U.S. Department of Agriculture, Economic Research Service.Find this resource:

MacDonald, J. M., Cessna, J., & Mosheim, R. (2016, March). Changing structure, financial risks, and government policies for the U.S. dairy industry. Economic Research Report no. 205. Washington, DC: U.S. Department of Agriculture, Economic Research Service.Find this resource:

MacDonald, J. M., Ribaudo, M., Livingston, M., Beckman, J., & Huang, W. (2009). Manure use for fertilizer and for energy. Mandated Report to the Congress. Administrative Publication No. AP-037. Washington, DC: U.S. Department of Agriculture, Economic Research Service.Find this resource:

MacDonald, J. M., & McBride, W. D. (2009, January). The transformation of U.S. livestock agriculture: Scale, efficiency and risks. Economic Information Bulletin no. 43. Washington, DC: U.S. Department of Agriculture, Economic Research Service.Find this resource:

MacDonald, J. M., & Wang, S. L. (2011). Foregoing subtherapeutic antibiotics: The impact on broiler grow-out operations. Applied Economic Perspectives and Policy, 33, 79–99.Find this resource:

MacDonald, J. M., Ollinger, M., Nelson, K., & Handy, C. (2000, February). Consolidation in U.S. meatpacking. Agricultural Economics Report no. 785. Washington, DC: U.S. Department of Agriculture, Economic Research Service.Find this resource:

Martin, J. R. (1979, December). Beef. In L. P. Shertz (Ed.), Another revolution in U.S. farming? AER-441. Washington, DC: U.S. Department of Agriculture. Economics, Statistics, and Cooperatives Service.Find this resource:

Mathews, W. A., & Sumner, D. A. (2015). Effects of housing system on the costs of commercial egg production. Poultry Science, 94, 552–557.Find this resource:

McBride, W. (1997, July). Change in U.S. livestock production, 1969–92. Agricultural Economic Report no. 754. Washington, DC: U.S. Department of Agriculture Economic Research Service.Find this resource:

McBride, W., & Key, N. (2013, October). US Hog production From 1992 to 2009: Technology, restructuring, and productivity growth. Economic Research Report no. 158. Washington, DC: U.S. Department of Agriculture Economic Research Service.Find this resource:

Mosheim, R., & Knox Lovell, C. A. (2009). Scale economies and inefficiency of U.S. dairy farms. American Journal of Agricultural Economics, 91, 777–794.Find this resource:

Mullaly, C., & Lusk, J. L. (2018). The impact of farm animal housing restrictions on egg prices, consumer welfare, and production in California. American Journal of Agricultural Economics, 100, 649–670.Find this resource:

O’Donoghue, E. J., Hoppe, R. A., Banker, D. A., Ebel, R., Fuglie, K., Korb, P., . . . Sandretto, C. (2011, January). The Changing organization of U.S. farming. Economic Information Bulletin no. 43. Washington, DC: U.S. Department of Agriculture, Economic Research Service.Find this resource:

Ollinger, M., MacDonald, J. M., & Madison, M. (2000, October). Structural change in U.S. chicken and turkey slaughter. Agricultural Economics Report no. 787. Washington, DC: U.S. Department of Agriculture, Economic Research Service.Find this resource:

Pew Commission on Industrial Farm Animal Production. (2008, April). Putting meat on the table: Industrial farm animal production in America. A Project of The Pew Charitable Trusts and Johns Hopkins Bloomberg School of Public Health.Find this resource:

Reimund, D. A., Martin, J. R., & Moore, C. V. (1981, April). Structural change in agriculture: The experience for broilers, fed cattle, and processing vegetables. Technical Bulletin No. 1648. Washington, DC: U.S. Department of Agriculture, Economics and Statistics Service.Find this resource:

Ribaudo, M., Gollehon, N., Aillery, M., Kaplan, J., Johansson, R., Agapoff, J., . . . Peters, M. (2003, June). Manure management for water quality: Costs to animal feeding operations of applying manure nutrients to land. Economic Research Service. AER-824. Washington, DC: U.S. Department of Agriculture.Find this resource:

Rogers, G. (1979, December). Poultry and eggs. In L. P. Shertz (Ed.), Another revolution in U.S. farming? Economics, Statistics, and Cooperatives Service. AER-441. Washington, DC: U.S. Department of Agriculture.Find this resource:

Schroder, J. J., & Neeteson, J. J. (2008). Nutrient management regulations in the Netherlands. Geoderma, 144, 418–425.Find this resource:

Smith, J. B. (2014, September 5). Bosque river dairies, Waco officials seeing progress, trust, and water quality issues. Waco Tribune-Herald.Find this resource:

Sneeringer, S. E. (2009). Does animal feeding operation pollution hurt public health? A national longitudinal study of health externalities identified by geographic shifts in livestock production. American Journal of Agricultural Economics, 91(1), 124–137.Find this resource:

Sneeringer, S. (2016, September). Comparing participation in nutrient trading by livestock operations to crop producers in the Chesapeake Bay Watershed. Economic Research Report No. 216. Washington, DC: U.S. Department of Agriculture, Economic Research Service.Find this resource:

Sneeringer, S., MacDonald, J. M., Key, N., McBride, W., & Mathews, K. (2015, November). Economics of antibiotic use in U.S. livestock production. Economic Research Report no. 200. U.S. Department of Agriculture, Economic Research Service.Find this resource:

Sneeringer, S., & Key, N. (2011, July). Effects of size-based environmental regulations: Evidence of regulatory avoidance. American Journal of Agricultural Economics, 93, 1189–1211.Find this resource:

Steinfeld, H., Wassenarr, T., & Jutzi, S. (2006). Livestock production systems in developing countries: Status, drivers, and trends. Revue Scientifique et Technique (International Office of Epizootics), 25, 505–516.Find this resource:

Sumner, D. A. (2018). Comment on ‘The impact of farm animal housing restrictions on egg prices, consumer welfare, and production in California’. American Journal of Agricultural Economics, 100, 670–673.Find this resource:

Thornton, P. K. (2010). Livestock production: Recent trends, future prospects. Philosophical Transactions of the Royal Society B, 365, 2853–2867.Find this resource:

U.S. Department of Agriculture. (2017, February). Cattle on feed. Washington, DC: National Agricultural Statistics Service.Find this resource:

U.S. Food and Drug Administration. (2017, December). 2016 summary report on antimicrobials sold or distributed for use in food-producing animals. Washington, DC: Center for Veterinary Medicine.Find this resource:

Van Grinsven, H. J. M, Tiktak, A., & Rougoor, C. W. (2016). Evaluation of the Dutch implementation of the nitrates directive, the Water Framework Directive, and the National Emissions Ceiling Directive. NJAS-Wageningen Journal of Life Sciences, 78, 69–84.Find this resource:

Wossink, A. (2004). The Dutch Nutrient Quota System: Past experience and lessons for the future. In T. Tietenberg & N. Johnstone (Eds.), Tradeable permits: Policy evaluation, design, and reform. Paris: Organization for Economic Cooperation and Development.Find this resource: