Mucha res y poco cerdo: el consumo de la carne en Colombia

Este articulo busca explicar por que los colombianos han comido mas carne de res que de cerdo en comparacion con otros latinoamericanos. Comienza examinando el desarrollo de una tradicion culinaria que favorece la carne de res. El eje central del argumento, sin embargo, es que la carne de res ha sido, historicamente, bastante mas barata que la de cerdo. Esta diferencia de precio esta ligada al alto costo del maiz, que suele emplearse para la ceba de cerdos, debido a la baja productividad de la agricultura colombiana. Otros factores que favorecieron a la carne de res incluyen una frontera agraria en retroceso, una poblacion de cerdos pequena, las ventajas de la ganaderia, la monopolizacion de la tierra, la influencia de la importacion de manteca de cerdo y el desarrollo de una industria de aceite vegetal.


SHAWN VAN AUSDAL dossier
In 2007, Colombia ceased to be primarily a beefeating nation (El Tiempo, 2007). 1 The change was significant. For generations, possibly even since the early colonial period, Colombians consumed much more beef than any other meat. For every pound of lechona, fritanga, chicharrón, or other form of pork that Colombians have savored since the late-nineteenth century, they have eaten between five and seven pounds of beef (see Figure  1). To give some perspective, in the United States, the ratio of pork-to-beef consumption has been about 1-to-1.5 since 1950 (Skaggs, 1986, pp. 166-167;. Even in Latin America, which has long been cattle country, Colombia's beef-heavy meat diet has been extreme. At least since 1960, Colombians have been, in relative terms, the largest beef-eaters in tropical Latin America (see Figures 2 and 3). (They have also been, in absolute terms, one of the largest beef consumers per capita in the region.) Consequently, Colombians have consumed much less pork, proportionally, than many other Latin Americans. Whereas pork has comprised between a fifth and a third of the meat diet in Brazil and Mexico since 1950, in Colombia it has hovered around ten percent (United Nations, 1962, p. 45; United Nations, 1964, p. 50; Jarvis, 1986, p. 2;. This article marks the end of an era by asking two related questions: why, historically, have Colombians eaten so little pork? And what accounted for the long predominance of beef? Although a variety of factors converged to make Colombia a beef-eating nation, I suggest that it was the historic high cost of pork that played a fundamental role. Much of this article, therefore, is an effort to explain why beef has been cheaper than pork. I argue that, in the case of Colombia since the mid-nineteenth century, it was not cheap, natural grasslands that made beef less expensive, as commonly suggested, but the low productivity of Colombian agriculture. What I consciously downplay here is the cultural or social status that beef may have had in determining meat consumption patterns. In fact, this article originated in a frustrated effort to identify pro-beef discourses in the first half of the twentieth century (Flórez, forthcoming). To explain why Colombians eat so much more beef than pork, I turn away from recent trends in food studies, which underline the symbolic aspects and cultural politics of food, to emphasize production and price (Watson and Caldwell, 2005). It is possible that the symbolic value of beef played a greater role than I allow, but overall I do not think that it was that critical.
Three general ideas structure this paper. Lest I get too carried away with my price-centered argument, in the first section I examine the development of a culinary tradition that favored beef. There are hints of a deeply-rooted tradition of meat consumption in Colombia, despite the obvious social inequalities and important variations over time and space. Since beef appears to have been the most common meat, a form of dietary inertia developed around it. Tradition, however, can only explain so much. In the second section, therefore, I turn to the comparative advantage of grass. My argument, as noted above, is that the historic predominance of beef largely stemmed from the high cost of pork, which in turn was a consequence of the high cost of corn and the low productivity of Colombian agriculture. Despite the recurring criticisms of Colombian ranching as extensive and inefficient, grass and cattle gave beef some advantages over other meats. In the last section, I examine how the agrarian structure and government policies also conspired against a richer tradition of pork consumption: the former by constraining peasant production; the latter by undercutting the market for lard.

A TRADITION OF BEEF
Tradition is a tricky word since many so-called traditions are actually practices of fairly recent origin that were "invented" to naturalize specific interests (Hobsbawm and Ranger, 1992). Nonetheless, in this section I suggest that part of the historic predominance of beef consumption in Colombia has to do simply with the development of a culinary tradition in which beef played an important role: Colombians have favored beef because it is what they grew up eating, what they learned to cook, and what they came to expect.
The Colombian taste for beef emerged within a wider Latin American tradition stretching back to the early colonial period. It is possible that the Spanish exported a penchant for beef from Andalusia along with the cultural practice of raising cattle from horseback on the open range (Bishko, 1952;Jordan, 1993). But key to the place of beef in colonial diets was the rapid proliferation of cattle in New World environments. Free from major predators, Old World diseases, and ecological competition, cattle multiplied quickly on the region's grasslands. The precipitous decline of native populations, in part a result of growing herds, opened up yet more space for livestock and helped create an abundance of animals relative to consumers (Crosby, 1972;Melville, 1994). In some areas, such as the Pampas and Northern Mexico, local residents could not eat all the cattle they culled. There, they extracted more value from hides and tallow than meat (Pilcher, 2006, p. 30;Barsky and Gelman, 2001, p. 59). Even outside these areas, beef consumption could be remarkably high. European visitors were often astonished at the large quantities of meat that Latin Americans ate (Pilcher, 2006, p. 18). Colonial Caracas, for instance, is said to have consumed 50 percent more beef than Paris even though it had only ten percent of the population (Rifkin, 1992, p. 49).
Not everyone, however, had easy access to beef or other kinds of meat. Although cattle adapted to a wide variety of environments, they did not multiply with the same fecundity everywhere. For instance, cattle herds expanded more slowly on the Llanos, the great natural grasslands of Colombia and Venezuela (characterized by climatic extremes of searing heat and flooding, abundant but poor quality grasses, and natural predators), than in the temperate and benign Pampas (Crosby, 1972;Rausch, 1984;Rausch 1993). In many other regions, vast tracts of tropical forest limited the geographic and biological expansion of cattle. The miners of northeastern Antioquia, therefore, ate cattle bred in the distant Valle del Cauca, fattened on the highland pastures of Rionegro, and driven to slaughter in the mines amidst lowland forests (West, 1952, pp. 112-15;Parsons, 1968, pp. 127-28). Also, the initial population explosion did not last indefinitely: as their pressure on rangelands increased, the growth rates of herds tapered. By the late-eighteenth century, a growing demand for cattle -from expanding human populations, economic growth, and increased trade-began to squeeze existing stocks, causing prices to rise rather substantially (Sourdis, 1996, pp. 44-45;Brungardt, 1974;Pilcher, 2006, pp. 27-28). Beef, therefore, was not always in great abundance and inexpensive. Jeffrey Pilcher (2006, pp. 16, 22) reminds us that while the Mexican elite dined on exaggerated quantities of meat, the rural poor retained a largely vegetarian diet that pre-dated the Spanish conquest. In Colombia, highland peasants also appear to have eaten little meat up to the end of the nineteenth century (Camacho Roldán, 1946, p. 131;Meisel and Vega, 2004, p. 12).
Nonetheless, there is scattered evidence to suggest that meat consumption in Colombia was fairly widespread. While the overall quantities may have been small, and consumption erratic, I suspect that it was sufficient to make meat -particularly beef-a key component of the national diet and culinary imagination. Robert West (1952, p. 112), in his study of mining in colonial Colombia, was surprised at the "large quantity of meat" that miners ate. Mining ordinances from the seventeenth century required that Indian laborers be given 12 pounds of meat each per week (West, 1952, p. 95;Calero, 1997, p. 147; see also Taussig, 1977, p. 403;Hamilton, 1993, p. 291, 312). Even if the ordinances were not enforced, the stipulation that daily rations include almost two pounds of meat underlines its abundance. The half-rations of meat stipulated in case of a siege of Cartagena in the mid-eighteenth century (for the militia, artisans, and workers) included six ounces of beef and two ounces of bacon (tocino) per day (Dorta, 1962, pp. 351-352). Fray Juan de Santa Gertrudis (Serra, 1994, pp. [67][68] remarked that for "ordinary people" of the coastal lowlands "the common food…is generally just a stew of beef jerky and…yucca, arracacha, sweet potato, cassava or ñame root and sapallo 2 ." Other eighteenth -and nineteenth-century accounts also note that meals outside the highlands were "almost always accompanied by a piece of beef, no matter how dry and hard it might be" (Hettner, 1976, p. 219; see also Vargas, 1944, pp. 11-12;Holton, 1967, pp. 25-26, 198;Striffler, 1994, p. 175;Hamilton, 1993, pp. 40, 62, 115, 335). For the highlands, Vargas (1944, pp. 136-137) calculated that an ordinary daily ration in a hospital in Zipaquirá should include one pound of beef and one ounce of bacon, but he acknowledged that many people would be unable to afford it. Urban laborers appear to have eaten some meat, but by most accounts highland peasants consumed little (Boussingault, 1994, pp. 365, 367).
The expansion of the export economy from the midnineteenth century did much to increase beef consumption and solidify its place in the culinary imagination. The demand for workers in the tobacco fields and on public works projects increased wages in both the lowlands and highlands and "introduced beef consumption to the working class" (Camacho Roldán, 1946, p. 167;Rivas, 1983, p. 212;Nieto Arteta, 1996, pp. 262-263). Towards the end of the century, Camacho Roldán (1946, p. 164) thought that beef consumption was "one of the items whose consumption has improved notably…" (see also Arboleda, 1905, pp. 96-116) 13, 1916 andJan. 14, 1916;Poveda, 1979, p. 120;Brew, 2000, p. 174). The department of Bolívar even provided prisoners in the jail in Cartagena with half-a-pound of meat per day. 3 A taste for beef also likely spread through the rations that soldiers received (and requisitioned) in the frequent civil wars of the nineteenth century. 4 By the early-twentieth century, the general pattern of meat consumption that would last for much of the rest of the century, both in terms of quantity and kind, was already well-established (see Figure 1). 5 The above consumption rates are somewhat misleading, though. They suggest, multiplied over the course of a year, that some Colombian laborers and even prisoners ate more meat, and considerably more beef, than most Europeans at the time (Holmes, 1916, pp. 271-73;see Hettner 1976, p. 93). Most Colombians, however, did not eat half-a-pound of meat daily. From 1915 to 1927, it was more on the order of one to one-and-a-half ounces of meat per day on a per capita basis (Departamento de Contraloría, 1930, p. 459). This discrepancy between daily rations that were significantly higher than the national per capita rate of consumption probably stems, in good part, from the temporary and seasonal nature of much work. It is possible that many Colombians obtained much of their meat in the form of rations while working for others. 6 Therefore, even if the average Colombian did not consume large quantities of meat, many did have at least periodic access to it. By the turn of the twentieth century, such recurring consumption helped beef to become a fixture -even if sometimes more symbolic than real-in the national diet and imagination.
There were, of course, plenty of variations in the development of this tradition of beef. If work-rations were an 3 AHC, Gobernación, Justicia (1905-1933, Folder 25, "…Marcial González…y Juan Grice han celebrado el siguiente contrato: 1905-6". 4 Martínez, 1990 important source of meat for many peasants, this probably reinforced differentiated consumption patterns by gender and age. (Although the employment of women and children in such jobs as the coffee harvest possibly did a good deal to include them in the circuits of meat consumption.) Regional differences also mattered. The more dynamic regions, such as Antioquia, could afford to provide better rations than stagnant ones, such as Boyacá. Lowland residents also appear to have eaten more meat than highlanders, at least until the 1930s or 40s (Hettner, 1976 What is still not clear, however, is why beef predominated over other meats. Some scholars have argued that beef has stood at the pinnacle of the food hierarchy throughout much of Western civilization (Twigg, 1979;Fiddes, 1991;Adams, 1990; see also Beardsworth and Keil, 1997, pp. 209-217). Could it be that the Colombian (and Latin American) tradition of beef is rooted in such a larger cultural complex? Even though hogs multiplied more rapidly than cattle in the early colonial period, did people, when given the chance, deliberately choose beef? There is some evidence of an historic hierarchy of meats in Colombia. Inns and steamboats in the nineteenth century did not serve fish because it was considered too cheap (Holton, 1967, p. 51). 8 Hettner (1976, p. 110) even remarked that in late-nineteenthcentury Bogotá, "pork is generally relegated to feed the lower classes". There might also have been some wariness about pork due to hygienic concerns (Littman, 1965). We must not assume, however, that beef was always the "elite" meat. In colonial Mexico City, with its heavy Spanish influence, mutton enjoyed this role; beef was cheaper and more plebeian (Pilcher, 2006 9 Since at least the mid-twentieth century, it has been well-to-do Colombians who, proportionally, have eaten the most pork (González, 1969, p. 41). It is not clear, therefore, that some deep cultural preference for beef can explain its predominance (see Orlove, 1997).
As I will argue in the rest of this article, there are a variety of reasons that converged to make beef the principal meat in Colombia. What I want to emphasize here is the role of tradition. From early on in some places, and certainly by the end of the nineteenth century in much of the country, meat had become a key part of a 'proper' diet. And since beef was the most-consumed meat, a culinary tradition developed around it. People from all classes, regions, and races acquired a preference for beef and created a repertoire of ways to prepare it. They likewise failed to develop a wide range of pork-based dishes, as a comparison between Mexican and Colombian cookbooks will show. Pork, therefore, had less culinary appeal. As the manager of one hog farm remarked, "people do not have the custom of eating pork" (Gómez Cuéllar, 1909, p. 196).
Too much emphasis on tradition, however, leaves little room for change (Mennell, 1996, pp.4-6). Since the mid-nineteenth century, there have been shifts in tastes and culinary practices: from salted or dried to fresh beef; from boiled to pan-fried; the slow spread of new cuts; the gradual diffusion of dishes once restricted to elites; the phenomenal growth of chicken consumption in recent years. But many of these are variations on a theme; the basic structure of the cuisine remains very similar to 9 In the colonial era, pork was also considered to be good for one's health (Saldarriaga, 2006).
what it was a century or more ago: a bit of meat amidst a plethora of starches. 10 Additionally, taste may play a role beyond the significance of tradition: what I (sarcastically) call the culinary determinism of beans. In many cuisines around the world, there is a propensity to pair pork and beans: feijoada (Brazil), cassoulet (southwestern France), split-pea soup (northern Europe), fabada (Asturias, Spain), pork-and-beans (U.S.), frisoles (Antioquia, Colombia). Could it be that Colombians eat little pork partly because they are not big bean eaters? Mexicans and Brazilians, for example, eat three to four times more beans than Colombians, and they eat proportionally more pork . In Colombia, the region that consumes the most beans -Antioquia-is also one of the biggest consumers of pork (Varela, 1952, p. 117;Restrepo, 1988, p. 97). It is plausible that Colombia developed such a strong tradition of beef partly because of the availability of a wide variety of other starches besides beans -potatoes, yucca, plantain, among many others-all of which do well cooked in beef -rather than pork-flavored water. This pork-bean link is not entirely consistent throughout Latin America. Chileans, for example, eat a fair amount of pork but few beans; and Nicaraguans eat lots of beans but little pork . By themselves, beans are not a determinant. Nonetheless, this idea does suggest the variety of possible factors at play behind culinary traditions and consumption patterns.

THE COMPARATIVE ADVANTAGE OF BEEF
While tradition and beans probably played some role in the formation of meat consumption patterns in Colombia, there is a simpler and more fundamental factor: price. Historically, beef has been significantly cheaper than pork. While the price data I have gathered so far is scattered, it does show a consistent, and often substantial, premium paid for pork. In Bogotá, between 1953 and 1965, pork was almost 20 percent more expensive than beef. In general, however, the difference was upwards of 40 percent and sometimes it was even higher (see Figure 3). Back in the eighteenth century, pork was four times the price of beef (Vargas, 1944, p. 90). 11 Given the substantial price SHAWN VAN AUSDAL dossier elasticity of meat consumption in Colombia, much of the historic preference for beef is likely just a reflection of the higher cost of the alternatives . After all, if pork costs 40 percent more than beef, the decision by a poor family of which meat to buy does not seem very difficult. For this reason, Gómez Rueda (1936, p. 499), head of the government's Department of Livestock, stated that "if the price [of pork] were lower than it currently is, which would allow it to compete with beef, its consumption would increase considerably".
A variety of other circumstantial evidence also points to the importance of price. At the beginning of the twentieth century, a number of large hog farmers from the Sabana de Bogotá noted that one of the obstacles their industry faced was the public's penchant for beef rather than pork. But they also realized that it would be difficult to increase pork consumption so long as it remained more expensive (Gómez Cuéllar, 1909). Second, studies also show that Colombians eat more pork when they have more disposable income (González, 1969, p. 41). Third, the one principal exception to the price premium for pork was in the Sinú Valley from the 1930s to the 1950s. There, hog raising expanded after the government restricted lard imports. Instead of shipping live hogs to the interior of the country, butchers slaughtered them locally and tins of lard were sent inland. Without a way to transport the meat to other markets, it had to be consumed quickly in the region. As a result, pork became cheap enough here to be "the meat of the poor." 12 Finally, at least since the 1960s, the price difference between pork and beef in Mexico and Brazil has been smaller than in Colombia, which may help explain why these two countries consume more pork (FAOSTAT, 2005).
We should be careful not to place too much emphasis on price alone, however. In Colombia, pork consumption sometimes has a quasi U-shaped curve in which people from poorer and wealthier groups consume proportionally more pork than those in the middle ( bought pork because of its property as a flavoring agent rather than to eat meat per se. 13 Nonetheless, if price does play a key role in determining meat consumption patterns, why has beef been historically cheaper than pork? Hogs are generally considered to be more productive than cattle: they convert feed into flesh more efficiently; they are more prolific; they grow and fatten faster; and they consume waste products (not only farm surplus but inferior agricultural products with little value, waste from agricultural processing, and kitchen scraps . Given the low opportunity-cost of many hogs, why would pork be more expensive than beef? The traditional answer is that Colombia, and indeed much of Latin America, has had a comparative advantage in natural grasslands. Ernst-Ludwig Littman (1965, p. 7), for example, noted that beef remained cheaper than pork even when the price of grains was favorable for raising hogs: From the biological point of view, this is a contradiction, since the efficiency of converting forage into meat is higher in hogs than cattle. The explanation is found in the existence of vast extensions of natural grasses in Colombia just as in Argentina, which can be used for extensive cattle production without large investments…. Ample land resources and the availability of labor give a decisive advantage to cattle raising at the cost of hog farming.
Likewise, Lowell Jarvis (1986, p. 10) stated that in Latin America the "abundant supply [of cattle] from low-cost pastoral resources has led to a tradition of high beef consumption by all levels of the population". Cheap grass, principally from natural grasslands, enabled Latin American cattle to out-compete hogs.
In Colombia, however, this argument runs into some problems. Here, cheap savanna land has been less criti- 13 It should be noted, however, that the price differential in Barranquilla at this time was not as large as in other parts of the country. It is likely that, once it surpassed a certain price, it was no longer worth buying pork.
cal to the national cattle industry than perhaps elsewhere in Latin America. Colombia's quintessential grasslands, the Llanos Orientales, were not that significant nationally until well into the twentieth century (García, 2003;Rausch, 1993). The center of Colombian ranching for the last century or so has been the Caribbean coast and Antioquia. Here, the predominate forage has not been from natural savannas but from "artificial" or planted pastures. While cattle mainly grazed on natural grasslands up to the mid-nineteenth century, since then the big expansion of ranching has largely come at the expense of the country's lowland forests. This transformation of forest into pasture, however, was often difficult, risky and expensive: grass was not cheap for the taking (Van Ausdal, MS). Trying to pinpoint the cost differences between producing grass and corn, or cattle and hogs, is not straightforward, however. Data, especially before 1950, is difficult to find, scattered geographically and temporally, and is not always easily comparable. One way to try to address the comparability problem is to calculate production costs in terms of the labor required to fatten hogs and cattle. Based on data from Antioquia in the 1950s, I calculate that it took roughly six or seven days of labor to produce the corn necessary to fatten one hog. 15 By contrast, in the Bajo 14 In the early 1970s, a development project found that penned hogs, fed improved diets, gained weight three times faster and with three times less feed, than "traditionally" raised animals. But because of the high cost of 'improved' feed, even with such productivity gains, profi t margins remained razor thin while the degree of risk greatly increased. Needless to say, the targeted peasants did not adopt most of the project's recommendations (CIAT, 1972). 15 The Caja de Crédito Agrario, Industrial y Minero (1955) estimated that it cost $383 to produce and market one hectare of corn in Antioquia. This was equivalent to over 100 days of labor in terms of prevailing wages. The average yield from this hectare was 1,500 kilograms of corn. If we consider only the actual labor costs of land preparation, planting, and weeding -and ignore land, harvesting, processing, marketing, and administrative costs-it took about 44 days to grow one hectare of corn. One day's labor, therefore, produced about 34 kilograms of corn. Historical observers suggest that it took between 200 and 250 kilos of corn, and two months, to fatten a hog, doubling its weight from roughly 50 to 100 kilograms (Ospina, 1940;Bernal, 1937).
Cauca in the early 1920s, it took only four days of labor to fatten a steer. 16 Obviously, this comparison does not tell the full story, but it does suggest that one of the problems faced by the Colombian hog industry was the amount of labor required to produce feed. One of the main advantages of raising cattle rather than hogs was that the substantial initial cost of buying or developing pasture land could be amortized over a relatively long period of time; by contrast, the land preparation and weeding costs needed to grow corn were a constant and heavy burden. A quick comparison with the United States provides some perspective on the amount of effort that went into farming corn in Colombia.  1946, pp. 183, 195) estimated that about half of the hogs slaughtered in Colombia (some 300,000 to 400,000) were fattened on corn, "each one of which consumes between two and four hundred pounds of grain" 20 . According to the manager of a large hog farm near Bogotá at the beginning of the twentieth century, corn or other grains and tubers (barley, fava beans, potatoes, etc.) were too expensive to be used at any other stage except fattening (Gómez Cuéllar, 1909, p. 196). The 300 to 400 percent difference between the price of thin and fat hogs likely 19 By low productivity I mean that crop yields were not very large in terms of inputs of land, labor, and capital. In turn, this helped make agricultural products expensive. Limited capital inputs (e.g., fertilizer), partly the result of an historic urban bias in agricultural policy, were one cause of low yields per unit of land and labor. (My thanks to one of the anonymous reviewers for highlighting this point.) Colombian agriculture faced additional challenges as well. For instance, the rugged topography of much of the country's farmland made mechanization diffi cult. The high cost of inputs also undercut a good deal of the increased productivity gains of mechanized agriculture (see Céspedes, 1979, Tabla 15). 20 Alejandro López (1915, p. 28) stated that in Antioquia, 100,000 hogs were fattened on about 200 kilograms of corn each. See also Monsalve (1929, p.148), who noted the maxim, "purchased corn does not fatten"; Ospina, 1940;Bernal, 1937. rested on the low opportunity-costs of breeding pigs and the importance of corn to fatten them. 21 Wylie, therefore, was probably correct to state that, overall, hogs ate little corn. Farmers generally allowed their hogs to roam freely, feeding on pasture grasses and in fallow fields or forests until the age of 10 to 12 months. Afterwards, however, enough hogs did fatten on corn (and other agricultural products) for it to become an important determinant in the price of pork and lard.
Moving hogs to market likely reinforced the use of corn to fatten them for slaughter. If they walk much without being fed, hogs will burn off their stores of fat and profit. This circumscribed the geographic area in which they could be profitably fattened for a particular market. 22 Hogs from outside this area had to be fattened locally.
The wide "hog-shed" of larger markets thus helped sustain the inter-regional movement of animals and the numbers that had to be fattened on corn or similar products. When frontier lands were relatively close to major markets, hogs were a convenient way to transform surplus corn (or other crop) into a mobile and marketable product. Once the distance became too great, however, raising fat hogs ceased to be profitable (COINCO, 1988). Similarly, extensive hog raising where land was cheap, whether by foraging, scavenging, or feeding on pasture grasses and legumes, was mostly limited to thin animals. Fat hogs needed a relatively easy way to get to market or had to be slaughtered locally. This is why hog farmers in the Sinú Valley -the largest hog-producing region in the 1930s-slaughtered the bulk of their own animals. It was more profitable to sell tins of lard to the interior of the country, even though this undercut the price of pork locally, than send live hogs.
Additionally, the high cost of corn (and other feed) also helped raise the price of pork by limiting the overall pig population. Because grains were expensive, hogs were principally a backyard activity that turned waste products into cash and took advantage of female and child labor. Since each peasant household could only generate so much waste, this limited the number of hogs they could profitably raise. As a result, cattle far outnumbered hogs. 21 Walsh (1977, p. 707) thought that the proliferation of small pork-packing operations meant that most farmers in the U.S. Midwest during the early -and mid-nineteenth century did not send their hogs great distances to be slaughtered. For longer trips they tended to take advantage of canals and later railroads. See also Cronon (1991, pp. 225-227).
In 1960, there was only one pig in Colombia for every five head of cattle (DANE, 1964, pp. 53-54). A team of livestock experts from the United Nations (1962, p. 5) found that there was good demand for pork in the 1950s, but that its "consumption [was] severely restricted by supply difficulties and high prices". Small hog populations likely helped boost prices.
Four other factors further limited hog supplies and sustained prices. First, the settlement of the agrarian frontier and the spread of pasturelands eventually undercut hog raising. Corn yields tended to be high in lands recently cleared of old-growth forest, and hogs were a convenient way to transport surpluses to market. Areas of colonization, therefore, were often important hog producers. The bonanza years were short-lived, however. And as falling corn yields pushed the corn-hog complex increasingly further from markets, it became harder to profitably use hogs to turn surpluses into mobile commodities (Parsons, 1968, p. 89). Furthermore, the introduction of African pasture grasses in the mid-nineteenth century helped changed the dynamics of pasture formation. These quickgrowing and livestock-resistant grasses -pará (Brachiaria mutica) and guinea (Panicum maximum)-helped prevent forest re-growth in recently cleared areas by rapidly forming a dense mat-like ground cover (Rivas, 1983, p. 36; Van Ausdal, MS; also see Parsons, 1972). They encouraged ranchers to develop new pastures out of the forest and farmers to plant grass in fallow fields. As a result, the once forested landscape -or patchwork of forest, fallow, and field-became increasingly dominated by grass. The development of new pastures initially stimulated corn production, which was used to loosen the soil before planting grass and help cover expenses. In the long run, however, the spread of permanent pastures stimulated cattle ranching at the expense of hogs. There is some scattered evidence to suggest that Colombians, at least in some regions, consumed more pork when there was more forest and less grass. In the mid -to late-nineteenth century, pork consumption rates appear to have been higher in parts of Old Bolívar than they would be in the twentieth century. 23 Pedro Nel Ospina could not find a market for his cattle in Ituango (Antioquia) in the early-twentieth century because the peasants of this frontier zone raised too many pigs (APNOyC, 200, p. 452). Pork consumption, though never very strong, may have slowly tapered off as colonos and cattle ranchers cleared the forests and planted grass (see Figure 1).
Second, the lack of elite interest in raising hogs limited their numbers. This reluctance may have stemmed from a degree of "repugnance" for the activity (Gómez Cuéllar, 1909, p. 193). But there were some practical reasons as well. For example, there were greater economies of scale in raising cattle than extensively-raised hogs (Poveda, 1979, p. 110; see also Van Ausdal, forthcoming). Many also thought that intensive hog raising was not worth the effort. Most dairy farmers on the Sabana de Bogotá preferred to throw or give away their leftover whey from cheese-making rather than use it to raise pigs (Gómez Cuéllar, 1909, p. 193). For hacendados without a ready supply of waste products that they could use for feed, raising hogs would require them to become farmers. Although the landed elite thought of themselves as agricultores, farming was something that, outside of a few products, many were unwilling to do: the risks of climate, pests, and overproduction were too great (Camacho Roldán, 1976, p. 125; Brew, 2000, pp. 210-212;Reinhardt, 1988; Van Ausdal, MS). Also, there were few economies of scale in farming staples until the spread of mechanization in the 1950s. 24 As a result, the landed elite generally left hog raising to peasants, a division of labor that the 1960 agriculture and livestock census clearly shows (see Figure 4).
A third factor that limited hog raising was the difficulty of generating low-cost feed from the waste of food processing industries. Colombian hog boosters encouraged entrepreneurs to take advantage of cheap by-products such as skim milk and whey from dairies, bran from millers, mash from breweries, oil seed cakes from vegetable oil mills, and slaughterhouse waste (Gómez Cuéllar, 1909;Medina, 1936, p. 444;Bernal, 1937;Ospina, 1940). While there were some steps in this direction, even into the second half of the twentieth century little progress had been made (León et al., 1975;Restrepo, 1988). One reason was the late urbanization of the country. Small urban markets, and low purchasing power more generally, limited the amount of available by-products. The geographic dispersion of waste material further limited the economies of scale that could have turned them into low-cost sources of feed. For example, without a system of cold chains with which to centralize slaughtering, SHAWN VAN AUSDAL dossier every municipality had its own slaughterhouse. Even in the second half of the twentieth century, many of these were too small to warrant processing their waste (United Nations, 1962, pp. 26-27). Even where there was a high degree of concentration, such as the vegetable oil industry in Barranquilla, high transportation costs discouraged the use of cottonseed cake as hog feed.
Finally, while traditional methods of 'extensive' hog rearing did not limit their numbers, they did limit the amount of meat that existing stocks produced. Even in the second half of the twentieth century, most Colombian hogs were hardy, self -reliant, disease-resistant creole breeds that had adapted well to local environments. But compared to improved breeds, they grew and reproduced slowly, and yielded less meat and fat (León et al., 1975, p. 16;Peñarete, 1958, p. 216;Ospina, 1913, p. 237). Furthermore, many hogs raised in this fashion received inadequate diets, which slowed their growth even more, lowered the speed at which they reproduced, and increased their mortality rates (Littman, 1965, 3;León et al., 1975, pp. 10, 32;CIAT, 1972, 92-93;Bernal, 1937, p. 936; Gómez Rueda, 1936, p. 554). Before 1950, boosters claimed that it was possible to raise hogs in six to ten months and fatten them in two (Bernal, 1937, p. 925;Ospina, 1940, pp. 40-41). Most Colombian hogs, however, were not raised under 'ideal' conditions. Even in the early 1970s, it could take 15 to 18 months before they were ready for slaughter (CIAT, 1972, p. 92;León et al., 1975, p. 10; see also Ospina, 1913, p. 237;Bernal, 1937, p. 936;Peñarete, 1958, p. 216). Slaughtering hogs at this late age meant that they did not produce meat much more efficiently than cattle. 25

THE POLITICS OF PORK
In the previous section, I argue that the high price of pork relative to beef limited its consumption. Pork was expensive because of the high cost of corn, used to fatten hogs, as well as a small pig population. These immediate causes, in turn, were rooted in the low productivity of Colombian agriculture, a small national market, the slow process of industrialization, elite disinterest, and the advantages of pasture and cattle. But just how 'natural' was the comparative advantage of beef? Was it primarily an economic issue? Or did politics, power, and policy decisions also have an impact? In this section, I look at the politics behind meat consumption in two different ways. 26 First, I examine how the political power of ranchers and their monopolization of much of the country's land negatively influenced hog raising. Second, I explore how the politics of lard imports helped to undermine pork consumption.
Two salient characteristics of the Colombian countryside have long been the inequitable land tenure structure and the vast majority of 'agricultural' land dedicated to raising cattle. Could the monopolization of land by ranchers be responsible for the country's beef-heavy diet? There are two ways this monopolization might have influenced meat consumption patterns. On the one hand, ranchers may have boosted the competitiveness of beef by raising cattle on the best agricultural land in the country. In so doing, they benefited from better pastures and more productive cattle operations. But more importantly, their control of the flat, fertile valley floors forced peasant agriculture onto more marginal hillsides, lowering its productivity and raising food (and feed) prices. This 'irrational' distribution of agricultural and pasture lands may have helped give beef an edge over pork (IBRD, 1956, p. 54;Currie, 1950). On the other hand, it is possible that the general monopolization of land also contributed to the competitiveness of cattle. One consequence of this monopolization was a land-hungry peasantry that was willing to clear forests and plant pasture in exchange for temporary access to land. In such arrangements, ranchers provided peasants with patches of forested land to farm for a few years so long as they returned it under grass. 27 These land-for-pasture exchanges aided the expansion of ranching and ultimately may have had some effect on the price of beef. Another consequence of this monopolization was to push peasants out to the agrarian frontier where they would undertake the hard labor of settling the forest. Ranchers later followed them to consolidate the lands that they had cleared (Fals Borda, 1976Borda, , 2002Negrete and Garabito, 1985;Reyes, 1978). Beef might have also benefited from the labor of these peasant colonizers through cheap land 26 For reasons of space, I do not address how the government supported the cattle industry -through research and extension work, subsidies, credit, and other programs-but did little to assist hog farmers. While such policies helped ranching expand geographically, they also contributed to the slow improvement of ranching productivity. While there are also some indications that the productivity of hog raising improved over the twentieth century, such gains probably lagged behind those of the cattle industry. This could be another way that beef maintained its competitive edge over pork. 27 Fals Borda, 1976 sales and the usurpation of developed plots. Lastly, the unequal land tenure structure squeezed the peasantry onto a small land base. Since peasants raised the bulk of the hogs in the country, the reduced size of their farms limited the number that each family was capable of raising and, therefore, the overall pig population.
Although it is likely that the inequitable distribution of land in Colombia negatively influenced hog raising, we should be careful not to overstate its impact. Any substantial expansion of hog raising would likely have depended on obtaining a cheap source of feed. Even though the peasantry might have been able to raise more hogs if they had greater access to land, there still remained the problem of fattening them for market. While they could have also grown more corn, without some way to substantially increase the productivity of their labor, hog and pork prices would have remained high. Additionally, without high-protein feed supplements, peasant hogs were not much more efficient than cattle at producing meat. And although a good number of ranchers profited from land-for-pasture exchanges or by cheaply 'buying' peasant clearings, the cattle industry did not depend on such practices (Van Ausdal, MS). There are limits to how much land usurpation subsidized the price of beef.
The second kind of political influence I want to address revolves around the question of trade and the place of lard in the Colombian diet. In 1944, the newspaper, Diario de la Costa, blamed the 'demise' of the coastal hog-raising industry on the 1936 trade treaty between Colombia and the United States: a 50 percent drop in the tariff on lard, it argued, unleashed a wave of U.S. imports that drove peasant hog producers out of business. 28 Up to the midtwentieth century, lard (along with tallow) was the most important source of cooking fat in Colombia and a key part of the hog industry. 29  Yet local supplies could not keep pace with growing demand, and by the end of the 1920s Colombia was again importing about half of the lard it consumed. 30 In an effort to reduce its dependence on U.S. lard, the government raised the import duty on lard by 500 percent in 1931. This measure considerably slowed imports but did not stop them. The following year, therefore, the government promulgated a sanitary regulation that effectively put an end to the trade. 31 Lard was an important component of the 1936 trade treaty between the U.S. and Colombia. However, while the treaty reduced the tariff by 50 percent, it did not address the sanitary restrictions. U.S. lard remained blocked from the Colombian market except when cooking-oil shortages during WWII prompted the government to temporarily lift the restriction. 32 The 1936 trade treaty, therefore, did not undermine Colombian hog producers. Nonetheless, the much longer history of lard imports prior to 1930 probably did discourage production by keeping a lid on prices (Camacho Roldán, 1973, p. 197; Restrepo Plata, 1912, p. 400; Díaz, 1996, p. 329).
(Though whether or not the country could have afforded much higher lard prices, without drastically reducing its consumption, is another question.) Ironically, the protectionism of the early-1930s eventually helped undermine hog raising and pork consumption. Despite the government's interest in reducing lard imports, it is doubtful that peasant hog-farmers could have effectively lobbied the government for protective trade policies. In fact, the government designed the measures to protect the nascent vegetable oil industry. By limiting lard imports, and reducing the duty on copra, the primary raw material in vegetable shortening, officials helped domestic manufacturers lower their production costs and grab a larger market share. Hog producers initially benefited from the measures. In the long run, however, they lost out as vegetable shortening (and eventually oil) displaced lard as the preeminent cooking fat. 33

SHAWN VAN AUSDAL
dossier Beef Pork Chicken

CONCLUSION
So why have Colombians historically eaten so much more beef than pork? The principal reason, I argue, is that pork has long been more expensive than beef; and that this price difference is rooted in the low productivity of Colombian agriculture. But the road to understanding consumption patterns is rarely straight and short. In this case, a variety of factors, in addition to price, converged to make beef -until recently-the king of meats in Colombia: the development of a taste for and tradition of beef; the culinary influence of other staples; a receding agrarian frontier; the lack of elite interest in hog raising; land tenure patterns; the difficulty of developing byproduct industries; a long history of cheap imports; the "modernization" of cooking fats. Although consumption studies have recently begun to emphasize the ideological and contested nature of food and diet, my focus here has been largely material: on what has influenced price and supply rather than the cultural politics of demand. Food is not just a matter of sustenance; a range of symbolic, cultural and political factors shape what and how we eat. Nonetheless, in the case of meat consumption in Colombia, the high cost of (corn and) pork did much to secure the long-held predominance of beef.