Shifting Cultivation:  From the History of Agriculture to the History of Food Systems

（中文译文）

 Peter A. Coclanis

     With globalization all the rage, agriculture doesn’t get a lot of media attention these days.  To be sure, now and then agricultural issues do make the news: for starters, one can point to recent battles over genetic modification of various cultigens, banana wars between nations in the Caribbean and nations in Latin America, denunciations of EU subsidies and U.S. farm bills by politicians in developing countries, predictable North-South debates at WTO meetings and, previously, at every round of GATT (General Agreement on Tariffs and Trade), and dire reports regarding the likely effects of drought, floods, disease, etc., on the supply of one or another commodity emanating from one or another part of the world. Generally speaking, however, in terms of reader interest, stories relating to agriculture can’t touch pieces on corporate misbehavior, currency devaluations, trade deficits, or the opening of a new auto plant. 

     On the face of it, the media’s relative disregard of agriculture is surprising.  Although farmers comprise only a tiny proportion of the labor force in developed countries, roughly half of the world’s workers are still employed in agriculture.[1]  If agriculture’s relative share of world output has been declining for centuries, food and fiber production has grown monumentally in absolute terms, and everyone will always have to eat.

     Why, then, the relative disregard of, and seeming disrespect for farming?  According to some, agriculture seems like “yesterday’s news.”  Farmers just seem too old-fashioned  and too traditional for our fast-paced, high-tech times.  At one time they may have been objects of veneration the world over, but in 2003 farmers seem a bit too elemental, too close to nature, too dirty and too sweaty for our increasingly sophisticated urban, high-tech lives.

     Similarly, the history of agriculture is also slowly receding from the scholarly scene, at least in the West.  Fewer and fewer students are interested in studying farming, being attracted instead to trendier fields.  For every student interested, let us say, in agricultural reform in eighteenth-century England, cotton production in the nineteenth-century American South, or cattle ranching in twentieth-century Argentina, there are ten interested in cultural history, women’s history, and the history of sex.  As a result, agricultural historians as a group are losing power and influence as they either die off or get longer and longer in the tooth!

     What is to be done?  Is there any way to reverse the above situation, at least insofar as it relates to agricultural history? In my view the answer is yes, but the solution will require scholars to reinvent themselves, to shift cultivation, as it were.  Simply put, instead of focusing so much attention on isolated farmers in the field, scholars most adopt a broader research purview and embed the study of farming in the modern period into an analysis of what might be called the “food system” as a whole.  In so doing, they will be able at once to develop a more sophisticated approach to agricultural production itself, and more explicitly to relate the farm sector to developments in the manufacturing and service sectors of the economy.  If we are lucky we may as a concomitant attract the interest of bright, ambitious young scholars, scholars who right now wouldn’t be caught dead pursuing research in agricultural history as traditionally conceived.

     But what do I mean by the term food system?  What activities are included under this rubric?  To answer the latter question first, all activities involved directly or indirectly in the production, storage, processing, financing, and distribution of outputs produced in the farm sector per se.  As such, the “food system” approach in some ways resembles the so-called commodity-chain approach championed by an increasing number of social scientists around the world.[2]  In this approach, a scholar interested in the athletic shoe industry, for example, would begin with rubber tapping in Malaysia or Indonesia and not end until he or she has adequately explained how and why poor African American youths in U.S. ghettoes buy so many pairs of expensive Michael Jordan model basketball shoes.[3]

     In principle, then, adoption of the “systems” approach to agriculture entails study of the process by which farm inputs are acquired, cultivation practices broadly conceived, the manner in which production is financed, how and where farm output is stored, the processing of such output, transportation logistics, and all levels of marketing and distribution.  Nor do these matters, even taken together, exhaust the list of relevant areas of inquiry.  Clearly, the state’s role, positive or negative, is crucial to the systems approach, particularly assessing the state’s ability to provide a stable context for production and exchange to take place, and its capacity to develop human capital generally and to foster and sustain the production and dissemination of specialized agricultural knowledge specifically.  In these matters, private and non-statal institutions and organizations obviously also play roles, whether we’re speaking of market stabilization (trade associations, co-operatives), risk reduction (futures markets, formal or informal crop-insurance schemes), or human capital development (private schools, NGOs, and international donors and foreign aid).

     To pursue the systems approach one need not necessarily study all of these aspects of agriculture, certainly not in a methodical way.  Just as a business firm need not be completely vertically integrated to achieve some of the benefits of the strategy, a researcher can achieve positive gains by approaching and conceptualizing agricultural problems in a relational, systemic, process-oriented way.  In this regard the words of Eugene Genovese, the great historian of American slavery, are very apropos:

           No subject is too small to treat.  But a good historian writes well on a

           small subject while taking account (if only implicitly and without a

           direct reference) of the whole, whereas an inferior one confuses the

           need to isolate a small portion of the whole with the license to assume

           that that portion thought and acted in isolation.[4]

To complicate our task a bit more, let me say that however much I appreciate the intellectual value added by the “commodity-chain” approach, and by the “chain” metaphor itself, other metaphors actually come closer to what I had in mind in writing this piece.  History is not linear and agriculture is not plane geometry.  Human behavior is more complex than that.  Rather than leaving you with the image of a chain linking together the various components of the agricultural system, I would propose a more intricate and elaborate metaphor, one that could be rendered graphically via a three-dimensional image of some type or another perhaps, or verbally by invoking the image of a web or, borrowing from physics, that of a field.  In each case, the idea is to move us away from the rigidity and constraints imposed by metaphors invoking chains.

     So what is the upshot of this appeal?  Other than for its potential to attract broader constituencies to matters agricultural, why move from the history of agriculture to the history of food systems?  Is this a distinction without a difference?

     In my view the answer is a resounding no: the difference, though subtle, matters a great deal.  Generally speaking, agricultural historians have traditionally focused on what farmers did while standing in their fields; historians of industrialization studied the various processing industries; business historians and students of marketing treated finance, transport logistics, and wholesale and retail distribution.  Seldom do these scholars talk to, or interact with, much less collaborate with one another.  Forty-five years ago two professors at the Harvard Business School, John H. Davis and Ray A. Goldberg, coined the term “agribusiness” to describe the high degree of interdependence that existed between the worlds of agriculture and business at the time.  Such interdependence, according to the authors, was a function of a plethora of technological, organizational, and institutional changes that over the past century-and-a-half had at once narrowed the function of the farm and transferred a host of functions formerly performed on farms to non-farm sites.  To Davis and Goldberg, “agribusiness” denoted “the sum total of all operations involved in the manufacture and distribution of farm supplies; production operations on the farm; and the storage, processing, and distribution of farm commodities and items made from them.”[5]  In other words, Davis and Goldberg’s term is pretty similar to the systems approach I am advocating today. 

     Deeply influenced by input-output analysis--the national-income accounting methodology pioneered by, and associated with future Nobelist Wassily Leontief--Davis and Goldberg were hoping to capture not only the complexity of the food-production sequence and the various and sundry functions, operations, and constituent parts thereof, but also the close interrelationship in modern agriculture between town and country, between factory and field.[6]  Although the authors, as indicated above, were concerned primarily with the modern era, when “[t]he functions of storing, processing, and distributing food and fiber have been transferred in large measure to off-the-farm entities,” I believe that one of their central conceptual insights--the disaggregation of  “farm” activities into discrete “agricultural” and “non-agricultural” tasks--is extremely helpful whether such activities are performed on-site or off-site, and whether we are looking at the agricultural sector five hundred years ago or “agribusiness” today.[7]

     Despite the usefulness of Davis and Goldberg’s concept, it never caught on either with the general public or with scholars--at least in the way the authors intended.  Indeed, it was not long before the concept became primarily associated not with the relationship between agriculture and business as the name implies, but as a synonym for large commercial farming operations, both in the U.S. and elsewhere.  This was unfortunate for several reasons, not least because the term often took on a negative connotation, but also because it impeded the development of a dynamic, integrated approach to the study of food and fiber flows both historically and in the present day.

     Up until now this paper has been largely exhortatory in nature: in terms of rhetorical strategy I have clearly been trying to convince you in a general way of the efficacy of the food systems approach.  In the remainder of the paper, however, I shall try to demonstrate some of the specific advantages of this approach by focusing on one part of the system, as it were--post-harvest activities--and by relating how and why such activities must perforce be studied closely if we would truly understand modern agriculture.  To foreground my argument, I shall argue that under appreciated, if not completely  overlooked innovations in food storage, processing, transport, preservation, finance, and marketing may have meant more in economic and social terms  than the developments agricultural historians have traditionally chosen to zero in on:  crop cultivation and the raising of livestock.[8]

     For example, for well over a century serious students of agriculture the world over have kept themselves busy trying to document historical changes in production indices such as seed/yield ratios, yields per acre/hectare, yields per laborer employed, etc.  In so doing, they have often undertaken ingenious, even heroic acts of data reconstruction via interpolation and extrapolation from incomplete data.  One of the most impressive of such efforts, by the great Dutch agricultural historian B.H. Slicher van Bath, produced such estimates for European agriculture century by century for the entire period from 500 C.E. to 1850 C.E., fully thirteen hundred fifty years.[9]  Specialists on the agricultural histories of other regions have written similar studies, for which all students of agriculture should be profoundly thankful.  Even if the indices produced in such studies (yields per acre, for example) are often misinterpreted--without additional information about labor and capital inputs we can’t assume ipso facto that rising yields demonstrate rising productivity--data on yields and changes in yields over time undeniably constitute important building blocks for any rich and dense analysis of agriculture.  My problem is not with studies of cultivation practices, farm technology, agrarian class structure, etc., but with the general tendency of scholars to focus on such topics at the expense of virtually everything else.  This tendency is particularly problematic because it rests upon the unproven or even untested assumption that changes in food production are more important than changes in other segments of the food system.  This may or may not be true, or it may be true in some cases but not in others, but we shall never know until we begin to devote greater attention to agriculture before any seeds are sown and after the crops are in.

     As stated above, I shall focus on post-harvest activities to make my principal point about the “systems” approach.  This said, it is nonetheless important to note that we still need far more work on pre-planting activities of all kinds.  Although we know a good deal about the types of inputs employed in many agricultural regimes at various points in time, we know far less about the specific manner or terms of their acquisition or employment, let alone about the economic costs and contributions of the same.  Whether we are speaking of land clearing, farm implements, fertilizers, tenure systems, credit arrangements, etc., it would help us immensely if we could gather sufficient qualitative and quantitative information to establish rough orders of magnitude about their relative roles in given agricultural regimes.  To give but one example of what I have in mind in this regard:  land clearing was the largest or second largest component of domestic capital formation in the United States from the early colonial period until at least the time of the Civil War (1861-1865).  Once part of the capital stock, cleared land, in combination with other resources, could be rendered into a flow of economically useful agricultural products or, if sold, could provide financial resources available for mobilization in other sectors of the economy.[10]  The number of American historians, even agricultural historians, who appreciate this is absurdly small.  And American historians are likely not alone in their disregard of such important socio-economic facts.

     But back to post-harvest activities, the main business at hand.  Let’s start with crop storage.  Practitioners in the field of agricultural development--and, of course, farmers themselves--have long acknowledged the importance of effective crop storage in the determination of profits or losses in any given year.  Indeed, estimates of the proportion of harvested crops lost “in the barn,” so to speak, are often exceptionally high.  In the case of rice, for example, D.H. Grist, one of the modern era’s greatest experts on the subject, has estimated that as much as 50 percent of the harvest was often lost to rats, insects, vermin, spoilage, etc., in rice-producing countries in Asia even in the second half of the twentieth century.[11]  Given all of the attention that scholars over the years have paid to crop yields, one would have thought that somewhat similar attention would have been paid to storage.  Unfortunately, this has not been the case.  For whatever reason, scholars, generally speaking, have largely avoided the subject.  Even in Europe and China, where a good bit of research has in fact been done on the history of crop storage (grain storage in particular), the literature on storage methods and facilities pales in comparison to that on yields, farm tools, rotation schemes, and the like.[12]  This may be due in part to the dearth of historical sources, but I suspect it is also due in part to a relative disinclination amongst scholars to study a topic that might be considered prosaic and commonplace.  In the case of the U.S., virtually the only time grain elevators--one of the great innovations in any “systems” account of agriculture--make history texts is when their owners are involved in litigation with farmers over the cost of storage.  Dryers, silos, and more recent developments in storage technology (such as the sealing of harvested grain in vacuum-hermetic PVC membranes) receive even less attention.  Why don’t more people realize that there can be a vast difference between gross yields and net yields, depending largely on the relative effectiveness of post-harvest storage technology?  Not merely in times past, but even today!

     Despite its monumental importance, food processing receives equally short shrift in what might be called the “master narrative” of modern agricultural history.  To be sure, processing generally receives perfunctory coverage in the specialist literature on individual agricultural commodities, and business historians have studied discrete processing industries such as grain-milling, sugar-refining, cheese-making, beer-brewing, and meat-packing among others.  What is lacking is an appreciation of the utility of integrating production and processing into general histories of agriculture, and a woeful neglect amongst agricultural historians of the ways in which innovations in food processing have affected, or, better yet, enhanced the material well-being of food producers and consumers alike.[13]

     In the leading textbook on the history of American agriculture, for example, there is no index entry for “food processing” (or any similar term) and only a few references to processing activities scattered inside.[14]  But over the past century-and-a-half innovations in processing--and in the closely linked area of food-preservation technology--have arguably been at least as important as innovations in production technology in shaping and conditioning farmers’ lives.

     Time does not permit a full account here of such innovations.  Suffice it to say that in the modern era, processing, which, simply put, entails the transformation of “relatively bulky perishable commodities into products that are more palatable, nutritionally dense, stable, and portable,” and which, in so doing, increases the value/weight ratio, was itself the object of revolutionary transformation.[15]  This is true pretty much across the board, that is to say, regardless of processing sector or product group.  When coupled with closely inter-correlated innovations in food preservation--the ability to store foods safely over increasingly long time horizons--via innovations in various heating, cooling, drying, chemical, microbial, and irradiation methods, the transformation alluded to above becomes more revolutionary still.

     No food so defines the modern era as meat, so let us begin our discussion of processing and preservation here.  Over the past two centuries, as per capita income gains and economic development more generally have spread to more and more parts of the world, meat consumption has grown dramatically in many areas, including in East Asia.  Such consumption was facilitated in large part by innovations great and small in processing and preservation, innovations ranging from mechanized disassembly lines and conveyor belts to brine injection technology.

      One innovation was by far the most important, however--mechanized refrigeration--which transformed, indeed, made possible both the modern meat industry and many other sectors of the modern food industry, most notably, the commercial dairy industry and products such as ice cream, cheese, and milk.  Despite the importance of this innovation to agriculture, one would be hard-pressed to find much mention of refrigeration or, more to the point, the refrigerated railroad car, in standard histories.  And virtually no one today outside of the small, esoteric field known as the history of technology has heard of John Gorrie, James Harrison, Alexander Kirk, or Ferdinand Carré, the pioneers of refrigeration in the mid-nineteenth century.

     Another innovation in preservation proved instrumental, if not vital to the meat industry, and important to the dairy industry as well: sterilization by bottling or canning, heating, and sealing.  This innovation occurred during the Napoleonic Wars, largely in response to military exigencies, and is generally associated with Nicolas Appert, yet another figure unrecognized today except by historians of technology.  One could make the case, though, that the food sterilization process Appert (and others) pioneered was ultimately to prove of greater importance to human history than anything the “Little Corporal”--Napoleon--ever did![16]

     Military exigencies—the need to feed large numbers of soldiers on the move or in the field over sustained periods—helped to induce several other key innovations in preservation over the past century and a half: canned milk, dried milk powder, factory cheese, irradiation and modern dehydration technology come immediately to mind in this regard.  And before leaving dairy products, the consumption of which have grown substantially in East Asia in recent years, one should at least mention processing developments such as pumping technology, mechanical churns (and later continuous churns) in cheese-making, Pasteurization, and the development of tests for butterfat content in milk production, and in butter production, the invention in the late 1870s of the centrifugal cream separator by the Swedish scientist Gustav de Laval.[17]

     One must have noticed by now that I haven’t yet touched on cereals and grains, arguably the most vital foodstuffs of all.  Here, too, however, innovations in processing and preservation have played vital roles in shaping and conditioning the production end of this segment of the food system.  To cite but one example: the development around 1840 of metal rollers for grinding grain helped to open up huge new opportunities for wheat farmers in eastern Europe and the United States, areas that produced mainly hard (high-gluten) grains that had proved difficult to mill with earlier stone-grinding technology.  Over time metal-roller milling was to spread to other grains (including rice) and to other parts of the world, often with equally impressive results.[18]

     Nor does metal roller milling exhaust the list of important innovations in grain processing, for we also see continuous improvements in grain-separation technology, faster analysis of mill batches for protein content among other things, and greater control over flour particle size and uniformity.  Not to mention the importance of new categories of foods made from milled grains, particularly breakfast cereals.[19]

     My primary goal in the discussion above has been to highlight the fact that it is a mistake to decouple food production from food processing and preservation.  And if I haven’t yet made my case, let me close the discussion by invoking the name J.R. Simplot, who during World War II developed a new method for dehydrating potatoes for the U.S. military.  He later made use of the method to become one of the richest men in the world.  He did so first by becoming the leading supplier of frozen French fries to McDonald’s (and other fast-food franchises), in so doing, helping to sustain the U.S. potato industry, particularly in Idaho and Washington.  With the income generated in agribusiness, Simplot then financed one of the leading-high-tech firms of the late twentieth century, the computer chip maker Micron Technologies, which is today the largest employer in the state of Idaho.[20]  How’s that for a linkage between town and country, between factory and field?

     As with processing and preservation, so with other post-harvest activities such as  transportation, finance, and marketing.  Numerous improvements and innovations in transportation--and communications, for that matter, which can be seen as the “transportation” of information--have kept the agricultural sector in a state of near perpetual motion over two centuries.  Such improvements and innovations range from the local--a new road, the clearing of a stream for traffic, a foot bridge--to those of global significance such as the invention of the telegraph, the advent of steamshipping, the opening of the Suez Canal, the establishment of long-distance “trunk” railroads (with refrigerated cars), and, more recently, the development of containerized shipping.  Taken together, they have everywhere acted at once to broaden and to tighten markets for agricultural goods, providing new opportunities and, alas, posing new problems to farmers the world over.

     Changes in agricultural finance have proven extremely important as well.  Two hundred years ago, Thomas Jefferson famously wrote that “merchants love nobody.”  Jefferson’s observation may be more or less sound, but so is its obverse, that is to say, “nobody loves merchants,” least of all agricultural historians![21]  Indeed, merchants of all types--money lenders, paddy brokers, bankers, speculators, exporters, etc.--traditionally come under heavy criticism from those that study farmers, who are deemed morally superior because they are producers rather than non-producing  “parasites” living off of the toil and sweat of others.  Nowhere has this been truer over the centuries than in China.  In recent years, though, such negative views have been challenged, which is all to the good.  Proponents of the so-called new agrarian economics--people such as Pranab K. Bardhan, for example--have argued persuasively that the high rural interest rates endemic in many parts of the world in times past may have been “efficient” (rather than “exploitative”) in light of the high information and transactions costs, risks, market failures, and moral hazards in such areas.[22]  Indeed, he and others (including myself) have argued that such rates may even be “efficient” today in many poor parts of the world.[23]  This question is debatable, as is the question of whether efficiency in and of itself is a worthy end in all cases.  The fact that traditional views are being interrogated is heartening, however, for it allows space for reinterpretations of such matters as risk management in agriculture via modern financial innovations such as futures markets, and techniques such as forward contracting, spreading sales, and hedging, techniques that over the past century-and-a half have spread to a wide range of commodities all over the world.[24]  Although such techniques are often disparaged and reviled, in my view they have on balance helped farmers immensely, and are certainly crucial to understanding the modern agricultural world.

     Food marketing, too, has changed radically over time.  The types of marketing channels, the number of specialized intermediaries involved therein, the diversity of food products, the types of tools and media employed in selling the same, and the feedback loops developed among food marketers, food processors, and food producers (including backward integration) suggest strongly that to properly analyze and interpret agriculture we would do well do to adopt the systems approach.[25]

     Clearly, all I could hope to do in this paper is to sketch out in rudimentary form an argument in support of a more integrated way to studying the history of agriculture over the past two hundred years.  For purposes of illustration, I have stressed the importance of post-harvest activities, but I might just as easily have written instead on pre-production activities such as the development of commercial fertilizers, whether natural or synthetic, and their acquisition by farmers, as well as innovations in seed and animal breeding, including most recently, of course, the development and employment of genetically modified organisms (GMOs).  And any truly holistic systems approach would incorporate advances in science such as the discovery of vitamins, the actions of the state, environmental/ecological issues, and food preparation and cooking, indeed, questions relating to consumption more generally.

     An impossible task?  Perhaps, but one worth striving towards, nonetheless.  The food  industry has been the biggest industry in the world for most of the past two centuries, and by some measures remains so today.  Agricultural historians have studied intensely one relatively small part of it, economic and business historians another, marketing specialists yet another.  As a result, we find ourselves in a situation that brings to mind the Indian fable about the six blind men and the elephant, each of whom describes the animal differently by seizing on a different part.

     I collaborate at times with a petroleum economist from whom I have learned a great deal, including the merits of analyzing an industry--in this case, petroleum--as a totality.  Oil people might specialize in questions relating to exploration, drilling, refining, transport, product development, finance, marketing, or trading, but whether one is positioned “upstream” or “downstream,” an oil specialist generally knows much more about the workings of the entire industry than an agricultural historian knows about goings on beyond the pasture or field. 

     The most famous essay in the scholarly literature on marketing and perhaps the scholarly literature on business in general was written by Theodore Levitt and appeared in the Harvard Business Review over forty years ago.  In the piece, entitled “Marketing Myopia,” Levitt pointed out that every industry was once a growth industry, but that many have nonetheless collapsed over time.  He argued--and this is the key point--that in every case of failure, failure came not because the market was saturated for their products, but because of failure of managerial imagination.  For example, in his view the railroad industry did not stop growing because the need for transporting people and freight declined, but because the need for transporting people and freight was not adequately filled by railroads, so was filled by other forms of transportation instead.  Railroad managers believed that they were in the railroad business, that is to say, rather than the transportation business, which kept them from creatively shifting with shifting times.  Levitt applies his argument to a number of other industries, and concludes by stating that an “…an organization must learn to think of itself not as producing goods or services but as buying customers, as doing the things that will make people want to do business with it.[26] 

     Levitt’s argument paralleled that of other important contemporary writers on management--Peter Drucker in particular--and it has remained a staple of management thought in the West ever since.  One need not accept the argument in toto to grant that it does seem germane to some of the problems confronting agricultural history and agricultural historians today.  Agricultural historians have generally assumed that they are in the farming business, and, as a result, have limited their scope of inquiry by and large to questions relating to food and fiber production. That we have learned much from such inquiries there is no doubt. Over the past two centuries, however, agriculture has become increasingly “industrialized,” and, in so doing, it has become increasingly linked to business.  Despite such links, which have made it increasingly difficult to draw the line between agriculture and business, most historians of agriculture continue to hold to their traditional ways.  In my view such stodginess and recalcitrance in the face of change has not only led to a relative decline in interest in the field, but to a relative decline in the insights practitioners are able to provide about their own subject matter.  What I have suggested in this paper is that we, as agricultural historians, are not in the farming business but in the food (and fiber) business.  To use Levitt’s framework, we need to go where the customers are, as it were, and adopt a food-systems approach to our subject.  By linking food production more closely to other aspects of the food business, we shall have a better chance of remaining relevant, viable, and vital over the next generations as agriculture becomes more and more “industrial” and as farmers the world over increasingly recede from the stage.