Rural Workers in Spanish America: Problems of Peonage and Oppression

During the past few years students from three continents have undertaken a great deal of new research on the rural history of Spanish America. Some of this research grows out of the current interest in social history which has inspired researchers to ask new questions of conventional sources, but much of the new work rests on the papers and accounting records of haciendas and plantations, some of which have been recently confiscated by agrarian reform agencies. Thus a cycle comes to a close as the Spanish American landed estate is given its death blow and then split open so that we may examine its historical entrails. With the beast laid out, the examining doctors have become more charitable about its long past life and there is no doubt that their work, which is just now appearing in print, will fundamentally change our understanding of rural society.¹

Within this general subject the story of rural working people has been particularly obscure. Lacking direct evidence, what little is known about peasants and peons has been necessarily inferred from the merest scraps in colonial archives and, for the nineteenth and twentieth centuries, our notions have come until fairly recently from the accounts of travelers and general texts. The voice of the historical actors itself is rarely heard. Nowhere is this sort of historical impressionism more apparent than in the discussion of debt, or debt peonage, and its role in agricultural labor. Unlike slavery or the systems of encomienda and repartimiento which were formal institutions with their own legislation and critics, peonage—the term is English and does not appear until the mid-nineteenth century—was usually an informal device of labor coercion. So informal, in fact, that although few would doubt that peonage ever existed, there is now a lively debate over just when and why it did.

I have two aims in this paper. The first is to examine some of the new research on Spanish American rural history. I try to do that in a comparative, or perhaps really in an allusive, way while at the same time placing the question of peonage within the gradual, patchy, and sporadic progression to freer forms of labor over the past two centuries. There are several elements in this problem and I shall try to indicate what the new research does not illuminate as well as those things it does. The second purpose is to suggest a partial explanation for changes in labor use. The context offered for this is the transition from noncapitalist to capitalist forms of agriculture which occurred—and continues to occur—at very different rates under widely varied circumstance. The history of working people involves exploitation and oppression, but it is also a record of give and take, choice and accommodation. My intention, then, is not to indict one group of men but rather to explain a system. By reducing villains to human dimensions, the stark reality of rural structure itself can be seen more clearly.

Questions of Peonage and Oppression

The academic work on which the general notions of peonage are based really begins with the 1935 essay by Silvio Zavala although journalistic and travel reports on the subject appeared much earlier. Woodrow Borah carried the discussion of Mexican peonage further based on his own work and on speculative interpretation of Zavala’s data.² At about the same time (1952), François Chevalier cautiously presented his own findings in La formation des grands domaines au Mexique—which, incidently, became much bolder and unqualified in the English version of the book.³ All of this work which dealt with the sixteenth and seventeenth centuries saw the origin of peonage in labor shortage and as part of an evolution from indirect exploitation (encomienda and repartimiento) which paralleled the “second serfdom” of Eastern Europe. All of the authors emphasized the extremely tentative and even speculative nature of their views and the ambiguity of scanty evidence.

A picture of generalized peonage nevertheless became standard in textbooks. A generation of students and the general public were taught by the forceful prose of L. B. Simpson’s enormously popular Many Mexicos that peonage was universally employed “to secure a cheap and constant supply of labor” throughout Mexico in the nineteenth century. Charles Cumberland also found debt peonage all pervasive, but he believed that the practice reached a peak in the eighteenth century.⁴ Enrique Semo’s recent book repeats the earlier view that peonage became a general practice in nineteenth century Mexico. In the absence of research most studies agree that the Mexican example is applicable to South America. Rodolfo Stavenhagen finds that peonage grew with the development of capitalist agriculture everywhere in Latin America.⁵

But have we been given a correct picture of debt bondage? We must be very careful, first of all, to distinguish between the two terms, for there can be bondage without debt and more commonly, debt without bondage. In places where landowners wield effective police control themselves or through local political leaders, no excuse, not even the legal fiction of debt is needed to bind workers. But where landowners do not have such power, what has been called debt may also been seen as credit; that is, as advances of cash or goods against the promise of future work. It is not enough merely to find cash, or more commonly rations from the hacienda store, marked against the worker’s account to assume that the advance carried coercive power. To establish the presence of a functioning peonage, there must be evidence of the landowner’s ability to restrict workers’ mobility.

About Yucatán and the southeast Mexican lowlands there is no disagreement: labor conditions were harsh; workers were imported by force; debt was systematically used to provide a legal basis for coercion; and plantation owners, aided by local police or the army were able to restrict workers’ movement and tie them to the estates. There is overwhelming evidence on these points although, it should be noted, mainly from literary sources. The combination of strong markets for tropical exports (sisal, rubber, sugar), a labor shortage, geographical isolation, and a progressive state willing to support the planters with force explains the virtual enslavement of masses of Mayas and Yaquis. In the Putumayo region of the Amazon headwaters, similar conditions led to notorious treatment of rubber gatherers at about the same time.⁶

But apart from these extreme cases, the ability of landowners to restrict their workers’ mobility is much less clear. Peru provides a good example of this, not only because the history of labor here is especially well developed, but also because of the way new sources have brought new interpretation. The older reports on working conditions tended toward the muckraking variety of exposé in which writers, outraged by the direction Peruvian society was developing in the early twentieth century, chose the worst and most abusive examples to make a general case. Peter Klarén’s earlier work on the origin of APRA, which provides in chapter 3 a summary of contemporary memoirs and reportage, reflects the conventional view of those sources that the new sugar plantations on the north coast imposed a harsh and cruel system of debt bondage through the enganche system of labor recruitment.

There is no doubt that workers were recruited from among the village and smallholder population of the sierra through the use of cash advances. Agents for the coastal plantations set up recruiting tables in the sierra during village fiestas or the Sunday market and offered from 30 to 100 soles in cash advances which at the going rate in the 1920s paid for 60 to 200 days of future labor. Those who received the cash advance signed a work agreement and the agents usually required a cosigner as well. The workers then went down to the coast to perform such unskilled and arduous work as canecutting or weeding. Their room and board were provided by the plantation and their wages were docked until the advance was paid off. In the early decades (1870s and 1880s) workers usually returned to their sierra homes for at least a few months of the year to tend their own land and renew family relations.⁷

Such, in bare outline, was the enganche system of labor recruitment. The older reports insist that it was harsh and tyrannical. The predominately urban observers found a six a.m. to six p.m. work day “bleak” and “oppressive;” the wage advanced in the sierra was described as a “lure” to entrap the innocent. Most writers who reported on the enganche system assumed that workers could be held for their debt and that once “hooked” into the system they were bent to their task and mercilessly and effectively hunted down if they tried to “escape.”⁸

It is true that the onset of capitalist agriculture attracted a score of outraged critics and also produced outbursts of labor troubles around Trujillo, especially in the years 1910-1920, but a clear understanding of the system itself has been distorted by more than the ordinary volume of ideological fervor and charged language. Recent research carried out in the Archivo del Fuero Agrario is beginning to change the older picture and raise a series of new questions. Where most of the prior studies implied that workers drawn from the families of smallholders and villagers were incapable of learning and each year seemed to stumble drunkenly into the recruiter’s grasp, it now seems likely that they freely and knowingly chose to work on the coast, took advantage of competition for labor, and knew how to drive up the amount of wages that plantation agents had to advance. Solomon Miller found another motive in sierra to coast migration: several men left “expressly for the purpose of changing wives.” The point here is that the closer the new sources enable us to get to social reality, the more there emerges a world of mutual adjustment and accommodation. Labor recruiters, for example, undoubtedly had to deal fairly with potential workers in order to establish a reputation which insured continuing success over the years. Klarén’s new research shows that recruiters got repeat business as their peons signed up year after year and then often asked for additional wage advances to be paid to families left behind in the sierra. Instead of being passive victims, it seems more likely that workers saw their chance and took it.⁹

Just how effective plantations were in holding workers to their obligations once they were on the job is still an open question. Unlike the policy adopted toward docile modern workers (we usually receive wages only after work is done), plantations had to tie up a great deal of cash in prepayment for labor. Certainly once a cash advance was made the plantation did not want the workers to default, and it is clear that attempts were made to keep workers on the job and seek their return if they left owing labor. All of this one might expect in Peru (or Pittsburgh), but it is also likely that because landowners had influence in local politics and courts, they would attempt to push enforcement beyond the limits of the law. Labor contractors in Peru around 1900 in fact were paid approximately twenty percent of the total wage bill to do just that along with their initial duties of recruitment and transportation to the coast. Obviously, the landowner would have preferred to offer the going rate of, say, 80 centavos a day and have the worker appear at the plantation gate and then stick to the job rather than have to pay 100 centavos through the enganchador. The twenty percent paid to the contractor might also be understood as extramural management costs or as the measure of friction or drag in an imperfect labor market. Once population grew and industrial work habits were accepted by the 1940s, the contractor’s role was no longer needed.

Other questions arise in the problem of enganche labor. How important, for example, were ethnic or cultural differences in resistance to or acceptance of wage labor; that is, were less hispanized Indians from the center and south highlands more vulnerable and more easily drawn into oppressive conditions than were the mestizo peasants of the north? Henri Favre’s recent study would certainly not support this, finding that for a variety of reasons the Indians in the Huancavelica region responded readily to the cash advances of labor recruiters. Favre sees the enganche system not as coercive or oppressive but rather as a means of orientation, regulation, and canalization of workers toward the cotton plantations on the south coast. In this case (he is talking about the period between 1880 and 1910), the Huancavelica peasants took advantage of the opportunity offered by recruiters and went down to the coast to work for wages. Many then returned to their communities where the outside earnings helped them preserve and reinforce their original way of life. Favre argues that these particular peasants were responsive because a breakdown in their system of interdependent, vertical agricultural niches forced them to seek work on the coast; consequently their experience may not be applicable to other zones.¹⁰

The degree of competition for workers among plantations, haciendas, mines, and cities is another serious omission in most studies of labor systems. The older reports drew a picture of sierra haciendas in which workers were immobilized by debt. At the same time, we are asked to believe that plantation agents recruited masses of these same people with no apparent objection from the hacendados. Is it reasonable to assume that haciendas and plantations would collude and not compete for scarce labor? Or, if workers were not scarce in the sierra, why would debt peonage have been necessary to hold them? There are local, national, and even international political questions involved here which with few exceptions are ignored. In the case of the Mexican north, Friedrich Katz shows that because of competition from Mexican mines and U.S. industry, hacendados were forced to increase peon wages and offer more generous sharecropping arrangements. In a recent study of another sugar economy in northern Argentina in the 1930s, Ian Rutledge argues that plantation owners bought or rented haciendas in the Jujuy highlands solely in order to extract labor services from the service tenantry or subrenters whom they then forced “with brutal methods” into plantation labor during the zafra.¹¹ Apart from the fact that this seems like an exceedingly expensive way of finding workers—one highland hacienda cost 41,000 pesos—Mr. Rutledge acknowledges that the Indian tenants were in fact free to leave the hacienda had they wanted to avoid plantation labor. A possible alternative to the Draconian interpretation of debt and violence presented by Mr. Rutledge is that given their possibilities, Indian tenants chose to work for wages on plantations and that the parcels of land guaranteed them by the hacienda were actually an incentive offered by the landowners to keep them from migrating out of the region during the dead season.¹²

Settlement patterns, tenure arrangements, the access rural people had to resources, and their own perception of the possibilities before them, are important elements in questions of peonage and oppression. The new research shows that people are able to make choices and bargain for advantage. Obviously the entire system was weighted heavily against ordinary peons and villagers. But men always work within limits. When considering the plight of our distant workers, it is not entirely inappropriate to ask if we ourselves are not bound by Household Finance Corporation or by the stock-sharing or retirement plans of the companies and universities we work for, which recruit us (bind us) through contributions and then threaten to withdraw their share if we try to “escape.” Given the nature of our society, we willingly choose such forms of peonage. The point is, of course, that we really do not know how much oppression or choice existed in rural Mexico or Peru and, unless the proper questions are asked, we will never find out. It is also fair to say that we may never learn even with the right questions.

Juan Martínez-Alier, one of the pioneer workers in the new archival material certainly asks the right questions. Expecting to find that tenants on the huge Cerro de Pasco and Fernandini livestock haciendas in Peru were bound by debt, he found instead that these haciendas were unable to dislodge tenants and, in fact, had little control over the amount of land or the number of sheep the tenants possessed. Debt was not needed to bind workers to the estate; the problem was quite the opposite, to expel them. In only one estate in the south highlands— the records deal mainly with the years between 1920 and 1960—did Martínez-Alier find evidence of debt, and then its use in labor coercion was not apparent.¹³ On these Peruvian haciendas, as on those of highland Ecuador, tenants had higher incomes and a more secure life than did day laborers and, consequently, they would have been foolish to leave the estate. Indeed, as several recent studies have shown, the ultimate threat against unsatisfactory tenants was often dismissal from the hacienda.¹⁴

Laura Maltby’s study of the records of similar haciendas shows that in the 1920s and 1930s, tenants were not exploited at the company store, but rather were sold items imported from Cuzco “at the purchase price or below it.” Missing hacienda livestock were charged against the responsible tenant’s account, but the estate was usually quickly reimbursed by the tenant with sheep or cash. There is no record here of debt nor were there legal or practical obstacles to mobility. Tenants could and did move from one estate to another, taking their flocks and seeking out the most advantageous conditions.¹⁵

The rural world Malcolm Deas describes (Cundinamarca, Colombia, for the years 1870-1910) is one of competition among haciendas for workers. Wages are bid up during the harvest; workers have unrestricted mobility and move easily to higher wages. Money is sometimes advanced in order to attract labor, but debt is not used as a device to hold men to the job. On one particular finca described by Deas, a dozen or so arrendatarios work for the estate and recruit extra workers from other regions to help out. It is a far cry from a smoothly working labor market, however. Day laborers and arrendatarios themselves must be hounded and cajoled to work, and there are many complaints of unreliability and inconstancy. “Even though the wage is high,” the mayordomo says of his workers, “one still must drag them to work and insist that they fulfill their obligations as if one were asking for [unpaid] labor.”¹⁶

Moving to the far south of Spanish America, the reality of landowner-worker relations is perfectly caught in the rustic language of one Rafael Herrera, an hacienda administrator in 1895 in central Chile. An administrator, Herrera tells us, must often be “strong and just;” the job requires much tact and one must “loosen the reins to tranquilize the people and make them understand with good and prudent declarations the way things must be done.” There are many times, Herrera continues, when one must “tolerate any number of demands [from the service tenantry] because on this hacienda there are so many places where a worker can go to live … one fires an inquilino and he says ‘Well, I’m not going’ and when he sees there is no hope of staying, what he does is move to another place or some settlement or along the road and there he mocks the sentence. One complains again to the owner [of the neighboring estate] on whose land he now lives and the man moves again and now besides mocking, he declares himself an enemy … this is why one must put up with impertinences of the workers and before firing one, try to arrange things with prudence.”¹⁷

Finally, let us return to Mexico where the study of peonage began and where the debate over its function is most developed. The tropical lowlands were notorious for harsh working conditions, but the center and north of Mexico were very different and one must be careful not to confuse these regions with the special circumstances found in Yucatán or Chiapas. There is no better guide to the Mexican material than Friedrich Katz whose recent work provides an overview and preliminary analysis. As he is careful to point out, however, few studies of the nineteenth century are based on hacienda records.¹⁸ Since Katz’ 1974 article, several new works have appeared. Two of these deal with years preceding the 1870-1930 period emphasized in this paper, but since they both present useful correctives to conventional views, I briefly note them here.

David Brading’s study of the Bajío (1700-1850), which is based on estate and notarial records, shows how population increase enabled landowners to diminish tenant perquisites in the eighteenth century and move gradually to a system of money rents and wage payments. But Brading inverts the older notions of debt to show that in fact the estates were more indebted to the workers than the other way around. Those workers who did have debts apparently left work with impunity.¹⁹ John Tutino’s research in the Riva Palacio papers deals with the mixed grain and livestock haciendas in the heart of dense Indian settlement near Chalco, where in the early nineteenth century perhaps seventy percent of the population lived in communities. A small number of specialized workers—shepherds, gardeners, stable keepers—were attracted to the haciendas through wage advances and by the “umbrella of security” which the haciendas provided in the form of rations and rudimentary medical care during famine and epidemics. The hacienda books record credit extended, but there is no evidence of debt bondage. Ordinary fieldworkers were recruited by the day or week in nearby villages with the help of local priests and through payment to labor bosses or contractors. These peons were organized into work teams and paid a cash wage which varied with age and experience. These workers, however, often chose to attend the numerous village fiestas or market days instead of working on the hacienda, and in any case they always gave their first priority to their own village plots. The haciendas’ harvesting was consequently delayed, but even though the owners or managers handed out cash advances, paid for and had rations prepared for the workers, cajoled and complained, there was little the hacendados were able to do.²⁰

Ward Barrett’s study of the Atlacomulco sugar plantation in Morelos at about the same time gives a similar picture of the problems in labor recruitment. In the early nineteenth century, hacendados had either to beseech in a “smooth and persuasive tone” to attract Indian workers or else hand out fairly large cash advances. In one case, 2,000 pesos were distributed, but most of the expected workers simply took the advance and never showed up for work.²¹ The research of both Tutino and Barrett undoubtedly reflects the immediate post-independence reversal of a trend toward greater landowner control of production which began in the later eighteenth century and then picked up again in the 1870s.

Some of the most elaborate and richly detailed research on haciendas was published in 1975 by Jan Bazant who had access to a number of private archives in the San Luis Potosí region. His study covers precisely the years—the last half of the nineteenth century—when peonage and landlord oppression were once believed to have reached an infamous peak. Bazant’s work reveals the familiar progression from service tenantry to wage labor. With population increase, the tenants’ positions deteriorated here just as they did a century before in the Bajío. Their real income was reduced through abolition of the special low price of maize sold to them and attempts were made to reduce the number of service tenants. The lack of alternatives toward the end of the century, which was made especially acute by the encroachment on village lands, meant that many tenants accepted reduced perquisites and stayed on.

Nevertheless, Bazant’s sources reveal a very human and varied picture of potosino rural society. Some peons owed money on account to the haciendas, others did not. The haciendas appear to have been unable or unwilling to restrict mobility, workers commonly left the estate owing money, and no effort was made to bring charges or return them. In the 1870s, far from being in debt to the hacienda, the tenants were owed back wages. As the century wore on, the hacienda work force seems to have been economically leveled out and further impoverished.²² The instrument of this was not overt coercion or debt bondage, but the inexorable grind of the market in a society where the less powerful were stripped of independent sources of income and had either to migrate or accept the fact of a reduced life.

In the Mexican north as well as in the south, economic growth in the nineteenth century created labor shortage, but the evolution of working conditions in these two regions seems to have followed quite different paths. In an early article François Chevalier argued that peon-hacendado interdependence, necessary in the face of Indian raids, gave hacienda workers in the north bargaining power and considerable freedom of action. Friedrich Katz points out that the United States provided a refuge for indebted workers and that mines and industry bid up the price of labor.²³ Haciendas in the north offered workers incentives for more output and introduced a number of paternalistic approaches such as medical care, schools, and better dwellings. Thus, where labor shortage in the tropical lowlands led to violence and peonage, in the north because of different circumstances it led to free labor and higher wages. Harry Cross’ recent dissertation based on hacienda account books and letter files from Zacatecas brings into question the conventional notions about peonage in this region. The main variables in the north seem to have been competition, differences in workers’ mentalities, and the absence of control sufficient to restrict mobility.²⁴ To this merging picture, however, there is a recent exception.

Charles Harris’ 1975 book is the most detailed examination ever made of a single latifundium. In contrast to most of the new work, Harris makes a strong case for an extremely oppressive debt peonage in Coahuila throughout the century of his investigation from 1765-1867. Working in the rich Sánchez-Navarro papers, Harris leaves no doubt that permanent workers often owed money to the hacienda as a result of salary advances, charges against their account in the hacienda store, charges for lost animals, fees for clerical services performed by the resident priest, and so on. He is equally detailed and emphatic in his contention that the Sánchez-Navarros were able to bring force against defaulting debtors and make them return to the hacienda to work off their obligations. The lack of quantitative control over the vivid stories of ill-treatment and the clear sympathies of the author make the argument on this point less convincing. As David Brading recently pointed out, one does have the impression that when it comes to peonage and oppression, Mr. Harris “may have encountered what he set out to find.”²⁵

If debt is one thing and bondage often another, how can we understand the common practice of advancing cash or goods to workers, especially when the amount advanced often far exceeds that needed to provide a pretext for coercion? In the case of enganche labor in Peru and Argentina, I have tried to show that plantations were forced to advance as much as six months’ salary in order to attract workers to field labor. The fact that workers could insist on so large an advance suggests that there was competition for their labor and that they had certain cards to play in bargaining with plantations. In place of force and coercion, wage advances more likely testify to the landowners’ lack of extraeconomic power and their need to play by the rules of a new and as yet imperfectly functioning labor market.²⁶ The same thing can be said of the coffee and cacao planters discussed by Deas and Kaerger or of Herrera, the frustrated Chilean mayordomo.²⁷

In the case of tenant or permanent workers’ debt in the more traditional haciendas, Herbert Nickel presents a closely reasoned and cautiously interpreted analysis of hacienda and notarial records in the Tlaxcala-Puebla region during the Porfiriato. Mr. Nickel finds that calpaneros (permanent residents) were given wage advances or credit at the hacienda store. At the same time, semaneros (seasonal workers) were given short-term salary advances as incentive for an agreed upon task (not to hold them). Toward the end of the Porfiriato there was a tendency toward standard work contracts, that is, payment for work done, for all workers. Nickel shows that salary advances or credit to the service tenantry was only one element in a concerted strategy which aimed at the systematic blocking of alternatives in order to obtain a constant and reliable work force. The strategy took into account population size, political strength of the Indian community, capital investment on the estate, and tenure patterns and ran the gamut from usurpation of village lands through religious intimidation to manipulation of bookkeeping. The strategy also included the sale of maize to permanent workers at rates well below the going market price, improved workers’ quarters, and occasional medical care in order to encourage the service tenantry to stay on the hacienda.

Mr. Nickel interprets the extension of credit beyond a certain amount, that is, an amount too great to be assumed by a neighboring hacendado who might wish to hire the worker, as an “indirect wage increase.” In Tlaxcala-Puebla, the daily wage had been kept constant at around two reales since the seventeenth century; but in the later nineteenth, with prices rising, hacendados advanced more credit to loyal workers. This advance could be withdrawn at any time and payment in such a form protected landowners because it did not generate a claim for general wage increases. Workers did want more credit (that is, debt) and in fact, high debt was equated with high status and conferred prestige. Nickel’s research shows that debts were rarely inherited; they were often waived upon the worker’s death, and many tenants were reluctant to leave an hacienda for fear of losing their source of credit in lean times. Service tenants frequently left an hacienda for all the normal reasons: disagreements with foremen, managers, or other workers, or simply for a better opportunity. Although landowners talked about not hiring a tenant who had left another estate, in practice they often did. If the worker who left had a debt against him, the competing hacendado usually reimbursed the previous employer.²⁸

Arturo Warman’s brilliant new book on Morelos fills out the picture of the privileged position of resident workers, the hips de la hacienda, in convincing detail. Usurpation of village lands by the hacienda together with population growth in the late nineteenth century meant that men lined up to obtain permanent work on the estates. They were then eligible for tiny salary advances which were always presented as a favor, an act of kindness and generosity, on the part of the land-owner in order to cement ties of authority and dependence. Along with favors came the implicitly and sometimes brusquely demonstrated threat of expulsion for sloth or disobedience. The resident workers were in any case a small minority of the total work force. Most labor was carried out by village peons hired by the week or task. Here the landowner did not need to advance credit nor bother with an hacienda store. The landless peon stood in line and took what he could get in the form of a money wage: “he was a free worker in the liberal sense of the word.”²⁹

Market Economy and Mentality

How can the new archival findings be reconciled with the older, still widely accepted picture of peonage? The role of debt can probably best be understood by placing it within the larger pattern of labor use and, especially, by considering the interplay between the landowners’ attempts to gain better control over production on one hand and the changing values and attitudes of rural people on the other. In figure 1, the various types of labor are arranged along a continuum according to the degree of direct control which landowners exercised over production. This scheme will serve as the point of departure for the examination of the landowners’ or the demand side of labor systems.³⁰

The first conquistadors and settlers were limited to indirect use of Indian labor since they only specified requirements and then were dependent upon native chiefs or the colonial bureaucracy to provide the labor. As land and its products became more valuable and the settlers were able to control the land, they also had more interest in regulating the labor aspect of production. This was done in a variety and combination of ways in the seventeenth and eighteenth centuries; evolution was neither unidirectional nor did it necessarily proceed in the sequence outlined here. There was overlap and regression of forms; but in general it is apparent that laborers were first attracted to estates through the offer of land or shares of produce.³¹ Under conditions of increasing profit, more labor service was demanded or renters were asked to pay more and in money; often, however, on account and not in cash. Finally, where markets were strongest in their influence, where profitability and calculability were greatest, and where capital was most easily available, landowners usually attempted, not always successfully, to transform labor types toward the wage labor pole of the continuum. Often although not necessarily, a phase was undergone in which intermediaries such as renters or labor contractors and recruiters were needed. This latter phase is especially apparent in the years after 1870 and the still incomplete transition to wage labor has come only in the past few decades.

From approximately 1870 on, a much stronger market was the main force for change in the internal organization of the hacienda system. As demand rises for commodities and then for land and labor, the estates become economically unstable and evolve into different forms, usually through some sort of family-size rental scheme (including sharecropping) and toward wage labor.³² In a broad global survey of rural class relations, Arthur Stinchcombe concluded that family-size rentals and sharecropping are most likely where: (a) land productivity and market prices are high; (b) the crop is highly labor intensive and little mechanized; (c) labor is cheap; (d) no economies of scale other than in labor exist; and (e) the period of crop production is less than one year.³³ To these economic features social and historical explanation may be added. Landowners, for several reasons, may resist the hard work involved in demesne management or in cases of especially uncertain crops or weather, they may prefer to let subentrepreneurs share the risk of farming. And, as mentioned above, landowners may understand that rentals or sharecropping are profitable and consequently see no point in moving to wage labor.

The effect of market penetration can be seen in the new research on Spanish America. Malcolm Deas’ study of the correspondence files of a coffee finca shows that this estate let its land to a dozen renters who put out the coffee trees, weeded and pruned them, grew their own food on estate land, and then were expected to work for the estate itself for two weeks of each month for cash. These renters in turn hired day laborers in periods of heavy labor need.³⁴ As the market for cotton in the Chancay Valley (central coast of Peru) rose after 1900, the most industrious peons were encouraged to sharecrop cotton in the system known as yanaconaje. They were lent money by the estate to buy seed and fertilizers and to hire day laborers from the sierra. In return, the yanaconas serviced the loan and paid their share of obligations in cotton. Erasmo, the yanacón whose biography has been put together by students from the Institute of Peruvian Studies, is typical of the petty entrepreneur who stood between the Peruvian coastal plantation and the precapitalist society of the sierra.³⁵ A similar system is described in great detail by Karl Kaerger’s invaluable two-volume work which, although published over seventy years ago, is included here because he asked the same questions that the new research does. In the coastal cacao plantations in Ecuador around 1900, owners provided capital and incentive to small planters (sembradores) who cleared land and planted new trees and did not receive payment until the trees were mature. These small planters employed day laborers and paid them with credit advanced by the landowners. The informality of the system offended Kaerger’s Prussian sense of order: he believed that in the end no one knew exactly who owed what to whom. The planters stayed around, even after the crop was in, and did odd jobs and occasional tasks for the owner.³⁶

In the Peruvian sierra during the later nineteenth and twentieth century a similar response was made to the rising wool market. Laura Maltby’s work on Picotani, a 54,000-hectare livestock hacienda near Puno, shows that some eighty colonos kept large numbers of their own animals (one owned over 2,000) which they sold independently of the hacienda. In return for grazing rights the estate received forty days of labor. Martínez-Alier describes tenants who were successful and relatively prosperous petty entrepreneurs on the huge Cerro de Pasco and Fernandini haciendas in the 1920s and even up to the 1960s.³⁷

A strong market for agricultural commodities in the Guanajuato mining region and Peruvian demand for Chilean wheat in the eighteenth century also led to systems of rentals and sharecropping similar to those produced by international demand after 1870. Renters farmed three-quarters of the fertile Bajío, and about the same time arrendatario was still the proper term in Chile for fairly well-off tenants who had not yet been reduced to mere service tenantry or inquilinaje.

What we have in all of these cases is an early stage of response to modern markets. The turn to sharecroppers and rental tenants makes sense when we recall that most of the rural people still lived in communities of smallholders, on the move seeking occasional daily work, or in loose squatter settlements in the interstices of various kinds of private property. Most of the rural people could not yet be brought easily into the orbit of the large estate. In the mid-nineteenth century less than one-third of all Bolivian peasants were attached to the haciendas; some seventy percent in Chalco (Valley of Mexico) still lived in communities in the early nineteenth century; and for all of central Mexico the figure is probably roughly the same at this time.³⁸ In Peru, coastal plantation owners must have believed it easier to reach 7,000 miles across the Pacific for laborers rather than attempt to pry them loose from their own sierra communities.³⁹

As demand for more labor began to grow, everyone complained of escasez de brazos (shortage of hands) and not without reason, for both the villagers and the marginal squatter were disinclined to work for the remuneration offered especially if, as in the case of villagers, access to their own land provided an acceptable way of life. If rental tenants and sharecroppers provided the haciendas with more income than before, it was not enough given the enormous possibilities now apparent as world demand grew, rail and steam lowered freight rates, and capital became more easily available. The next step taken by landowners was to apply pressure on the tenants by reducing their perquisites and pressing for greater labor service while at the same time negotiating for a larger part of sharecropping arrangements. Following this, landowners attempted to gain more direct control over production by moving toward wage labor systems.

The problem now presented, from the landowner’s point of view, was that the ordinary rural inhabitant was inclined to sloth, unreliable, lacked ambition, or at least was not yet adequately responsive to wage incentives. For their part, the tenants were reluctant to work more for less, the smallholder had his own and acceptable way of life, and the day laborer—who might himself be drawn from the families of either tenants or villagers—was unwilling to work at the pace and intensity now demanded. Under these circumstances the landowners, usually supported by the national goverment in which they had much influence, devised an entire range of strategies to bring people with precapitalist mentality into the labor market. These strategies were applied until mechanization and the massive demographic growth of recent decades shifted the labor-to-land ratio to the landowners’ advantage and made possible the full development of capitalist agriculture.

An underlying feature of the transition to wage labor was that the large landowners and their agents, and for that matter, the petty entrepreneurs, believed that without hunger or the need to cover expenses, the ordinary rural inhabitant simply would not work for another person. These attitudes are often cited as examples of troglodytic mentality or racist attitudes held by a colonial or neocolonial elite; but in fact, they are very similar to seventeenth and eighteenth-century attitudes in Britain and Western Europe. “Every one but an idiot knows that the lower classes must be kept poor or they will never be industrious,” was Arthur Young’s informed view of the east of England as late as 1771. And as D. C. Coleman points out in the same article, John Law’s hypothetically typical economy in 1705 assumed the existence of 1,000 people idle half the time and another 300 “poor and idle who live by charity.”⁴⁰

We should not forget that until at least 1750 in England and in many places until much later, the industrial entrepreneur considered his labor force largely impervious to monetary incentives, reluctant to work in the way which suited the boss or indeed to enter his employ at all. Eric Hobsbawm quotes Townsend in 1780: “The poor know little of the motives which stimulate the higher ranks to action—pride, honour and ambition. In general it is only hunger which can spur and goad them on to labor.” Even in the advanced economy of eighteenth-century England, wages for the unskilled working class were set only in the crudest way by the market.⁴¹ The purpose of wage-fixing policy was to fix the maximum, not minimum, rates, and Hobsbawm writes that not until the last half of the nineteenth century were the laissez faire rules of the game thoroughly learned.⁴² Before then, wages were fixed by customary standard; skilled workers got roughly double that of unskilled, for example. Employers aimed to hire at the lowest wage—to pay the lowest total wage bill for a given mass of workers—on the assumption that more pay would not lead to incentive for greater output. J. H. Plumb makes the point that it was among the Quaker industrialists in mid-eighteenth century that this attitude first began to change, and it is this growing notion that “free” workers, if treated well, if given incentives and needs, would produce more, that underlies the innovations of such men as Josiah Wedgwood and Robert Owen and the abolition of black slavery.⁴³

Given this understanding of the most advanced capitalist economy in the world, we should not be surprised to find similar ideas expressed among the eighteenth and nineteenth-century peninsular and creole elite. The Bishop of Quito was entirely representative of current opinion when he argued in 1797 that the best way “to get esa gente [that is, the common Ecuadorian peasants] to work, to eliminate their sloth, reduce their drunkenness and erase the memory of their Yncas,” was to insist on taxation.⁴⁴ Innumerable examples of this outlook can be found into the nineteenth century, and although there is no systematic study of either worker or entrepreneurial mentality, it seems to me that it is not until the last third of the nineteenth century—that is, a century after a similar shift had occurred in Western Europe—that there emerges gradually a new attitude toward the common worker.

Given the prevailing notion that men would work only out of necessity, it naturally followed that needs should be created; and this, in the later eighteenth and in the nineteenth century, takes two general forms. In the first place, the colonial devices of tax, clerical fees, and forced distribution of merchandise were increasingly employed in the later eighteenth century. The tribute was the most convenient form of creating obligations and the standard practice was for the entrepreneur, hacendado or obrajero, to pay the cash tribute for his workers while extracting labor service from those liable for the tax.⁴⁵ All of these expedients—tribute, fees, forced purchase of merchandise—were designed to bring the ordinary person into the economy at a time when the subtle hard-sell consumerism of modern capitalism was unknown, the wares of city life had not yet penetrated into the hinterland, and the modern lust for possession had not yet become general. Everyone complained that workers had exceedingly minimal needs and consequently would not work beyond the point where their bare necessities were satisfied.

The other principal element in the precapitalist strategy was to deprive potential workers of alternatives to estate labor, or, to put it another way, to remove restraints in the formation of a free labor market. An early feature of the new strategy was the Bourbon assault on charity or on the kind of welfare which was believed to encourage idleness and sloth. Thus we see in the later eighteenth century an effort to confine the incorrigible, sweep the streets of mendicant indigents, put them into workhouses, and limit welfare and alms to the truly deserving poor. Some pressure was brought to bear on both the Church and individuals to reduce the indiscriminate handing out of alms. This is all standard eighteenth-century fare and its application in America would make an interesting study, but as yet there is no good recent work on Spanish America comparable to Callahan’s on Spain.⁴⁶ It would be easy to overstate the effect of these reforms but we may already detect here some pressure, whatever the magnitude, directed at undercutting the floor of self-sufficiency by blocking the alternatives of workers in order to encourage them to enter the labor market. A much more effective measure was the nineteenth-century liberal challenge to the Church which inadvertently weakened charitable and welfare functions in addition to diminishing its economic and political role. The impact of the liberal reform on welfare has not been measured, but both liberals and Marxists coincide in seeing the reform as progressive. However the social cost, especially its impact on the lower classes who had depended on the Church and on the widespread practice of almsgiving, may have been important. Paradoxically enough, the hacienda itself may have provided a certain measure of social welfare by its tolerance of underemployed residents and their often unproductive dependents.

Much more important in the process of threatening the foundations of independent peasant existence was the absorption of village lands by private haciendas. In Mexico beginning in the 1870s, in Guatemala where the reduction of Indian lands was accompanied by anti-vagrancy laws, in Bolivia where two-thirds of the rural population became dependent upon haciendas, and in fact throughout the Andean spine, the resources and means of independent livelihood of a great many rural people were reduced.⁴⁷ All of this is a familiar story. The process of separating peasants from their independent livelihood, which included access to hacienda land, village lands, or artisanal production, was rarely carried to the point of complete destitution. Peasant resistance was a factor in some cases and, where the state was unable or unwilling to put itself squarely on the landowner’s side, the costs of expulsion or appropriation were too high for individual landowners. Again, there is also a question of social welfare. The pre-1870 hacienda had undoubtedly sheltered, fed, and underemployed far more people than required for production, and because the patron-client relationship “implied the obligation of the patron to assist his dependents … the hacienda probably had a beneficent rôle.”⁴⁸ The more cost-conscious landowners understood that village communities and their extended family networks provided important services for the wage-earning poor. Day laborers on the estates could return to the subsistence sector which fulfilled functions of social security which the hacienda was less and less willing to undertake.⁴⁹

In any case, the more common practice in Spanish America was not the complete destruction of communities—they still survive everywhere—but the reduction of them to the point where diminished resources forced villagers or their dependents into the labor market. And of course it was not just the reduction of village lands, but this combined with population growth which propelled rural people into the labor market. After the 1940s, the scramble for any kind of job by millions of landless rural people eliminated labor shortage and made the previous forms of compulsion unnecessary.

The other side of the question, that is, the change in condition and attitude of rural people, is more difficult to gauge, and the new research on individual estates does not help very much. What people thought or felt about working for others or even for themselves is not recorded, and so far the scholar with an interest in this kind of mentality has not come forward. Workers’ attitudes and emotions still must be inferred either from a careful reading of their employers’ opinions of them or from their own actions.

The change from the natural or seasonal rhythms of ordinary agriculture to the more disciplined and timed labor required in industry or in industrial agriculture has been noticed by students of other societies, and we may begin, to obtain a frame of reference, with the work of E. P. Thompson and others around the journal Past and Present. Thompson describes the natural rhythms of “task oriented” work in agriculture. Nature demands that the grain be harvested before the thunderstorms set in, seafaring people must “integrate their lives with the tides,” sheep must be attended to at lambing time and so on. Thompson then makes three observations: first, that task orientation is more “humanly comprehensible than timed labor” because the “peasant or laborer appears to attend upon what is an observed necessity.” Second, he notes that “social intercourse and labor are intermingled” and consequently there is no great sense of conflict between labor and “passing the time of day.” Third, “to men accustomed to labor timed by the clock, this attitude to labour appears to be wasteful and lacking in energy.”⁵⁰ The innumerable Spanish American travel accounts, written in large part by Anglo-Americans with little rural experience scornfully decrying sloth, are familiar to us all; and in more recent times, other features of this impression appear as a new post-industrial urban generation of young scholars is shocked by the ordinary scene in any farming society of dawn-to-dusk labor. The distortion and misunderstanding work both ways: the peasants are “incurably lazy” when seen in the off-season or on rainy days; and “crushingly exploited” while only working an ordinary farming day.

Thompson’s ideas are not, of course, based on Spanish American experience but much of what he says is applicable to rural or village life everywhere. Robson Tyrer’s study of the Quito farming and obraje economy in the seventeenth century, long considered to be one of the most exploitative and harsh systems in Spanish America, shows that people probably worked no more than half the days of each year in the obrajes and during the rest of the time tended their flocks and farmed.⁵¹ Charles Harris writes that many workers in eighteenth-century Coahuila worked only two months of the year (although the hacienda fed them over a full twelve months).⁵²

Actually, little attention has been paid to the length of the working day, week or year, either by contemporaries or modern academics. But as labor demands were stepped up in the 1870s, reproaches against ‘St. Monday’ (the common practice of taking Monday off) became more frequent; there were more complaints about absenteeism. In the 1880s in Chile, a time of strong market demand, workers even during the harvest season rarely worked more than twenty days of the month. In Malcolm Deas’ 1900 coffee finca, where the modern devices of fines for tardiness and prizes for production were clearly present, the administrator still lamented the lack of discipline or the desire for gain. “I don’t understand these people,” he confessed, “they’re really ‘Indios’; even with a good wage one has to drag them to work.”⁵³ Karl Kaerger, with experience in the agriculture of Prussia and Germany, was scornful of the casual and indefinite modes of labor (Unbestimmtheit) which he thought, “so totally suits the South American character.”⁵⁴

These attitudes should not be dismissed out of hand as mere racism or insensitivity of the exploiter. Rather, we are seeing in all of these cases, at a time when markets and the prospect for gain were strong, a clash between the values of entrepreneurs culturally akin to the time-oriented employers of the industrial world who, especially by 1870, were beginning to insist on reliability, discipline and constancy in their workers.⁵⁵ Opposed to this were those very different values of rural people, villagers, peons, and seasonal workers. The growing number of complaints and the tone of intensity and frustration indicates the mounting problem in this clash of values. I believe that it is the period after 1870, down to the 1930s, where the transition from “preindustrial” to modern attitudes takes place—the analogous shift in Britain is a century earlier—and it is in this context that the role of debt may fruitfully be examined.

A Possible Framework for Understanding Peonage

Let us now bring the various elements of our problem into sharper focus. The scheme below arranges the two principal features in the argument along vertical and horizontal continua.⁵⁶ Along the left side is the range of workers’ attitudes, the varying degree of their incentive toward gain. Implicit here is their changing perception of need and the acceptance and internalization of discipline. Along the top, ranging from much to little, is the degree of direct control over production on the part of the large landowners or rural entrepreneurs. Implicit here are change in capital investment with consequent division and specialization of labor, the landowners’ power to limit or block alternative ways of life for other rural inhabitants, and the landowners’ increasingly rational or profit-maximizing attitude. At one extreme pole in this scheme, where there is little incentive for gain on the part of the workers and little control over production by a rural elite, the historical record generally shows a subsistence economy with little linkage to the market; the rhythm of work turns around seasonal demand. At the other pole where workers are keen to earn ever more and landowners have control (the case say, of California agribusiness, Tucumán sugar, and the Peruvian coast in the 1950s and 1960s), one finds a nearly pure wage labor system with overtime pay and bonus for extra output.

Between these poles are areas of ambiguity and conflict. Where workers emerge with strong incentive toward gain while landowners have but little control, smallholders or villagers may hold on to their land and there are usually varieties of rentals for cash or crop or sharecropping on the estate. The system of yanaconaje which came into existence on the Peruvian coast in the late nineteenth century and endured until heavy capital investment in the late 1940s; the coffee workers described by Deas; Brading’s late eighteenth-century Bajío; and Bazant’s nineteenth century acasillados are all illustrative examples. As landowners increase their control through expropriation of village lands or when they benefit from a shift in the labor-to-land ratio, there is a tendency to reduce renters’ perquisites, increase labor requirements, and eventually move to a wage labor system. Obviously, local or national political support is important in this process; without either, the transition to wage labor remains only a goal. Martínez-Alier’s studies show how unsuccessful the attempt was in a number of highland Peruvian haciendas where neither the Cerro de Pasco corporation nor the Femandini corporation had sufficient control to expel tenants.

The greatest conflict in our scheme comes when the market rather quickly opens enormous opportunities for landowners—through, for example, overseas demand, new domestic markets, the introduction of rail or new roads, or new sources of capital—while there is still a general resistance to wage labor and where precapitalist attitudes are still the rule. As landowners’ control increases, they first, as we have seen, put the screws to the service tenantry and sharecroppers. Increasingly however, others, including villagers and the rootless or casual worker, must also be brought into the production process. How is the landowner to make the wage mechanism work? Advertisements are put in papers, but the word is slowly and imperfectly spread and only a few workers appear at the gates of the new sugar central or sisal plantation. Or, if they do appear, they work only to cover their own needs (defined by them very modestly) and then disappear, or they return for village fiestas or to work their own fields. One measure of how unresponsive or inconstant rural people were to wage offers can be seen in the price paid for Chinese laborers. In Cuba and Peru between 1847-1874, over a quarter of a million Chinese were employed at a cost of $340 to $500 each.⁵⁷ It is at this point—in the interstices between subsistence and rental arrangements at one pole and wage labor at the other—where other devices and mechanisms are employed by the landowners to obtain a work force. The power of landowners to block alternatives and their reluctant but eventual willingness to pay sufficiently high wages for higher quality labor (the United Fruit Company early initiated this practice) did bring workers into the wage labor market.⁵⁸ But where the demand for labor was especially strong and potential workers dragged their heels, more direct measures were used. Landowners hired labor contractors or enganchadores whom they paid as much as twenty percent of the wage bill for recruitment, transportation, and labor management. The contractors often found it necessary to give part of the total salary in advance to potential workers; in other cases landowners extended credit during slack times or charged corn rations, merchandise, or even clerical services against a tenant’s or worker’s account.

Until recent years, most students and observers have assumed that debt meant bondage. Because of this assumption and because the sources then available did not reveal the inner workings of rural estates, we have by and large an unreal and one-dimensional picture of rural society. New research based on sources closer to social reality now provides a basis for reevaluation. An explanatory context of course is still fundamental. To help with this, I have suggested the possibility of examining the change in values which occurred in the transition from noncapitalist to capitalist forms of agriculture.

From about 1870 onward, because of capital investment and technological change there was in many regions of Spanish America a fundamental change in the way men made their living. Alongside the quickening pace of economic life, the majority of rural people still lived a quiet existence, their attitudes toward work shaped by the rhythms of ordinary agriculture and the limited economic horizons of village life. The consequent disjunction between the demand and supply of labor brought about conflict and accommodation. A great many rural people were wrenched out of one social and mental world and ended up in another.

The years 1870-1930 represented a transition which in many ways is analogous to the period a century earlier in Great Britain and Western Europe. During these sixty years, landowners and rural entrepreneurs managed, gradually and incompletely, to tighten their control over production. This was done only rarely with the kind of Draconian measures described in Barbarous Mexico, or by the American envoy in Veracruz, or the novels of Ciro Alegría or Jorge Icaza.⁵⁹ More commonly labor contractors, enganchadores, paternalistic measures, wage increases, wage advances, and in rare cases debt bondage were used to obtain the kind of labor needed by new types of plantations and haciendas.⁶⁰ Reliable and productive workers were demanded at a time when rural people were not yet fully responsive to wage incentive; when, in other words, a free labor market did not yet exist. All of the landowners’ devices were accompanied by only partially successful efforts to block alternatives to wage labor. Most of the new research is consistent in its rejection of debt as a controlling feature of labor. The closer we get to social reality, to the everyday workings of society, the better we understand that rural people are not merely passive victims; rather, they make choices, work out of self-interest. They and landowners alike make compromises and strike accommodations which are often mutually beneficial. To be sure, the world within which the relatively powerless make choices is narrowly limited and, in some cases, the indirect limitation of choice comes dangerously close to direct coercion.

By the 1940s, new labor habits were formed, men were alienated from their work, and the values appropriate to a smoothly working wage labor system became more common. This process is not yet complete. In those areas where labor was not needed, where markets were weak, where capital was not attracted, pockets of rural people were by-passed. But everywhere else we can see the breakdown of community, the creation of a rootless and alienated mass, and the triumph of the consumer society. For having accelerated this integration of rural people into modern economic life, we may thank the modernizing landowner, the effective contratista, and the rural workers’ own capacity for spiritual self-destruction.

Author notes