8
Gamonales, Colonos, and Capitalists

The decades between the late 1850s and 1920 were the heyday of the altiplano livestock estate. In an increasingly hectic drive toward expansion, old and new landholding families came to claim a greater share of the region's resources than they had since the conquest: the proportion of land, livestock capital, and labor held in the hacienda sector reached a peak during the first third of the present century. For this offensive against the Indian peasantry to bring long-term success, the control over the expanded resource base needed to be secure and effective. As hacendados tried to set down stable and profitable relations of production in their expanding domains, imperceptibly slow but no less real developments began to change both labor relations and technical aspects of livestock operations. By the early twentieth century the long-term stability of yanaconaje, with its seigneurial overtones, was giving way to a labor tenancy that was, at its edges, more mobile and took on contractual overtones. Modest capital investments showed the path toward more intensive livestock production. But this was not a transition from feudal haciendas to capitalist livestock ranches. Most hacendados were unwilling, and the few who tried were unable, to change the fundamental traits of the labor tenancy system and convert to modern, capital-intensive methods of production. This traditionalism was to have long-term implications for the power constellations in the region's agrarian structure.

Labor Recruitment

Beginning in the late 1850s, the number of colonos on Azángaro's estates grew rapidly. In 1829, just after independence, at most some 5,600 persons lived as colono families on the province's estates, no more than 14 percent

of Azángaro's rural population. By 1876 this number had nearly doubled to more than 9,800, accounting for just under one-fourth (23.4 percent) of Azángaro's rural population. Between 1876 and 1940 the number of colonos and their families grew more than threefold to 31,651. In the latter year colonos made up 35.8 percent of Azángaro's rural population; their share may have been higher during the 1920s.^[1] The resident population on some of Azángaro's largest estates grew even more spectacularly between the 1860s and 1920s (table 8.1).

This wave of new colonos was the last phase in a process of "fixing" an adequate, stable labor force on altiplano livestock estates that began in the late sixteenth century. For the 350 years before the early 1900s, the number of yanaconas , as labor tenants were still often called even after independence, had been insufficient. During the colonial period hacendados had to rely on supplementary mitayos even for routine tasks such as herding and shearing. As late as the 1860s estates drafted nominally free laborers from neighboring communities.^[2]

Until after 1900 hacendados jealously guarded their colono labor force. Administrators and renters were admonished by the owners not to do anything that might cause colonos to abandon the hacienda; rather, if possible they should procure additional labor tenants. But by the early twentieth century an expanded colono labor force had greatly diminished the need for outside labor. Most estates could now carry out all routine tasks of the livestock cycle as well as their limited agricultural operations with the resident labor force. Hacendados employed only small numbers of outside wage laborers for special projects, such as constructing sheepfolds or digging irrigation ditches.^[3] Until 1920, then, the tendency was toward diminution of old forms of wage labor on the estates.

Some of the increase in the number of colonos was due to population growth among earlier hacienda residents: the children of colonos staying on the hacienda and establishing new households as shepherds. But between 1876 and 1940 population growth on Azángaro's haciendas was nearly three times larger than that of the peasant communities, pointing to the incorporation of people from the outside. The lion's share of these newly incorporated colonos were peasants and their families from adjacent communities whose land the haciendas had acquired. Routinely the notarial bills of sale legalizing these land transfers noted that the sellers had promised "to serve the buyer as yanaconas."^[4] The peasant families stayed on their land, converted into a cabaña of the estate, but now had to fulfill the labor obligations associated with their new status as colonos. Most colono families were incorporated into the estates with their land. At the peak of hacienda expansion around 1910 some thirteen thousand to fifteen

thousand persons belonged to such first-generation colono families, more than a third of the total population of the hacienda sector.

But not all colonos settled in an estate with a family, nor did all stay there for the rest of their lives. By the early twentieth century a complex pattern had developed in which a core of colono families resided in one and the same hacienda for a lifetime, perhaps for generations, whereas other colonos worked on an estate for a few years and then moved on to another estate. On Hacienda Picotani in 1909 six patronyms accounted for forty-one of the sixty-eight shepherds employed by the estate. Fifteen years later, in 1924, only ten of the twenty-seven patronyms from 1909 were still present, but the core six still accounted for more than 50 percent of all shepherds.^[5]

Long-term colono families usually belonged to a few kin groups contributing numerous shepherds to the estate, whereas individual colonos

without strong family ties came and went over the years. Mobile colonos were often unmarried young men with little livestock of their own who switched estates in search for the best working conditions. As many of the small, newly formed haciendas had only scarce pasture resources, and their owners, with little capital, tended to pay the lowest wages and meager food subsidies, fluctuation of colonos was strongest there.^[6] By the early twentieth century, then, a group of colonos—still quite small—emerged who were highly responsive to different labor conditions. They had probably come from landless or land-poor families in the communities and remained poor on the estates. Until the 1920s haciendas functioned as a catchment basin for these rural poor, whose only alternative consisted in migrating to urban or mining areas.

The mechanisms by which hacendados induced peasants to work for them included a complex mix of coercion and economic and ideological elements.^[7] This mix largely reflects the ambiguous bases of land sales by peasants to hispanized landholders, characterized in chapter 6 as a blend between a free contract and imposition by force. Certainly force or the threat of force could play an important role in turning peasants into colonos. Community peasants at times explicitly rejected "being turned into yanaconas," taking legal steps to block the maneuvers of powerful hacendados.^[8] Evidence for such campaigns comes largely from the 1860s and 1870s, when pressures on community resources had not reached a critical level. Rather than a deterioration of their economic condition, they feared a loss of status, suggesting that colonial corporate caste stratification still influenced peasant thinking during the decades before the War of the Pacific. At least until that date peasants judged the condition of yanacona as lower in relation to hierarchies within the communities.

Toward the end of the nineteenth century economic reasons began to play a greater role in peasants' resignation to becoming hacienda colonos. Sale of their land often stood at the end of a series of debts. Dependence on the cyclical wool export trade, tighter political control exercised by local gamonales, and growing scarcity of land in the communities made peasants more vulnerable to debt entrapment strategies. Once a hacendado had succeeded in affecting their land, acceptance of the status of colono, with continued usufruct of at least part of that land, was preferable to staying on as landless peasants in the community or searching for wage labor in distant places. What mattered most to peasants was the continued usufruct of their land; hence their transformation into labor tenants was perhaps less drastic than it seems from our vantage point, fully accustomed as we are to thinking in terms of private property and the sanctity of contract. The peasants-turned-colonos who continued to work their erstwhile land

for their families' subsistence certainly did not conceive of this transformation as ceding absolute dominion over the land to the new patrón.

As Gavin Smith has observed for the central Peruvian sierra, even communities "with a long history of confrontation and independence" were shaped by "an equally long history in which Andean peasants have been expected, in their turn therefore expected to offer services to superiors who often went under kin terms or pseudo-kin terms as taita or padrino/patrón."^[9] Indeed, the fact of being incorporated as colonos into a hacienda did not automatically alter the type of services that the peasant previously had to render in his community. He and his family had been obliged to herd sheep, plant crops, transport goods to market, and work as domestic servants for the governor or the priest, just as they now would be expected to do for the patrón of the hacienda.^[10] Gamonal exploitation was not exclusive to the estates; it was replicated in the authority structure that tied communities to local officialdom. Acceptance of the colono status ideologically signified a switch of patrón, not a totally novel form of subordination.

On the surface, the drive of altiplano hacendados to expand their labor force appears overwhelmingly successful. But a closer look at the colonato , as it had evolved by the early twentieth century demonstrates how fluid, fragile, and unwieldy this labor regime had become. Rather than a broad-based transition to a more tightly controlled regime of wage labor, the struggles over the nature of colonato presaged a new stalemate between the social forces in the rural altiplano.

The Colonato and Paternalism

The settlement pattern on estates closely resembled that of the rest of the province. Even today the traveler has difficulty distinguishing between territories belonging to haciendas and those forming parcialidades. The caserío constituted the estate's nerve center; here orders to shepherds originated, and central functions and labor processes were carried out. By 1900 caseríos had changed little in their physical appearance since the early years after independence. Improvements usually took the form of enlarging the building complex without changing its essential features. More rooms were added around a second or third patio. Some "progressive" hacendados invested in corrugated tin roofs to replace the old thatch. In a few instances special quarters were built for preparing cheese and butter, without, however, any notable investment in modern dairy-processing equipment.

Although many small fincas were managed by the owner, most large haciendas were under the direction of a mayordomo or, after 1900, an

administrador . Until the 1870s most mayordomos were recruited from among trusted shepherd families on the estate itself, but after the War of the Pacific a growing share of administrators were sons of small estate owners and other members of the altiplano's middle strata whose families belonged to the network of friends and clients of the hacendado. In a few cases large haciendas employed natives of Arequipa or other coastal towns or foreigners as administrators. Besides free room and board in the caserío, their remuneration consisted of a salary perhaps ten times that of shepherds and a fixed percentage of the hacienda's yearling sheep, which they could sell or raise on estate pastures at will.^[11] With few expenses on the estate, some administrators accumulated savings that they invested in land themselves. While administrator on the Molina family's important Hacienda Churura in Putina during the early years of this century, the Arequipeño Luis Gutiérrez put together a finca of his own through purchases of land from a nearby peasant community.^[12]

Hacendados usually left their administrators considerable freedom in the daily management of the estates, partly because they had limited understanding of the intricacies of livestock raising and partly because the wealthiest among them increasingly preferred to reside in fine houses in Puno, Arequipa, or Lima, pursuing political or administrative careers, engaging in liberal professions, or tending to business interests. They personally took charge of the disposition of the hacienda products, of land purchases and sales, and of other legal matters; but the administrator controlled the timing of major tasks in the annual livestock cycle, the assignments and hiring of shepherds, and the rudimentary ledgers registering the livestock capital and accounts with the colonos. Inevitably all kinds of conflicts developed between owners and administrators. The belief that "the administrators did not have any interests in the prosperity of the hacienda and cared only for their prosperity" was widespread.^[13] Where a relation of personal trust between owner and administrator did not prevail, the position could become volatile. During his first five years as owner of Hacienda Picotani, the Arequipeño Manuel Guillermo de Castresana hired and fired four administrators.

At times an administrator could function as lightning rod for the owner. In his daily contact with the shepherds the administrator would discipline them, at times with cruelty, while in the estimation of the colonos the hacendado remained the taita (father), the final arbiter over right and wrong, the dispenser of protection. When the administrator was socially close to the Indian colonos, he could become a broker who might transmit the interests of the "hacienda community" to the owner. In 1925 one mayordomo penned the complaints of colonos, who could not write, against

the administrator of Picotani, an act that promptly cost the mayordomo his job.^[14] Mayordomos helped organize the celebration of the haciendas' patron saint and partook in the devotion to them. "I have promised to become ccapero [one of the officials] for the fiesta, I am going to serve our Lady of the Rosary in Toma and am daily asking her blessings for your health, my Sr. Patrón," wrote Pedro Balero, the mayordomo of Hacienda Toma, to its owner, Juan Paredes, in 1874. "I sure hope you'll also come to see our fiesta."^[15]

Below administrator and mayordomo every hacienda employed a number of subaltern officials drawn from among the shepherds. The quipu , "generally the shepherd with the greatest knowledge and best behavior," was directly in charge of many livestock operations—selection, classification, assignment of pastures, slaughter, and wool clip.^[16] The quipu continued to work as shepherd, leaving the care of the flock to his wife and children. He received a salary about twice that of shepherds and rations of food and stimulants during his presence at the caserío; he was also assigned some laborers, perhaps young sons of other shepherds, to harvest his fields. Haciendas further employed rodeantes , sons of colonos who still had not received a flock of their own and normally lived in their parents' household. Paid rations and a salary up to twice that of shepherds, they took orders from the quipu. A type of circuit riders, they guarded moyas and external hacienda boundaries from intruders and conveyed orders to the shepherds.^[17]

In 1928 Emilio Romero remarked that "generally, the Indian organization [of the estates] has continued. All the positions are Indian and they have their prerogatives and jurisdictions within the territory of the hacienda."^[18] A certain similarity with the hierarchy of offices in the Indian communities may indeed be noted. Even though quipus and rodeantes were handpicked by the administrator or owner, they had to enjoy the respect of most shepherds to be effective. In Picotani, for example, all three quipus in 1909 belonged to the preeminent kin group, the Mullisacas.^[19] In contrast to most large estates on Peru's coast, altiplano livestock haciendas had not yet developed a clear separation between "white-collar" employees, identified with management, and manual workers. The day-to-day functioning of the estate rested on the smooth cooperation of a hierarchy of Indian supervisory officers mediating between owner or administrator and the colonos, just as communal officers did vis-à-vis district authorities.

The colono family formed the basic unit of production of the altiplano livestock estate.^[20] They lived far removed from the caserío in the center of their cabaña or tiana , the sector of pastures assigned to them for herding hacienda flocks. Families were rarely switched from one cabaña to another,

and some, on the older estates, may have lived in the same spot for generations. Their residences, one or more low, windowless, thatched-roof adobe cottages, differed in no way from those of community peasants.

Once or twice a year the head of the colono family was assigned a cargo , the flock of hacienda livestock that he would drive to his sector of pastures. He would be responsible for the flock's well-being until he had to render accounts six to twelve months later. In the meantime supervision by the hacienda administration was sporadic. Once every week or two the quipu or one of the rodeantes would visit the cabaña, check whether the estate's flock was doing well, and pass along instructions about upcoming tasks and events in the caserío.

Besides the year-round work of herding, the colono family was obligated to provide labor for a broad range of other tasks during the annual production cycle. The head of the household or an adult son had to participate in the wool clip in February and March and the slaughter in June and July at the caserío, the two major annual labor tasks lasting up to a month each. At least one family member had to take part in the agricultural labors of the estate: plowing, sowing, weeding, and harvesting tubers, quinua, and barley, produced on most haciendas as provisions for the shepherds and the owner's urban household and as fodder for cattle and pack animals. They were obligated to transport hacienda products to merchants in towns or to warehouses at railroad stations and to bring back provisions on the return trip. Such trips could remove the colono for a week or more from his own household labors. Colonos commonly had to use their own llamas or mules for this service, without any recompense for loss or damage. Men and women of the colono family had to work as pongos and mitanis in the caserío or the urban residence of the owner for up to one month per year. In that function they were the general-purpose servants, cooking, washing, cleaning, running errands, and fixing things around the house. The colono family had to send one worker to the estate's extraordinary projects (faenas), building or repairing irrigation ditches or fixing the stone walls around corrals and moyas.^[21]

All labor services were unregulated. There were no written contracts.^[22] This vagueness of the quantity and quality of labor obligations presented opportunities for both the labor tenants and the hacendado to interpret obligations in their favor. Conflicts could develop on a wide range of issues. By the early 1990s, an era in which the colonato expanded so rapidly that on many estates the majority of labor tenants had arrived during the previous decade, the weight of custom began to lose strength in the definition of rights and obligations.

The remuneration of the colonos theoretically consisted of three distinct elements: rights to the usufruct of hacienda resources, wages, and gifts and subsidies. The first of these was by far the most important for the colonos, constituting the basis of what Juan Martínez Alier has called their income as peasants.^[23] They had the right to maintain their own livestock (huacchos ) on the pastures of the cabaña they occupied, and they could plant crops for their own family subsistence. Traditionally no limitations had been placed on huaccho flocks and crops of colonos, and this remained the case on most estates in 1920. The huacchos and the colono's flock of hacienda animals were pastured together, convenient for the colono but loaded with problems for the hacendado. It was easy to exchange animals between the two flocks; frequent crossbreeding kept huaccho and estate animals equally scroungy and unimproved; and the size of estate flocks was limited by the simultaneously pastured huacchos, since losses on flocks of more than a thousand animals grew exponentially.^[24]

Huaccho flocks consumed a significant share of the haciendas' pasture resources. The labor tenants' livestock accounted for 25 to 50 percent of all animals maintained within an estate's borders, in some cases even more.^[25] This was perhaps the strongest indication of the limited control of a hacendado over the colonos. Colonos might collectively produce as much wool and other livestock products as the hacendado did. It was on some of the smaller estates, put together since the 1880s, that pastures remained scarcest and the weight of huacchos was strongest. Some small fincas appeared to be little more than agglomerations of peasant estancias loosely held together by the authority of the gamonal owner, who extracted just enough commodities and labor from his "domain" to underwrite his social aspirations.^[26]

With little control exercised by the hacendado, the size of huaccho flocks varied substantially, both within one estate and between estates. They were the most important factor determining the well-being and social status of colono families. On some estates in Muñani and Putina a few families owned as many as one thousand or even two thousand OMR, including herds with hundreds of alpacas, making them no less affluent than kulak families in the communities.^[27] The most prominent colono families, such as those of Guillermo and Aniceto Mullisaca in Hacienda Picotani, had resided on their estates for generations; the clan of the Mullisacas had been members of the ayllu Picotani, which Kuraka Diego Choquehuanca transformed into an estate during the 1760s.

But in contrast to such "kulak colonos" most colono families owned between twenty and one hundred head of livestock; some may not have

owned any. Because well-being depended largely on the strength of their "peasant economy," colonos with little stock and scarce opportunities to trade were poor. "Generally the pay for the shepherd is not sufficient to satisfy the most pressing needs of his family and he finds himself forced to steal the products of the hacienda," commented one agronomist as late as 1932.^[28] Administrators of Hacienda Picotani regularly worried about the estate's "poor colonos" whose families did not have enough to eat. Women shepherds, widowed or single, and young single male shepherds were found most often among the poor on the estates.

What allowed this tremendous differentiation in the ranks of the colonos? We can only speculate. Affluent colonos needed to have access to plentiful pastures from the hacienda and sufficient labor from their kin group. This condition in turn required mutually reinforcing privileges and respect from the hacendado and fellow colonos. The situation of affluent colonos, mediating between the patrón and other shepherds, may have been akin to that of high communal authorities.^[29] Poor shepherds, in contrast, lacked access to sufficient labor power and thus found it difficult to maintain sizable huaccho herds, grow enough crops for family subsistence, and dedicate time to produce craft goods for home consumption and barter. Perhaps their condition was made more permanent by being assigned a cabaña with pastures of limited extent and poor quality.

Wages and subsidies and gifts remained secondary and merely supplementary for the colono families' subsistence economy, yet they had important implications for the nature of the labor regime. Differentiation between wages and gifts and subsidies was largely fictitious before 1920. Wages for yanacona shepherds had been decreed at least since the ordinances of the Viceroy Duque de la Palata in 1687.^[30] By the mid-nineteenth century some estates were calculating remunerations to colonos based on the number of livestock herded, although payment was mostly in the form of subsidies (avíos ) of food and stimulants. Other estates alotted avíos as a fixed monthly amount, without reference to a wage.^[31]

This ambiguity continued into the twentieth century. A slowly increasing number of haciendas nominally paid wages, now calculated on the basis of the length of the colono's service per year.^[32] But payment continued to be largely in chuño, maize, coca leaves, alcohol, and dried meat. The national debate about the "feudal exploitation" of Andean "serfs" led to a series of protective laws by congress beginning in 1916, largely reasserting legislation on the books from the 1820s prohibiting forced labor and requiring payment of wages.^[33] These laws only made the duplicity of altiplano landlords more of an open secret, however, without changing the

basic practice of payment in kind for some thirty years after the first law's passage in 1916.

During the early twentieth century, then, two issues remained confused and unresolved: whether wages would be paid, and, if they were paid, whether they would take the form of avíos or of money. As estates were expanding rapidly and commercial interchange intensified, the very nature of the labor regime began to be at issue. Would the altiplano haciendas move toward rural wage labor or see a continued evolution of the seigneurial, paternalistic regime of labor tenancy? Every attempt by hacendados to alter a particular obligation, privilege, or form of payment of colonos and every demand or defiant stance of resistance by colonos formed part of the evolving definition of this labor regime.

Nominal wages increased between the mid-nineteenth century and the early decades of the twentieth century. During the early 1840s Juan Paredes paid an average of 27.5 pesos annually to the shepherds of Hacienda Quimsachata in Azángaro, most of it in foodstuff and stimulants.^[34] In 1908–9 Hacienda Picotani, which had recently passed into the possession of the Arequipeño financier and businessman Manuel Guillermo de Castresana, paid between 38.40 and 62.40 soles bolivianos annually, equivalent to between 48 and 78 pesos. In 1917–18, at the height of the wool boom, Hacienda Sollocota, owned by the heirs of the recently deceased José Sebastian Urquiaga, still paid a uniform 38.40 soles (presumably bolivianos) annually, or 48 pesos.^[35] The increase in nominal wages since the 1840s amounted to between 74.5 and 157.8 percent. However, with prices for chuño, maize, and coca leaves rising by more than 100 percent during this period, in real terms Picotani's highest wages of 1908–9 lay only some 50 percent above those paid in Quimsachata in the early 1840s, while the lowest wages lay below those of the 1840s.

Disbursements of wages in the form of goods or cash were made on an irregular basis, as the colonos needed them, although on some estates a pound of coca leaves and an arroba of maize were distributed to each shepherd every month. Once a year, in September, after the slaughter and the recount of the estate's livestock capital, the administrator settled accounts with each shepherd. Advances drawn during the previous twelve months were added up, and the shepherd then received the remainder of his or her annual wage. If the shepherd had drawn more than the allotted wage, the difference would be carried over to next year's accounts. On Hacienda Sollocota, in 1918, the colonos on average took more than half of their annual wages at the September settlement of accounts, mostly in the form of chuño and other foodstuffs. Although the timing and com-

position of their pay varied according to family size and the strength of their own peasant economy, cash disbursements were insignificant for most. The Sollocota data suggest that the major function of wages was to help tide colono families over during the six months prior to harvest, when food reserves from their own previous harvest were running low.^[36]

Advances included not only goods or money received by the colonos toward their annual wage but other debts to the hacienda as well. The major source of such debts concerned fallas de ganado , the animals from estate flocks for which shepherds could not account at the annual recount in August or September.^[37] Unless the colono brought the hide or carcass of a dead sheep to the caserío shortly after it had died, he or she was held responsible for its loss. Its value would be subtracted from the yearly wage, and the colono was expected to pay with wool or a sheep from his or her own flocks as compensation. Fallas were a constant bone of contention between colonos and hacienda administrators. Shepherds claimed, with good reason, that on the open range losses to wild animals or rustlers were unavoidable. Administrators and owners, perhaps with equal justification, were convinced that colonos took better care of their own animals than they did of those of the estate and occasionally slaughtered hacienda sheep for their own consumption. They also accused colonos of appropriating hacienda lambs for their own flocks before they could be marked and of handing over dead huaccho lambs as belonging to the estate flock.^[38]

Debts from fallas could be considerable, and estates found it difficult to collect them from the shepherds. Between September 1907 and September 1908 the administrator of Hacienda Picotani charged colonos with losing 846 sheep, mostly lambs from one day to six months old. Some colonos were responsible for up to 100 fallas. The shepherds disputed most fallas, and Picotani's administrator Adrian Fischer was concerned that if he did not manage to take huaccho animals in recompense for the lost hacienda sheep, the shepherds would "laugh [at me] and never comply [with orders]." One year later Fischer had received only 274 sheep and nine soles bolivianos cash in payment of these fallas.^[39]

On Picotani colonos were regularly in debt not only for fallas but also for extraordinary avíos, such as the gallon cans of alcohol that they bought during festivities. The administration used such debts to gain access to the colonos' own wool production. In special ledgers, separate from the wage accounts, the administration entered the amounts of wool equivalent to the value of such debts, according to fixed barter ratios between alcohol, maize, or falla stock and wool. For example, for one can of alcohol (probably a gallon) the shepherd would owe two quintales of sheep wool. On October 1, 1909, the end of the agricultural year 1908–9, fifty-six colonos of

Picotani owed the estate 118 quintales of sheep wool, 4 quintales of alpaca wool, and 12.5 pounds of llama wool. Two well-off colonos with substantial huaccho herds owed about 11 quintales of sheep wool each, whereas colonos whom the administration classified as poor owed from a maximum of 4.5 quintales down to 12.5 pounds of sheep wool. Only ten of the colonos employed at the time owed no wool. In two cases the debts had been contracted during the preceding agricultural year, 1907–8. Although one of these cases involved the substantial amount of over 8 quintales sheep wool, Fischer considered them as "lost in the balance for 1908, deducting them from the profit for that year." Such debts, then, were to be paid with the wool clip from the following year; if still unpaid after twelve months, they were written off.^[40]

"You cannot imagine how much labor it is costing to get the indiada [Indian shepherds] to pay up the fallas and wool debt," Fischer wrote to the owner, Manuel Guillermo de Castresana, on August 30, 1908. "I have given them until the tenth of the coming month to cancel their wool accounts and bring in the fallas." Three weeks later, on September 20, less than half of the wool debt had been paid up in one sector of the estate. Fischer found that much of the credit granted by his predecessor was recorded inaccurately and amounted to absurd charges, with several colonos disputing their debts.^[41] It was unlikely that he would be able to collect the totality of the substantial wool debts arising from credits granted during the preceding year. Two years earlier his predecessor had managed to collect just over half of the amount Fischer now hoped to collect.^[42]

Debt peonage was not a feature of labor tenancy in altiplano livestock estates by the early twentieth century. Hacendados systematically used credit to force peasants to relinquish their land and become colonos on their estates, but once this end was achieved, credit served more limited goals in the correlations of power within the estate. Some hacendados used it, with limited success, as a lever to gain control over the labor tenants' own wool production. More generally, credit was required to stabilize the livelihood of colono families and thus increase their willingness to stay on the estate year after year.^[43] In the altiplano the hacienda store, the notorious tienda de raya of Mexican cereal estates, was unknown. To be sure, some hacendados used the exchange of goods with their colonos as an extra source of income, charging dearly for foodstuffs and paying little for the colonos' wool; in effect such hacendados replicated the relations between wool traders or urban shopkeepers and community peasants.^[44] But other hacendados charged their colonos at cost for maize, coca leaves, and alcohol and paid prices for colonos' wool on a par with those paid by itinerant traders in the communities or in district capitals.^[45] Debts were not used

systematically as a tool to retain colonos by force; rather, owners and administrators were constantly bothered with the difficulty of collecting debts. As José Sebastian Urquiaga, owner of Hacienda Sollocota, wrote in 1916, the shepherd could "withdraw from the estate at his convenience; all he has to do is turn over the hacienda flock as he received it."^[46]

Hacendado control over the labor and resources of their colono families was strictly limited. Indeed, the colonos' strategies to assure their families' subsistence involved virtually the same range of activities as those of the community peasants. The colono's wife and minor children fulfilled the most time-consuming obligation toward the estate, herding the sheep entrusted to them. They perhaps bore the brunt of any increased work requirements imposed by the transition from community peasant to colono. The father was free to pursue more autonomous income-earning activities and to cultivate social ties with fellow colonos and friends and relatives in nearby communities. Colonos marketed a wide range of rural crafts, the most important being textiles; pottery, hats, and other items were also marketed. But their premier commercial activity concerned sale or barter of wool and other livestock products from their own huaccho flocks.

Even hacendados who sought to capture these goods to enhance the trade with their estate's direct "demesne" production failed to cut off the colonos' market ties. In Picotani, for example, the hacienda hardly received any of the valuable wool from their large alpaca herds (amounting to 7,741 heads in 1929), and the labor tenants exchanged as much as half of their sheep wool directly with third parties.^[47] Where hacendados sought to force colonos to sell them their livestock products, colonos still found ways to entrust the wool to friends and relatives in nearby communities for later sale.^[48] Itinerant traders, buying up wool, tallow, and hides and peddling alcohol, coca leaves, aniline dyes, and borsalino hats, included stops at colonos' cabañas on their circuits.

Most colonos periodically left the estate to barter or sell their livestock and crafts products for family provisions. One of Picotani's well-to-do shepherds, Benito Mullisaca, on one occasion sold the estate 200 quintales of salt, a crucial commodity for livestock operations; he had brought the salt from Lake Salinas, in the central plains of the province.^[49] Just as in the communities, then, exchange activities of colonos shaded into trade proper. In some cases colonos maintained fields in distinct ecological zones to produce the maize needed to supplement the avíos. Many of Picotani's shepherds left the estate after the conclusion of the wool clip for one or two weeks in April to harvest their maize in the ceja de la selva of Sandia, at times causing considerable delays to hacienda work projects.^[50]

Yet this impressive autonomy of colono households was only one side of the coin. Hacendados sought, with varying degrees of success, to tie their labor tenants into a hierarchical, paternalistic order of authority. As was true for paternalistic authority in most premodern agrarian structures, it relied both on benevolence—generosity and protection—and on malevolence—punishment and withdrawal of favor.^[51] In altiplano livestock estates paternalism was cast in the idiom of Andean asymmetric reciprocity, with its emphasis on festive common labor and gifts, as well as in the idiom of Catholic patriarchy, stressing the concern of the father for the spiritual and material well-being of his children.^[52]

The patrón's generosity and protection was designed to succor the colonos' life from birth until death, at work and in celebrations, on the estate and in their dealings with outsiders, much like an invisible shield. Hacendados became godfathers of the labor tenants' children at the occasion of their baptism. They loaned money or directly contributed to the costs of funerals and, at least until 1880, were asked to give their blessing to marriages between children of colonos. Hacendados sought amicable relations with parish priests so that they would attend to the spiritual needs of the hacienda's residents and officiate at baptisms, marriages, funerals, and the annual celebrations for the estate's patron saint.^[53]

When all colonos worked together at the caserío or in the fields, they expected and received a copious, warm meal and perhaps a handful of coca leaves and a swig of alcohol. During the slaughter some estates allowed their colonos to keep part of the innards, reason enough for the whole shepherd family to attend this task. Colonos in charge of cows were allowed to keep the milk one day every week. And on certain occasions the administrator distributed gifts in the name of the patrón, a vara or two of cloth, for example. When colonos conducted business outside the confines of the estate, transporting hacienda products to town or the railroad station, exchanging goods for their own household economy in markets or the ceja de la selva, they were under the protection of their patrón. The hacendados used their influence and connections, as well as their lawyers, to free colonos grabbed by the Guardia Nacional or the army on their recruitment sweeps and to get them out of jail when accused of minor infractions. When colonos were attacked or livestock was stolen from them, the hacienda administration pressed charges and sought restitution on their behalf. Conflicts between labor tenants were routinely settled by the hacendado rather than before courts.^[54]

But such protection and favors were always conditional and could be withdrawn at the whim of the hacendado. Administrators could give troublesome colonos difficult or time-consuming assignments, impound

animals from their huaccho flocks, or fail to intervene on their behalf before the authorities. Some hacendados and administrators resorted to floggings against colonos who had failed "to show respect" or refused to carry out an order. And numerous estate owners and administrators did not stop short of sexually abusing the wives and daughters of their labor tenants, on the estate or when they served as mitanis in their townhouses.^[55] In case of resistance or negligence by colonos "the reprisals are immediate and consist of material punishments and exactions," as the Cuzqueño indigenista Francisco Ponce de León noted. Mayordomos and foremen were well armed, and "with such constant threats, the patrón maintains his real and effective authority over the submissive and defenseless Indian."^[56]

No doubt, colonos viewed their hacendado as a powerful personage whose protection they needed and whose wrath it was prudent to avoid. Many of the more traditional hacendados, the gamonales, adapted the construction of their power to the cultural and material environment of the Andean peasant world. They practiced and championed a traditional Catholicism and showed a certain respect for peasant ritual. Celebrations of the estate's patron saint, weddings, and funerals served to reenact and actualize the paternal ties binding the patrón to "his men." The colonos rendered homage to their taita (father), who in return demonstrated generous affection for his hijitos (sons). Alberto Flores Galindo and Manuel Burga have shown in the case of Azángaro's José Angelino Lizares Quiñones how legends and rumors could imbue a gamonal with an aura of magic, associating him with supernatural events and shamanistic powers. Use of violent force, although vital for such all-encompassing visions of the patrón, constituted merely a tool of last resort in the construction of his authority.^[57]

At the same time hacendados viewed themselves as outposts of modern, European, civilization in a sea of Indian barbarians. In communications among themselves, with their staff, and with public authorities, the Indians were "lazy," "shameless," and "abusive" in their appropriation of hacienda resources and "refractory to learning." One administrator repeatedly referred to the incorporation of community peasants into the hacienda as "conquests." Often, although by no means always, such images of the Indian colono were suffused with a vicious racism. Andrea Pinedo, widow of the uniquely powerful and abusive gamonal Pio León Cabrera, embittered about the brutal sacking and destruction of her husband's Hacienda Hanccoyo in 1917 by colonos and political enemies, a few years later referred to Indians as "cannibals" and "helots," overcome "in the lair of their ancestral vices by slovenliness and their innate apathy."^[58]

There was, to be sure, an overlap between the visions of the hacendado as traditional, quasi-magical lord and as representative of modern civilization. In both constructs it was his right, even his duty, to rule the Indian colonos with a "firm hand." An owner of a livestock estate in Cuzco's Quispicanchis province admonished his administrator in 1904 that he should "correct [the colonos] with toughness so that they mend their ways and cease being rascals."^[59] Both constructs formed part of the traditional hacendado's self-image. Fully versed in the Andean peasant world, their language, customs, norms, and codes of behavior, he felt superior to them by dint of his education (even if slight), his modern life-style, his social intercourse with the notable citizens of the province, or even birth.

His persona, the Janus-faced construct of his power, was at once a product of the Andean peasant world, one contemporary version of accepted forms of domination and subordination, and the representation of Western civilization, whose mission it was to conquer the barbarian Indian peasants and convert them into "industrious laborers." Authority derived from such constructs, even if all encompassing on the surface, was highly fragile. For the colono the generous and protective taita could easily be unmasked as the exploitive gamonal. Both images were "true." On one level at least, what Eugene Genovese has written of antebellum Southern masters may be applied to altiplano hacendados at the apogee of their power: "The most flattering self-portraits of the slaveholders and the harshest and most exaggerated indictments of their critics have much in common. . . . They were tough, proud, and arrogant; liberal-spirited in all that did not touch their honor; gracious and courteous; generous and kind; quick to anger and extraordinarily cruel. . . . They were not men to be taken lightly, not men to be frivolously made enemies of."^[60]

Paternalistic authority, even if claiming to hold sway across all domains of the subjects' lives, is inherently limited and ill defined. So much more so in the setting of altiplano livestock estates, with their dispersed settlement and labor patterns and the long history of peasant autonomy. In the relation of domination and subordination between hacendado and colono part of the "transcript" was hidden, to use the language of James Scott. Both needed to get along with each other on a daily basis and were aware of the roughly defined customary boundaries of reciprocal obligations and rights. Hacendados could not afford to lay bare their contempt for the Indian labor tenants in their day-to-day dealings as they did when among themselves. Colonos, while "investing" in demonstrations of deference and respect toward the patrón, and, in principle, accepting their subordination, sought to "hide the track" of their resistance and, on most days,

kept anger or contempt felt toward administrator or hacendado to the relative safety of conversations with family and friends.^[61]

Hacendados, eager to stabilize their expanding work force, needed to limit the severity and frequency of punishments and other measures of social control. Picotani's administrator Adrian Fischer, when reporting his success in recruiting new shepherds, noted the dire consequences of neglecting such caution: "Pascual Miranda from Tarucani also has a flock [on our estate] already. That finca increasingly finds itself without Indians, just as Sollocota, and I believe the reason is that both Don Victor and César Ballón apply the stick with cruelty."^[62] And Juliana Garmendia, owner of Hacienda Ccapana, a livestock estate in Cuzco's Quispicanchis province, repeatedly admonished the administrator "to treat the Indians with prudence; I notice that they are desperate because, they say, you beat them, and they want to leave the finca; I am alerting you to this, because that would be against our interest."^[63]

The Economics of the Estate

The altiplano livestock hacienda operated at a low level of productivity, and efforts at improvement remained feeble during the early decades of this century. Most owners sought to maximize the net income derived from their estates by minimizing monetary outlays for production and commercialization.^[64] In some regards such a strategy was exactly the opposite of that of modern capitalist agrarian enterprises seeking to optimize the return on invested capital. The labor regime, with its stress on minimizing the wage bill, was merely one, albeit central, aspect of this strategy. The same low productivity—low capitalization approach characterized the quality and pasturing of the estate's stock, its installations, and its means of transportation.

Most of the sheep roaming altiplano pastures around 1900 were descendants of the Spanish merinos brought to Peru during the sixteenth and seventeenth centuries. Contemporary observers agreed that Puno's sheep population, degenerate through lack of selective breeding, combined many of the characteristics in the species most undesirable for efficient ranching operations: low weight; coarse, irregular, and short fibers; and little wool. To boot, many sheep came in hues from black to shades of brown and gray, "dirty colors" that at times found no market in Europe.^[65] Whereas purebred merinos or Corriedales or South Downs, typical of Australian or Argentine flocks, weighed 60 to 112 kilograms and easily produced 7 kilograms of wool annually, Peru's criollos weighed about 15 kilograms and

produced 1 kilogram of wool.^[66] Obviously these sheep also produced little meat, a less important product for altiplano ranchers. Descriptions of Puno's cattle were no more flattering.^[67]

One of the advantages of large estates was that they could aim at economies of scale through the efficient separation of flocks and herds and their optimal rotation through different pastures. A typical small to mid-sized finca, with a livestock capital of 1,500 OMR and six or seven colonos, perhaps employed one labor tenant each in charge of one cattle herd and one flock of alpacas, undifferentiated as to age and sex. The remaining four or five shepherds employed to watch the stock of sheep would not be sufficient to form optimally differentiated flocks. In contrast, a large estate such as Sollocota, with some fifty colonos, maintained highly differentiated flocks: ewes placed with rams for fertilization; ewes with lambs up to six months old; yearlings separated by sex; rams; ewes between weaning of their last lamb and new fertilization; one- and two-year-old wethers; lambing flocks (pregnant ewes after separation from the rams); slaughter animals (three-year-old wethers and five- to six-year-old ewes). In addition, Sollocota kept fourteen differentiated cattle herds, two flocks of alpacas, one flock of transport llamas, and two colonos in charge of the riding horses.^[68]

Ideally such a careful regime of flocks and herds differentiated according to annual reproduction cycles should have produced certain productivity gains—through higher rates of reproduction and efficient use of pastures—even without any further investments. Thus, ideally, agglomeration of land into large estates should have brought productivity increases. But the gains seem to have been small. In 1909 Picotani's 20,844 adult sheep produced about 2.5 pounds of wool per head, only slightly more than the average for all estates.^[69] No estate could easily escape the complex set of conditions that kept productivity low. They were inextricably intertwined with the logic, or the "rules of play," of altiplano society.

Hacendados pursued a strategy of maximizing their livestock even to the point of risking fodder shortages.^[70] Ranchers relied nearly exclusively on unimproved natural pastures,^[71] and availability of fodder fluctuated sharply from year to year and from season to season. Many estates strove to bring their livestock capital to a level at which the carrying capacity of the pastures would be fully used during the season with plentiful fodder. Toward the end of the dry season, between September and November, fodder became scarce and mortality rates increased. In Sollocota's ledgers the following entries appeared for the animals entrusted to the shepherd Agapito Montesinos in 1906: "October 15, 2 ewes dead (because of weak-

ness [por flacas ])"; "November 19, 2 ewes dead (because of weakness)"; "November 22, 1 ewe and 2 rams dead (because of weakness)"; "December 10, 4 ewes and 1 ram dead (because of weakness)."^[72]

Haciendas set aside some of the best pastures, those maintaining moisture longest, for the dry season and constructed stone fences around these moyas. Some pastures produced fresh fodder months after the last rains in March or April through small irrigation works.^[73] But these moyas had "only a small extension," often insufficient to tide all animals over the long months without rain.^[74] And the shepherds did not always respect the moyas. They had the "shamelessness," lamented one administrator, to put their own huacchos on the fresh pastures, leaving the grazed, exhausted pasture for the estate animals.^[75]

Agronomists and livestock technicians had no doubt that hacendados "preferred quantity to quality." "Don't we see here livestock ranches with numerous flocks of animals that are nearly always skinny and little developed?" the veterinarian A. Declerq from Peru's National Agricultural School asked in a lecture before the Sociedad Nacional de Agricultura in 1907. "What is the cause of this state of affairs? It is only the disproportion between the number of animals and the quantity of fodder."^[76] This choice of "quantity over quality" both expressed and reinforced a vicious circle like those encountered by other Third World economies, from which it is so difficult to escape: because the productivity of each animal was low, hacendados aimed for maximum stock, which in turn perpetuated low productivity. To choose significantly lower levels of stock heightened the danger of invasions of temporarily vacant pastures by neighboring hacendados and community peasants, initiating a downward spiral of hacienda resources: a further need to reduce livestock capital, in its turn inviting further invasions, and so on. Under conditions of high risk, insecurity, and disputed rights to resources, the efficient estate maximized livestock capital, even if this approach lowered productivity.^[77]

This rationality informed the strategy of stock reproduction adopted by most hacendados. The German agronomist Karl Kaerger calculated from the livestock inventory of an estate near Juliaca, a few kilometers southwest of Azángaro province, that only forty live lambs were produced per one hundred fertile ewes annually.^[78] Statistics for several estates in Azángaro confirm these low rates of reproduction (table 8.2). By comparison, on the Hacienda de Hermanas, one of the vast ranches belonging to the Sanchez Navarro family in the state of Coahuila, Mexico, even in 1847 more than ninety lambs were born to one hundred ewes. In the western United States lamb crops of range-herded sheep averaged nearly 80 percent by the early

twentieth century, and those of farm sheep, tended in enclosed paddocks, could be as high as 130 to 150 percent.^[79]

The major cause for low lambing crops on altiplano estates was mortality among lambs.^[80] As many as 50 percent of lambs died during the first six months after birth, most of them during the first few weeks. Another 25 percent of lambs fell to loss or theft.^[81] The lambs were born in the open range, during day and night, during lambing periods that lasted for a month or more. Sheepfolds to protect pregnant ewes and newborn lambs from frosts and hail were virtually unknown. Administrators complained that the shepherds "completely neglected" to care for the newborn lambs or appropriated unmarked lambs for their huaccho flocks. Some lambs were not accepted by their mothers, others were accidentally separated from them, and many ewes were too weak to feed their lambs. Lacking intensive supervision, many of these lambs died of starvation, if they were not too sickly to survive in the first place.^[82]

With such meager results in each lambing period, altiplano hacendados attempted to maximize the increase of their stock by scheduling up to four lambing seasons each year: the Christmas, March, San Juan's Day, and

Lapaca (Quechua for "stealthy") lambings. Ewes that had not produced a lamb at one of the two major lambings, Christmas and San Juan, were again placed with rams three months later. Many hacendados tried to get two lambs per year from each ewe, expecting at least one lamb to die. They were continuing a practice that the Jesuits had already sought to abandon—without clear success—on their ranches two hundred years earlier. Livestock technicians were advocating only one annual lambing season, around Christmas when the weather was more benign. Experience in Australia and Argentina had shown that this was the way to maximize lambing crops, allowing the shepherds to concentrate attention on each lamb and permitting the ewes to recover their strength.^[83]

Stock reproduction was further hampered by the failure to practice selective breeding, by diseases and adult livestock mortality, and by the short life span of sheep. Not only did most hacendados fail to introduce new blood through the purchase of purebred or improved rams, but they also neglected inbreeding through selection of the best ewes and rams from their own flocks. Ewes were culled only because of old age, thus keeping unproductive and infertile animals that depressed the lamb crop. Some 10 percent of adult sheep died annually on most estates because of both malnutrition and diseases.

Bacterial infections were rare in the dry and cold altiplano climate, but parasites such as the ovine ticks (garrapata ), scab mites (sarna ), and various species of worms attacking the sheep's lungs and stomach were endemic, stunting growth, reproductive capacity, and the weight of fleeces. In 1924, after nearly twenty years of striving for improvements on Hacienda Picotani, one of the altiplano's most "reformist" estates, its new administrator was shocked to discover that nearly all its sheep were infected with up to four varieties of worms in addition to being "covered with garrapatas." He warned the owners that if nothing was undertaken immediately, much of the stock might be lost.^[84] A number of widespread poisonous weeds, especially the "zenkalayo , of mortiferous effects for the animals," contributed to the precarious health conditions of many flocks.^[85]

"The tough and nutrient-poor fodder" that dominated Puno's pastures meant that sheep had to be slaughtered at five years, "at most at six, but sometimes even at four years," because they had lost their teeth.^[86] This meant that ewes, first served by rams at eighteen to twenty-four months, could be used only three to four years for reproduction. With annual lambing crops ranging from 20 to 50 percent, each ewe on average produced as few as 0.6 and no more than 2 surviving lambs during its entire life span. These were dangerously low rates of reproduction, which toward the lower end of the range did not guarantee replacement of culled sheep. For

example, the mid-sized Hacienda Huito in Santiago de Pupuja, belonging to Dr. Alejandro Cano Arce, judge on the bench of Puno's superior court, in May 1896 had two flocks of old ewes and two-year-old wethers, altogether 287 animals, which would be slaughtered in July. At the same time the estate had 326 lambs born since the previous June. With a comparatively high lambing crop of 52.2 percent Huito barely managed to increase its stock by less than 3 percent over the year.^[87] Lambing crops below 40 percent could be insufficient to replace existing stock.

The stability of estate livestock capitals was precarious in the altiplano. On average, the region's aggregate sheep, alpaca, and cattle population grew at a slow rate of perhaps 3 percent annually during the century after independence. But many estates saw their stock decline over several years, as did peasant smallholders. Lax management of the flocks, instability of borders, endemic livestock rustling, years of drought, sharp frosts, and hail, a livestock epidemic—all could disrupt the precarious balance of flocks and herds. A downward spiral could result, at the end of which small and mid-sized estates remained without livestock capital and the largest estates lost up to two-fifths of their capital during a span of several years. Such unstable, downward-spiraling livestock operations explain the frequent cases of estates sold without stock or with stock considerably below their pastures' carrying capacity. The need to restock was common for the precarious, low-productivity economy of altiplano estates.

Multiple annual lambings required frequent and highly complex movements of sheep between various flocks. Under these conditions rudimentary accounting and lax maintenance of livestock ledgers made optimal disposition of the flock nearly impossible. "The majority of estates," wrote Vicente Jiménez in the first practical guide for altiplano ranchers, published in 1902, "keep only sporadic annotations without system or order. The owners are . . . satisfied with lambings irregularly registered by the mayordomo, or with a verbal report by a mere quipu and at times even with the declarations by the shepherd himself, never complete or precise. As a result the livestock capital remains stationary, if it does not suffer losses."^[88] Multiple lambings, scarcity of fodder, diseases, high mortality rates, the pervasiveness of livestock rustling, and border insecurity tended to cancel out productivity advantages theoretically achieved by large estates through economies of scale.

But low productivity was not identical to inefficiency. Under the precarious, low-productivity conditions of the altiplano livestock economy, efficient operations made the difference between flourishing estates with high income-earning potential for the owner and decaying, disintegrating estates. Efficient livestock estates were characterized by a stable or growing

livestock population making full use of available pastures; enough colono families to minimize the need for external labor; enough food production for supplying the colonos; well-established marketing networks, including cheap sources for food supplements (maize, coca leaves, alcohol), preferably through barter for the estate's own surplus crops (potatoes) and livestock products; control and supervision of the labor force through at least minimal bookkeeping; and, last but not least, safe and stable borders.

The two major threats to efficient operation of seigneurial livestock estates were posed by rising costs of production and by disintegration. Costs reached critically high levels if the estate failed to produce sufficient crops to feed its resident labor force, if it needed to hire outside labor on a permanent basis, if flocks could not be kept stable and needed to be replenished through purchases, or if disputes over land required expensive litigation. Estates were threatened by disintegration when they were considerably understocked, when control over shepherds was especially weak, and when serious disputes about hacienda lands arose either within the owners' family or with neighbors. All these factors, often interrelated, could cripple estate operations and lead to a serious decline of its income-generating capacity. The owner might be forced to sell, or the hacienda might shrink in size or become atomized. Clearly then, a great difference existed between efficient and inefficient seigneurial livestock estates. But the criteria for efficiency had little to do with capitalist, profit-optimizing modes of operation.

Available sources do not allow us to calculate the rate of return on altiplano livestock estates with any degree of confidence. It varied according to the conjunctures for livestock products and from one estate to the next. Returns depended on the estate's efficiency, on the type of tenure (that is, whether the hacienda was held in fee simple or emphyteusis or was rented), and on the level of debt incurred in the acquisition of land and livestock capital. Large estates usually rendered a greater rate of return than small ones did, not because they were more productive through economies of scale, but because they could more easily achieve the kind of efficiency described above. Their owners were more powerful and could protect their colonos more effectively against abuses by authorities, traders, and the military, crucial for stabilizing a resident labor force. They could use their considerable labor power to fight border incursions and livestock rustlers; on many large haciendas the ratio between resources used for "demesne" production and those given over to the colonos also tended to be more favorable to the owner. Because power was such a vital ingredient in the altiplano's livestock economy, large landowners faced fewer risks of seeing their stock decline or control over land diminish through external and internal siege, with obvious consequences for the rate of return.

The very notion of profit or rate of return renders misleading results for the type of economy prevailing in the altiplano during the early twentieth century. Labor and land were not fully commodified, and the goods produced on an estate did not all enter the market.^[89] Calculating market prices for capital, inputs, and labor, as well as for the output of the hacienda, thus exaggerates the amount of capital invested, the costs of production, and the total value of production.^[90] If, by contrast, calculations are based merely on the money actually paid and received by the hacendado for the various factors of production and for the estate's output, the results for individual haciendas are no longer comparable. (For example, one hacendado may have purchased his or her estate, whereas another may have inherited or agglomerated land through trickery.) Because the notion of rate of return continued to be so inappropriate for altiplano estates, it is not surprising that accounting for profits and losses was practiced hardly at all. What hacendados cared about in 1910—just as they had in 1830, when José Domingo Choquehuanca differentiated Azángaro's social classes according to wealth—were net revenues, the difference between current gross revenues produced by the estate and the current costs of production.

The meaning of the following widely differing estimates of rates of return, then, should not be exaggerated. In the first years of this century the French traveler Paul Walle visited livestock haciendas in the central sierra, where some hacendados had undertaken considerable efforts to improve their estates. In spite of the high mortality and low productivity of the livestock, ranchers there claimed rates of return (intérêt ) of up to 25 percent and not below 10 percent.^[91] Altiplano estates rarely reached Walle's maximum rates of return, except perhaps during the boom years of World War I. During the decade before 1915 they may have hovered around 10 percent. In 1909 Hacienda Picotani was worth about 80,000 soles m.n., more than any other estate in Azángaro province. It sold about 8,000 soles m.n. of wools that year, and additional sales of animals on the hoof, dried meat, tallow, hides, butter, and cheese probably totaled about 4,000 soles m.n. The estate paid wages and salaries, mostly in kind, of 3,498.93 soles m.n. to its permanently employed colonos and administrators. Other expenditures—for livestock medicine, salt, a small number of new stock, legal fees, and incidental wages for special construction projects—were small, perhaps no more than 1,000 soles m.n. Thus, total costs of about 4,500 soles m.n. represented slightly more than a third of the value of production, 12,000 soles m.n. Picotani's net revenues of 7,500 soles m.n. for 1909 would have brought a rate of return of 9.4 percent.^[92]

José Sebastian Urquiaga, the large landholder from Azángaro, calculated that in 1916, early in the World War I export boom, an altiplano hacienda of "ten square miles" (about 25,000 hectares) with a livestock capital of 10,000 head of all types and ages, or 7,500 OMR, and average quality pastures, would have cost between 33,750 and 41,250 soles m.n. He suggested that its "net rent" at that moment, "when wool achieved its highest price due to the European war," would amount to 5,000 soles m.n. annually, a rate of return of between 12.1 and 14.8 percent.^[93] Because Urquiaga sought to downplay hacendado exploitation, this was perhaps a conservative estimate. In any case, such an average hacienda's "net rent" would have increased over the next two years as a result of even steeper increases in the price of wools.

Ten years later, in 1929, when costs had increased while prices had not recovered high wartime levels, Carlos Barreda, an agronomist and spokesman for reformist ranchers in Puno department, suggested much lower rates of return. "Rigorously calculating the production of a mid-sized estate which is managed by the owner, . . . the utility barely reaches 6 percent of the capital employed." Even this low rate of return was possible only by paying "extremely low" wages to colonos. If shepherds were paid the legally mandated minimum wages, "no hacienda would produce a utility or interest on capital," and at a "rational wage" of 30 soles m.n. per month, "most haciendas would go bankrupt." Barreda concluded that the profit of most altiplano estates "is due . . . to the work of the Indian shepherds converted into utility."^[94]

These widely divergent estimates of rates of return, from 6 percent in 1929 to perhaps 15 percent in 1916, in part reflect different conjunctures. More important, however, they demonstrate the imprecision of the measure. With accounting continuing at a rudimentary level in most estates, hacendados simply did not think in terms of capitalist rates of return. It is worth repeating that it was the notion of net revenue that mattered to them, the simple calculation of the difference between the value of the estate's marketed commodities and the cost of production and other regular expenditures. The data confirms for the altiplano the notion well established for Andean estates and haciendas in other parts of Spanish America, that low labor costs were crucial for maintaining large net revenues.^[95] Other costs were even lower. Expenditures for tools, improved stock, seeds for pasture, and building materials were so small that they rarely figure in estimates of hacienda production costs.^[96] Thus, the fixed capital of the estates remained minimal, limited largely to the simple adobe constructions of the caserío.

Under such conditions rising wage bills would have an immediate, devastating effect on net revenue. Greater outlays for labor allowing a more controlled, regularized work process and making use of increasingly specialized skills made economic sense only if they went hand in hand with capital investments, such as fencing and improved stock and pastures, assuring higher productivity per unit of labor. Initiating such a process was out of the question for most hacendados. They lacked sufficient capital, the credit system was not designed to proffer it, and their control over the estate's resources remained strictly limited. Endemic land invasions and livestock rustling by neighboring hacendados and community peasants and the colonos' stubborn defense of their autonomy made it less than certain that improving landholders could reap any benefits from capital investments. Under the circumstances most old and new hacendado families adopted the least risky strategy for increasing net revenues: they expanded their haciendas, ran more livestock on them, and settled additional colono families, while keeping outlays for land, labor, and inputs of production at a minimum. They did so by making full use of kinship networks, clientalism, and especially their strengthened position vis-à-vis large segments of the altiplano peasantry resulting from many peasants' dependent insertion into dendritic commercialization channels and the provincial elite's greater power of repression.^[97]

It would be too simple, however, to identify the low-productivity, "seigneurial" operation of most altiplano livestock estates during the early twentieth century as mere traditionalism, a continuation of how things had been in 1850 or even 1780. Ironically, it was the very structures blocking the emergence of capitalist livestock enterprises that offered the opportunity to hundreds of modest families of mestizos, whites, and even a few affluent Indian peasants to acquire small to mid-sized haciendas. Given the economies of scale in highly productive, capitalized ranching operations, a tidy transition to agrarian capitalism, with secure and well-defined property rights, skilled rural wage laborers, and improved stock grazing on fenced pastures would have resulted in a predominance of latifundism in the altiplano perhaps akin to the vast estancias of Argentina's fertile pampas. The continued and indeed enhanced strength of clientalism, paternalism, and violence as sources of power paradoxically fostered the emergence of middling landholders—owners of small to mid-sized haciendas as well as affluent peasants—rather than blocking it. While locked in a fierce contest over the distribution of resources that reproduced the main colonial divide between Hispanic conquerors and a subordinate Indian peasantry, most gamonal hacendados and Indian peasants joined in re-

jecting attempts to transform the altiplano along the lines of agrarian capitalism.

Attempts to Modernize Livestock Haciendas and the New Stalemate

So far I have stressed how during the two phases of expansion between the late 1850s and 1920 altiplano estates continued to be marked by low productivity and reliance on highly autonomous labor tenants. But after 1900 politicians, intellectuals, agronomists, veterinarians, and even a few practical ranchers wrote and spoke about the need to modernize technical, economic, and social aspects of livestock ranching in Peru. Until the 1920s those not too familiar with the day-to-day problems and the brutal conflicts in the altiplano showed great optimism about the possibility of achieving such modernization in a relatively short time, and they were encouraged by a brief spate of government interest in improving stock raising between 1917 and 1923. The height of the modernizers' enthusiasm was reached during the early 1920s, just when the altiplano haciendas had entered a severe crisis.

In December 1920 Colonel Robert Stordy, a retired colonial officer, delivered a lecture before the Royal Society of Arts in London. He had been tapped by the Peruvian Corporation and President Augusto Leguía to direct an experimental sheep ranch to be established in the department of Puno and had just returned to England after a six-month inspection tour of Peru's livestock raising zones. Stordy fired the imagination of his illustrious audience about the future potential of wool production in this distant land that conjured up "visions of unbound wealth, mystery and fable." "I venture to assert that the breeding and rearing of livestock should prove more valuable to the Republic than its mines," Stordy told the listeners. "In the great Peruvian Cordillera there lies a future pregnant with possibilities; . . . in the practice of sheep husbandry and the conservation and scientific development of the alpaca and vicuña hair industry there are commercial prospects of considerable extent and value." Stordy envisioned the investment of foreign (especially British) capital in altiplano ranches and land, with a core labor force of "a large number of ex-officers and men of the British Army who might be glad to avail themselves of the amenities of an outdoor life, . . . whose habits of discipline and appreciation of work would make them desirable members of the community, and who could . . . do much to raise Peru to that position among the wool and hair producing countries of the world which her natural advantages dictate." All that was needed to set this vast transformation in motion was for the

Peruvian government "to advance her best interests by throwing open large tracts of land for colonization on an extensive scale."^[98] In advancing this ingenious plan to solve Britain's problems with unemployed veterans and develop Peru's economy in one stroke, Stordy failed to appraise his audience of the fact that, at the very moment when he spoke, a bitter conflict over increasingly scarce land was raging between hacendados and Indian community peasants in the region that he wanted opened for colonization. Based on racist stereotypes, his insistence on British (especially Scottish) colonists as shepherds because of their "habits of discipline and work" implied the need to remove the Indian peasants from their land to achieve his bountiful vision.

Others, although not sharing Stordy's racist calls for European colonization, foresaw a no less grandiose transformation of Peru's sheep and alpaca industry. Professor L. Maccagno, a zootechnician at Lima's National School for Agriculture and Veterinary Sciences, reported in 1924 on what he believed to be a growing disparity between world wool production and consumption and concluded that "Peru is the country best equipped to resolve the crisis. . . . No region in the world presents greater advantages than Peru [for increased wool production]." Maccagno believed that massive investments in ranching, both by national and foreign capitalists, was favored by the "existence of many estates of enormous extension" with "excellent conditions . . . for systematizing and improving pastures and stock."^[99]

Confidence in a rapid and profound transformation of Peru's livestock estates into highly productive capitalist enterprises was fueled by the collapse of the wool market in 1920–21. This crisis both demonstrated the vulnerability of the seigneurial estate and the exhaustion of its growth potential and encouraged radical indigenistas to put forth the opposite vision of a rejuvenated Andean peasant community recovering its dominance (see chapter 9). During the first two decades of this century, however, calls for change had a less visionary ring to them, advocating instead gradual improvements toward the desired transformation. To be sure, fundamental critiques of the "semifeudal" hacienda had been voiced even before World War I, and highly productive agrarian enterprises employing proletarianized shepherds were already seen as the solution by many upper-class commentators on rural affairs. But during the heyday of the altiplano estate, politicians and professionals fully expected the hacendados themselves to undertake the required transformation.^[100]

Against such expectations, change in the altiplano came painstakingly slow, and by the late 1920s it was evident that it would not soon touch the great majority of livestock estates. The following discussion of investments

in improved livestock operations and attempts to change the labor regime thus concerns only a handful of haciendas in Azángaro and, out of more than a thousand estates, a few dozen in the whole Peruvian altiplano. A tiny minority of "improving landlords" sought to limit colono autonomy and to increase their reliance on wages, strengthen the administrative-supervisory structure of their estates, and invest in more productive stock, pastures, and tools.^[101]

Encouraged by the intensifying national debate about feudal exploitation on serrano estates, colonos demanded payment for extraordinary services such as pongueaje and alquilas. When such payments were legislated in 1916, hacendados responded by charging yerbaje , a pasturage fee for the colonos' huaccho flocks, designed to cancel out the need for additional cash payments. Although such fictitious wages and countervailing fees lent a more contractual aspect to labor tenancy, they did not change the estates' day-to-day labor process.^[102]

The "improving landlord" needed to be willing and have sufficient capital to increase the shepherds' wages in order to tackle the crucial issue in the transition to modern, productive livestock operations: a significant reduction in the colonos' use of hacienda resources and a complete separation of estate and huacchos flocks. A 1926 report about Hacienda Huacahuta in Melgar province, owned by the Bedoya and Revie Company, clearly indicated this nexus. The present difficulties could be overcome "through the system of enclosed paddocks. . . . Under this new system the Indian will be obliged to pasture his own flock on his own land, leaving the patrón free to improve and develop his hacienda for the benefit of the country and himself. The rancher will be able to pay much higher wages to the Indian, freeing him from the state of semislavery in which he finds himself today."^[103]

After 1906, when Manuel Guillermo de Castresana acquired Hacienda Picotani, he began a slow and cautious campaign to limit the usufruct of estate pastures by colonos. In 1908 a differential wage scale was introduced according to which the poorest colonos earned nearly two-thirds higher wages than the richest colonos, a bonus for running only small huaccho flocks on hacienda pastures. The colonos with the largest flocks of huacchos were obliged to sell some of their slaughter animals to the estate at low prices "in order to compensate a bit for the pastures that they use."^[104] In September 1911 administrator Carlos Esteves proposed to give the estate's critical lambing flocks to the poorest shepherds, rewarding them "according to their needs, with a bit more avíos. . . . They are more careful in shepherding the flocks and have less need for extensive pastures [for huacchos]." Castresana and his administrator systematically favored

poorer shepherds over the entrenched colonos with large huaccho flocks by shifting the composition of their remuneration toward wages and avíos. Esteves suggested to Castresana that they follow this approach "to subdue the rich colonos a bit, or rather to punish them. It would not matter to you if two or three [rich] shepherds get annoyed; it will be much worse if the advance of your interests continues to suffer. The stock of the hacienda will never increase if they [the rich shepherds] keep on utilizing estate pastures and oppressing [sic ] the estate's animals as they do until today."^[105]

Despite Esteves's bravado, Castresana had to walk a tightrope in his endeavor to recover hacienda resources from the colonos and instill greater attention to the estate's livestock in them. As Picotani was expanding rapidly between 1906 and the war years, it needed more shepherds. The richest colonos formed the most stable part of the work force, and alienating them could undermine the stability of Picotani's livestock operation. Although Castresana taxed the shepherds' use of hacienda pastures, he never ordered an actual reduction of huaccho flocks.

The rhythm of change on Picotani picked up only after Castresana's death in 1924, when the hacienda passed into the possession of his nephews and nieces, the children of Eduardo López de Romaña, Peru's president between 1899 and 1903. On the recommendation of the Gildemeister family, owners of Peru's largest sugar estate, they hired a German agronomist as Picotani's new administrator. Having previously worked on a large corporate ranch in the central highlands, this man was in a hurry to make Picotani more productive. Eager to expand the estate's stock, he doubled the average number of sheep in each colono's flock by 1927. He required additional labor from the colonos, increased falla charges against them, and tried to abolish the food supplements. Shepherds unwilling to comply were routinely flogged. The colonos repeatedly protested against this erosion of their customary rights and working conditions to the owners, who warned the foreign administrator about the danger of an "Indian uprising," especially in Picotani where "the colonos formerly enjoyed consideration and good treatment." They ordered him "to change those radical extreme measures of punishment for prudent measures that defend the interests of the hacienda and at the same time are not abusive of the Indians."^[106] But by 1928 the situation had come to a head. More than a third of the shepherds abandoned the estate, animal mortality increased, and the quality of wool and meat declined because of overgrazing. The remaining colonos plotted to blow up the caserío. The owners now agreed that the German administrator had to go, bitterly complaining to the Gildemeisters for recommending a man so inappropriate for the delicate task of directing an altiplano livestock estate.^[107]

By the early 1930s Picotani had managed to recover its work force, achieve more stable flocks averaging about seven hundred animals, and improve productivity per shepherd and per sheep.^[108] But the crucial matter of reducing huaccho flocks and strictly separating them from those of the hacienda had still not been achieved. In fact, huaccho flocks continued to grow, from 11,648 sheep and 7,741 llamas and alpacas in 1929 to 9,971 sheep and 14,722 llamas and alpacas in 1943. In that year the owners, after delaying for over a decade for fear of the colonos' reaction, finally approved a plan to castrate huaccho rams and designate a special flock of hacienda rams to breed huacchos. The goal was to improve the blood of huacchos so that accidental crossbreeding would be less damaging to improved hacienda flocks. During the years until the agrarian reform in the early 1970s, huacchos were reduced to about one-eighth of their number in 1943. Although the number of colonos declined by nearly 50 percent, from 66 in 1935 to 35 in 1969, the estate's stock was reduced by no more than 20 percent, and productivity per animal probably continued to rise.^[109]

As early as the 1920s some colonos acquired specialized skills, for example, sheep shearing or supervising of lambings.^[110] Basic medical service had been provided to the colonos since before 1920, but Picotani established a primary school for colono children only in 1946, more than three decades after the first schools opened in peasant communities.^[111] As the owners sought to push back the colonos' peasant economy and create a more skilled work force, they increased wages, first between 1908 and 1911, then during the late 1920s, and again after World War II. Between 1906 and 1948 nominal wages went up elevenfold, perhaps double the increase for retail food prices.^[112] As cash wages became more important, Picotani's owners had to relinquish their attempts to block direct commercial ties between the estate's work force and outside traders.

Efforts to change the traditional colonato regime, with its high degree of autonomy of shepherd families and their lack of modern sheep-ranching skills, began in earnest only during the years of the World War I bonanza. Colonos resisted these efforts almost immediately, seeking to block any erosion of their autonomy. In 1917, with the bloody attack on Pio León Cabrera's Hacienda Hanccoyo in the mountainous border region between Azángaro and Sandia provinces and a little-known uprising in Hacienda Huasacona, Muñani district, colonos rose up against their patrones for the first time.^[113] During the 1920s they organized militant protests on numerous haciendas in Azángaro and adjacent provinces in Puno and southern Cuzco. Often in contact with the broad movement of community peasants sweeping the southern highlands, the colonos nevertheless pursued their own agenda. They protested against limits on huaccho flocks,

steep fees for hacienda pastures, forced breeding programs designed to change the quality and color of their huacchos, and restrictions on the sale of their livestock products. "Unionist" demands, typical for rural workers, were not absent but secondary. They included limits on the size of flocks, no reduction of food subsidies, and compensation for special services such as alquila and pongueaje.^[114]

Riots, protests, and participation in rebellions by colonos began to block and channel changes in the labor regime. By 1930 the grand designs for transforming the region's sheep and alpaca industry into a high-productivity business based on wage labor had failed. The direction and speed of change on modernizing estates became steadier, slower, and more predictable. The colonato would not be abolished, but use rights of the shepherds would gradually be limited and huaccho flocks reduced and separated from hacienda flocks. The weight of wages in shepherds' total remuneration increased, and colonos acquired special skills. By the 1950s the labor system on modernizing estates had become contractual. On a few estates, such as Posoconi in Asillo, labor unions submitted lists of demands to management in the name of all colonos. There the old paternalism, if it survived at all, degenerated to mere folklore.

This gradually evolving contractual colonato retained some basic features of the old while removing obstacles for highly productive, capital-intensive ranching.^[115] Not unlike other labor-tenancy regimes on modernizing estates, such as its namesake on Brazil's coffee fazendas , it continued to keep the cost of labor down and reduced the hacendado's risks during cyclical crises caused by declining commodity prices, livestock epidemics, or harsh weather. These landlord gains were achieved after decades of gradual, piecemeal changes and then only by a small group of modernizers. Before 1920 they had hardly made a dent in reforming the altiplano estates' labor regime.^[116]

Isolated technical changes preceded reforms of the labor regime on modernizing estates, but their effectiveness necessarily remained limited. Attempts to improve the breed of sheep through the importation of rams from Europe and Argentina began as early as the 1840s, and after the 1890s they became more regular.^[117] But before 1920 they showed minimal results. The few imported purebred animals quickly died from diseases, and no infrastructure existed to keep their improved-breed offspring separated from the great majority of degenerate animals.^[118] Prodded by the experimental sheep ranch that began operating in Chuquibambilla, Melgar province, in 1921, some modernizing estates, among them San José, Sollocota, and Picotani, began to import large numbers of purebred rams from Argentina. By the early 1930s Picotani maintained a few flocks of purebred

Corriedale and Rambouillet sheep and larger numbers of three-fourths and half-blooded sheep; even these animals produced 70 to 150 percent more wool than did the unimproved criollos.^[119] During the 1920s some estates finally altered reproductive cycles to one annual lambing season, devoting special care to the survival of the newborn lambs.^[120]

Beginning in the early years of this century, a few improving hacendados invested in technical facilities, such as sheepfolds, shearing knives or scissors or, on one estate, even a hydraulic machine for the wool clip, sheep dips to combat parasites, and wool-washing installations and presses.^[121] Use of trucks to transport hacienda products to railroad stations, eliminating conflicts with colonos over alquila services, began only in the 1920s. During the mid-twentieth century truck transport revolutionized the altiplano's marketing system, benefiting haciendas and peasant communities alike.^[122]

Some hacendados undertook to increase the efficiency of their estate's administration, indispensable if they hoped to tighten control over colonos and introduce more regulated, "scientific" methods of livestock raising. Instead of the rudimentary ledgers registering avíos and the livestock added to or subtracted from the shepherds' flocks, modernizing haciendas now kept numerous accounts, meticulously recording the disposition and productivity of the different types of stock, the results of the annual wool clip and slaughter, the dispatch of hacienda products to market, remaining products and supplies on hand in the warehouse, the disposition of the flocks, and payments to and debts of the colonos.^[123] Improving landlords hired more quipus and rodeantes from among their colonos. A handful of administrators now had some training in agronomy; others at least possessed basic skills of reading, writing, and arithmetic, a far cry from the semiliterate mayordomos still the norm on altiplano estates as late as the 1870s. It remained rare for an owner's son to study agronomy before taking over the parental estate.^[124] Increasingly, sons of large landholders from the altiplano sought an education preparing them for urban careers, as they were loath to be directly associated with the management of their parental estates, deemed feudal and antiprogressive by public opinion even before the onset of the crisis in 1920.

Technical change came slowest where it most directly interfered with the labor regime. This was true especially of the introduction of artificial pastures and of fencing, measures of strategic importance for increasing productivity yet also inimical to the colonos' autonomy. Around the turn of the century Alberto Gadea, the director of Puno's Colegio San Carlos, experimented with new, more nutritive strains of pasture, and more hacendados planted barley as additional livestock feed.^[125] But these examples did not catch on. Any serious effort to tackle the increasing problem of in-

sufficient feed required fencing in order to protect the young shoots of improved grasses or barley until they matured. Indeed, during the war boom several altiplano hacendados began to fence in pastures, preparing to plant high-nutrition grasses.^[126] They encountered immediate and adamant resistance by the colonos, who understood full well that fences would lead to exclusion of their huaccho flocks from the estates.

As late as the 1960s fences remained exceptional in the altiplano. The great majority of hacendados, of course, had never even attempted to invest in seeds for high-nutrition grasses or in fencing, in part because they lacked funds or credit to do so. The obstacles to augmenting fodder by these means presented the most immediate motivation for hacendados to acquire additional land during the years of rising demand for wool, allowing them to keep larger flocks. Lack of fencing also forced hacendados to keep the size of flocks within conventional limits to avoid loss or predation. With increasing livestock capital, estate owners inevitably had to employ more shepherds.

The Andean Hacienda Revisited

Our current understanding of the Andean estate as it operated between the late nineteenth century and the 1960s owes much to the pioneering work of three economists, Juan Martínez Alier, Shane Hunt, and Geoffrey Bertram.^[127] All three rejected the widespread image of Andean seigneurial estates as feudal or semifeudal enterprises whose inefficiency and apparent waste of resources reflected the social parasitism and lack of entrepreneurial spirit of the hacendados, latifundists who relentlessly and brutally exploited their Indian "serfs." All three authors sought to demonstrate that the persistence of labor tenancy made economic sense from the perspective of both hacendado and colono as long as certain assumptions prevailed. In concluding this chapter, I summarize Bertram's discussion of the hacienda, which draws and expands on the work of Keith Griffin, Martínez Alier, and Hunt, and then highlight how my own discussion goes beyond the 1970s revisionism.

Bertram suggests that because of sociopolitical conditions the opportunity costs of land were lower for the large landholder than they were for the peasant. Andean hacendados attempted to monopolize land in order to manipulate other markets, especially those for labor and credit. Wage labor does not automatically represent a more advanced type of relations of production than labor tenancy does. Because both hacendado and colono were rational actors, the choice between the two systems "rests upon the efficiency and profitability in production for external markets, in the case

of the landowner, and the comparison between the cash wage offered and the value of usufruct access to land for the colono." Labor tenancy will prevail as long as the value of a year's access to a plot of land for the colono is higher than the cash wage for a year's labor on the hacienda, and the value of a year's labor is in turn higher than the opportunity cost of the same plot for the hacendado. Under these conditions both hacendado and peasant are better off than they would be in a system based exclusively on payment by means of a cash wage.

The peasant enterprise, employing the labor of the whole family and redistributing the output, constituted the underpinning of the Andean hacienda. Colonos preferred it to wage labor also because it provided security and employment to all members of the family and a safety cushion in difficult times. "Unless there is full employment in the wider system," guaranteeing "any member of the peasant enterprise employment at the prevailing wage rate," the peasant enterprise will use its members' labor even when its marginal product is substantially below the prevailing wage rate and will tenaciously defend the access to land on which it is based. Bertram, following Martínez Alier, saw proof for this assertion in the fact that expulsion of a colono from a hacienda was often considered punishment.

This system comes under pressure through exogenous change, especially through increases in the price paid for hacienda products. The hacendado will seek to raise productivity and to work more of the land directly, in "demesne." With a sufficient increase in the value of estate lands, the hacendado will try to abandon the labor tenancy regime altogether in favor of wage labor. In an open economy—one in which the landlord does not automatically possess the type of power requisite for keeping his or her opportunity costs for land and labor stable—the balance of class forces ultimately determines the capacity of the landlord to carry out such changes. "The same forces which led the hacienda to seek to expand could also produce a countervailing (though not equal) strengthening of the peasantry's motive and ability to resist." The notorious slowness with which innovations such as enclosures, specialization of labor, new crop types, improved pastures, and new stock breeds have been introduced on serrano estates during the twentieth century was due not to the "feudal" mentality of the large landlords but rather to the strength of Indian opposition. Bertram concludes that the "hacienda has thus been an agent of progress (in the sense of modernization of the rural sector) while the peasantry has been conservative." Peasants calculated that the benefits of modernization would accrue primarily to other groups while they would have to bear the costs.^[128]

How does this analysis fit into the preceding discussion of altiplano livestock estates? In order to stress the economic rationality of the Andean agrarian structure, Bertram and Martínez Alier confound and conflate the issue of the incorporation of peasants into the estate with the issue of alternative labor regimes, labor tenancy or wage labor. The expansion of the hacienda would be explicable by the same shift of relative factor costs as the expansion of hacienda demesne production. Indeed, these authors see both processes as essentially one and the same thing: in reaction to increasing commodity prices hacendados reclaimed old lands of the hacienda to which they possessed titles dating to colonial times but which they had not worked because of a lack of a market. This construct clearly does not explain the great majority of land transfers from peasants to hacendados in the altiplano during the late nineteenth and early twentieth centuries. It allows the authors to disregard issues of ideology and culture involved in the incorporation of peasants and their land into haciendas. From the perspective of the peasants, this process represented a totally distinct and much less cataclysmic change than did any attempt to switch from labor tenancy to wage labor, even though, from the perspective of the hacendado, both could be understood as reactions to changing relative prices.

Most authors writing on the Latin American hacienda, including Bertram, Hunt, and Martínez Alier, assume that estates held land far in excess of productive needs, primarily as a mechanism to pry loose labor from the Indian community sector. But this assertion is difficult to prove, and the case of the altiplano seems to contradict it. Land not regularly used by its putative owner was likely to be invaded by peasants or neighboring hacendados. When the market improved for livestock products after the 1850s, hacendados required both more labor and more land. The acquisition of land cannot be explained simply as a ploy to incorporate more labor into the estate. At the same time, the incorporation of more labor cannot be said to have been the consequence of the monopolization of land by hacendados. In fact, by the late 1910s, when the extraordinary expansion of the hacienda sector had largely run its course, additional labor requirements were, it appears, met less by transfers from the communities to the haciendas and more by shifts of colonos from one estate to another.^[129]

During the phase of expansion the willingness of community peasants to work as colonos on estates should not be explained primarily by hacienda monopolization of land. More important factors were the increasing commercial dependence of peasants on gamonales eager to use their credit and compadrazgo relationships as a means to incorporate both the land and the labor of their clients into new or expanding estates; demographic pressures from within the peasant communities, which did not automatically derive

from hacienda expansion itself; and finally the greater concentration of effective force in the hands of provincial hacendado elites. Altogether the significance of the shift from freeholder to colono, from the ideological perspective of the peasant, was perhaps less dramatic than commonly assumed. When peasants passed into the sphere of hacienda control, together with their families, their livestock, and their land, they were changing one patrón, the district governor or some other authority, for another, the hacendado. Gavin Smith has captured this shift aptly for the central Peruvian highlands by suggesting that hacienda expansion may be understood as part of a community coming under the sway of an hacendado, while not necessarily severing ties with and participation in the trunk community.^[130]

Not only did hacienda colonos and community peasants have more in common and, indeed, show a greater overlap than usually assumed, but they also showed a similar kind of internal differentiation. The minority of colonos with no or very little livestock, and no or very little family labor to rely on, enjoyed considerably fewer benefits from their peasant enterprise than most others did. They were more open to the lures of modernizing estate owners when these sought to shift the balance of remunerations from usufruct rights to wage payments. Poor colonos relativize the notion of Martínez Alier and Bertram that dismissal from the estate was considered punishment. A small number of colonos in the altiplano were quite mobile by the early twentieth century, moving between estates in search of the best remunerations. Martínez Alier's argument that the situation of labor tenants was more favorable than that of community freeholders thus has to be specified both in terms of the period and the specific strata of colonos and types of estates.^[131]

Only the massive transfer of land from the peasant sector to the estate sector between the late 1850s and 1920 created more favorable animal/land ratios in the estate sector than existed in communities of peasant freeholders. But even by 1910 or 1920 peasants integrated as colonos into newly formed small estates often had no more pastureland available for their own flocks than they had controlled previously as community freeholders.^[132]

To understand the resilience of seigneurial livestock estates based on labor tenancy, we need to go beyond a consideration of commodity, labor, and land markets. Changes in these markets were a necessary condition for the transition toward capital-intensive agrarian enterprises using wage labor. But they were never sufficient to bring about these changes. The cost of capital, the high-risk environment, the neocolonial mode of constructing power, and peasant resistance to such power "explain" the reluctance of

owners of seigneurial estates to undertake their modernization even when other relative factor costs favored such change.

Long-term credit for capital improvement projects continued to be unavailable in the altiplano during the early twentieth century, except for small amounts raised locally at usurious rates. Hacendados constantly feared attacks on their livestock, invasions of their pastures, and destruction of their installations. The massive transfer of land from peasant communities heightened these risks. Although the Peruvian state had sought to strengthen the legal guarantees for private property after the 1850s, this attempt proved largely ineffective in the altiplano. The same landholders who demanded guarantees for rural property owners relied on clientalism, the peasants' acceptance of charismatic authority, and, as an ultimate resort, violence to expand their holdings.

What Bertram called "class constellations" were closely connected to this high-risk environment; indeed, such constellations underlay the lack of definition of rights in resources. Most colonos resisted the improving landlords' attempts to banish their huaccho flocks and turn them into rural proletarians, and not merely because their own peasant enterprise generated more income than they could trust wages to do. Wealthy colonos such as the Mullisacas in Picotani, original residents of the place who for more than a century had raised their alpacas and sheep and exercised far-flung trading activities under the umbrella of the hacienda, but also many of the thousands of peasants recently incorporated into estates, did not consider the hacendados to be owners of the land they were working. They accepted the right of the hacendados to exploit the resources of their estates and to their family's labor, as long as the hacendados respected their rights to the usufruct of hacienda pastures and crop lands. The hacienda sector was highly successful in expanding its share of resources, including land, between the 1860s and 1920, but estate owners did not have the same free disposition over these resources that a capitalist entrepreneur enjoys.

It is misleading to speak of the hacendado as the progressive, modernizing element in the altiplano's rural economy of the early twentieth century, and the peasant, either in the community or on an estate, as the conservative force. Most hacendados, like most of the peasants, clung to labor-intensive, low-productivity methods of production not because they were intrinsically conservative but because they offered the most secure way to keep a tenuous hold on resources, especially land. Eventually, a minority of peasants, like their counterparts among the large landholders, sought to invest in more intensive exploitation of their estancias. There was no iron law that condemned them to becoming either proletarians or

increasingly archaic peasants on shrinking plots of poor lands; they were not automatically condemned to exemplify the notions of either Lenin or Chayanov, as David Lehman has remarked.^[133] The blocked development of productivity on the altiplano's haciendas, just as among the community peasants, cannot, in the last instance, be explained in terms of relative factor costs. It was the consequence of a neocolonial society in which rights to resources were contested and fragmentary and in which liberal notions of property and contract did not become fully accepted. As long as this colonial heritage remained strong, hacendados and peasants alike would waver between a profit-optimizing, high-risk strategy of investing into more intensive livestock production and the conventional risk-averting path of capital-extensive production embedded in social relations of hierarchy (patron-client relationships) and solidarity (within networks of family and friends). The seigneurial hacienda and the communal peasant economy formed two sides of the same coin.

The implementation of modern stock-raising methods on a broad scale presupposed their acceptance by the majority of hacendados, colonos, and community peasants alike and required multifaceted changes on individual estates. The successful introduction of such improvements as selective breeding, fencing, artificial pastures, and reductions of stocks in order to achieve optimal productivity per animal implied not merely technical changes but a major reorientation in the relations of production, intimately tied to the system of social stratification. Questions concerning the capital intensity of production were intertwined with the labor regime and the land tenure pattern. Selective breeding made sense only if hacendados could separate their own livestock from that of colonos. Providing sufficient fodder for hacienda livestock implied, at a minimum, limiting access of the colonos' huaccho flocks to estate pastures or even switching to a system of wage labor without usufruct rights. Reducing stock capital presupposed a land tenure pattern with universally recognized and respected borders.

Tied up with the capital-extensive agrarian complex of the altiplano was a whole sociopolitical system of domination, a value system and a life-style that gave even rather marginal hacendados a favorable place in provincial society. A livestock economy in which market forces reign unimpeded would have undermined this world, in which hacendados commanded respect as much for their family origin, social status, public office, and patterns of consumption as for their economic resources.^[134] Introducing a system of wage labor would weaken the clientalistic, paternalistic structure of Azángaro's estates and society at large. Although the most affluent hacendados might have benefited from agrarian capitalism, the majority of estate owners would have seen their social status threatened.

Recently historians writing on large landed estates in Latin America during the "age of export economies" have sought to stress how much the institution changed during the decades between the continent's more intensive insertion into world markets through railroad and steamship and the Great Depression of 1929–32. Simon Miller concluded an article on Mexican cereal estates by arguing that "far from being a 'feudal' anachronism of artificial and foreign origins, the arable hacienda of the Mesa Central was in fact a dynamic and appropriate adaptation to nineteenth century Mexico and capable of significant capital accumulation."^[135] Likewise, Michael Jimenez suggests that before 1930 coffee hacienda owners in Colombia's Cundinamarca state were "travelling far in grandfather's car" by consolidating recently formed or expanded estates and adapting them to changing labor and market conditions.^[136]

We have certainly come a long way since the first third of this century, when indigenista, populist, liberal, or Marxist critics of the Spanish American hacienda decried it as feudal and archaic, an obstinate, anachronistic remnant of a long-passed colonial age. I have tried to show how the altiplano estate of the early twentieth century was the product of struggles over land and labor between various strata of hispanized landholders, community peasants, and labor tenants. It adapted to changing markets, transportation systems, commercial hierarchies, and power constellations in the province and nationally. It would not make much sense to view the altiplano hacienda at its apogee, the boom years between the 1890s and the end of World War I, as a mere remnant of the colonial regime. After all, half of the haciendas in 1920 had not existed a hundred years earlier. Most of their owners did not come from a closed, preexisting stratum of colonial large landholders; rather, they were the recent beneficiaries and protagonists of struggles with the peasantry, a self-made group of hacendados who skillfully and often ruthlessly exploited newly arising commercial and political opportunities.

Yet if we look inside the hacienda, at its labor regime and technical-economic aspects, we cannot avoid the conclusion that change came exceedingly piecemeal and slow, that hacendados adapted to new commercial opportunities primarily by expanding the scope of long-standing practices. Why did altiplano livestock ranchers travel less far in grandfather's car than did owners of Colombian coffee fincas or central Mexican cereal haciendas?

Part of the answer may lie in the nature of the production process. Altiplano ranching was highly decentralized, with different flocks and colonos' cabañas often miles apart. Centralized or centrally coordinated labor processes occupied no more than eight to ten weeks every year. Entrenched decentralized production processes such as these would have been difficult to turn around under any circumstances. Altiplano ranchers

also had a certain amount of bad luck. In contrast to their Colombian colleagues, they were hit with a severe crisis and a period of volatile markets immediately on the tails of the great expansion of their landholdings, leaving them little time for internal consolidation and modernization. Hacendados were also hampered by the high cost of capital and difficulties of transportation, even though by the 1920s more roads had been constructed in Puno department than just about anywhere else in Peru.

But the fundamental problem was the very nature of the altiplano's hierarchical and segmented neocolonial society. By neocolonialism I am not referring primarily to any dependence that might have continued to tie the altiplano, now by strings of commerce rather than sovereignty, to an overseas metropolis. Rather, I am referring to the revitalized strength of a colonial mind-set polarizing society into Indians and Spaniards or whites; to the ambiguity, among all social groups in the altiplano, between clinging to the security in hierarchical or communal associations and taking advantage of new commercial opportunities; and to the willingness, or perhaps the perceived inevitability, of actors among all social groups to rely on force and violence in the pursuit of personal or group interests.

This neocolonialism of the altiplano's late nineteenth and early twentieth centuries socioeconomic and political matrix allowed large landholders to expand the land formally under their control but blocked the universal recognition of property rights and, hence, full rights of disposition for the hacendados. It allowed hacendados to integrate Indian laborers into their expanding domains at minimal costs but blocked the transition toward a wage-labor regime. Neocolonialism facilitated the cycle between the 1860s and 1920 in which hispanized landholders came to control an increasing share of altiplano resources, but it also constituted the basis of Indian peasant resistance against that expanded control.

The practices of paternalism, coercion, and violence, through which the hispanized provincial elite defined the community peasants and colonos as Indian and subordinate, reinforced the Indians' own perception of their identity as distinct and taught them the continued usefulness of communal solidarity and of maintaining their peasant livelihoods. In a real sense neocolonialism informed both the strength of the landlord offensive and the strength of peasant resistance against it. The peak of hacienda expansion nearly coincided with the hacendados' attempts to initiate internal transformations on their domain and the peasants' "internal" and "external" resistance against both. When this resistance was followed, in 1920, by the collapse of the booming market for wools, hacendados and peasants, the minority of modernizers and the majority of "traditionalists" among both groups, entered a new phase of stalemate.