Chapter 4. Promoting strategic linkages between the farm/non-farm sectors: the Peruvian case - Javier Escobal and Victor Agreda

INTRODUCTION

The central objective of this chapter is to analyse the linkages between commercial agriculture and agro-industry in two of the most important agricultural valleys of Peru. Recent institutional innovations in these regions have altered the relationship between potato, cotton and asparagus farmers and agribusiness. The innovations have centred principally around management experience, a critical factor in producing for agribusiness. The developments have clearly benefited participating producers, but the direct and indirect effects on employment and consumption patterns in the area are less straightforward; local upstream linkages are weak and indirect effects have largely leaked out of the region. Innovation has tended to accentuate income concentration and capital accumulation among those who are already wealthy. The exception - small-scale cotton producers - illustrates the possibility of smaller farmers purchasing management experience. “Unfriendly” linkages between the agricultural sector and agro-industry have otherwise forced small agricultural producers to develop a set of off-farm activities that are not linked with the agricultural activity as a strategy to complement their incomes.

DESCRIPTION OF THE STUDY AREAS

In 1997, a private research centre in Peru - GRADE - undertook two surveys of 30 households each in the valleys of Chincha and Mantaro. Group interviews based on a structured questionnaire were carried out to collect supplementary contextual data from five groups of 15-19 farmers each. These primary data were supplemented with information from the 1994 Peruvian Agricultural Census and the 1997 Living Standards Measurement Survey, and from interviews with inform-ants in the zone. The data collected comprised qualitative information on spatial preferences for input acquisition, contractual arrangements and income sources, and quantitative information on outputs and inputs related to the main crops examined. The survey did not collect information on consumption expenditure.

Chincha valley

The first case study was carried out in the Chincha valley, 250 km south of Lima by the coastal highway. The Chincha coastal zone is one of the most important agricultural valleys of the Peruvian southern coast; it has been linked to the cotton export market for more than a century. The zone has abundant aquifers for irrigation and plenty of flat cropland. It is dominated by the intermediate city of Chincha, whose rapid growth in the past two decades has been based on agro-industry, fishing and non-metallic mining. Chincha is near to other intermediate cities in the sierra and serves as their link to the coast.

Approximately 8 000 farms in this valley cover about 81 500 ha, of which 31 000 ha are prime agricultural land owned mainly by large landholders. Average plot size is 6.4 ha. As can be seen in Table 1, however, more than 70 percent of the farms in the valley have less than 5 ha but control only 12 percent of the land; 24 percent have less than 0.5 ha per farm.

In the Chincha valley, two types of agricultural units predominate: large and medium-sized farms, or fundos, managed by entrepreneurs using modern irrigation techniques, and small-scale farmers, or parceleros, former cooperative associates, which constitute the majority. The cutoff farm size to distinguish between parceleros and fundos is 20 ha; parceleros have on average 6 ha, and the fundos 69 ha.

Compared to the small farmers, large farmers are more commercialized and use more irrigation. Large farmers irrigate 95 percent of their land; small farmers irrigate 70 percent. Small farmers mainly irrigate with river water using the gravity method; large farmers irrigate from wells using the drip method and thus have greater water control but greater capital investment. Large farmers use more chemicals and machinery and rent more of their land; they dedicate more of their land to asparagus - 65 percent for large farmers, 8 percent for small farmers - and less to cotton - 18 percent for large farmers and 78 percent for small farmers. The rest of the small farmers’ land is mainly under maize for home consumption.

TABLE 1
Farm and land distribution by land size (%): Chincha

Source: 1994 Peruvian Agricultural Census.

Home-based non-farm activity is undertaken by 22 percent of farm households. The probability of partici-pation increases with farm size, probably because larger farmers have more cash to meet capital entry requirements. The activities are mainly small-scale process-ing of cheese and yoghurt, machinery rental, commerce and cottage manufactur-ing. Employment outside the home is undertaken by 30 percent of farm house-holds in the Chincha zone. Here the probability of participation decreases with farm size. Small farmers work off-farm in agricultural wage labour and non-farm activity with low entry requirements in terms of education and financial capital; examples are construction, fishing, commerce and transport, as shown in Table 2. Large farmers tend to operate larger-scale and capital-intensive non-farm enterprises and engage in education-intensive non-farm salaried employment. The seasonality of off-farm labour is concentrated during the summer, when off-farm activities peak outside the planting season.

TABLE 2
Share of total off-farm activities: Chincha

Source: 1994 Peruvian Agricultural Census.

Mantaro valley

The second case study was carried out in the Mantaro valley, one of the most important agricultural areas of the Peruvian sierra. The valley comprises two distinct agro-ecological zones. The first is located on the floor of the valley and contains the prime agricultural land and most of the population. Most land in this area is irrigated and used for commercial purposes by farmers and by communities specializing in potatoes, white corn, vegetables and cultivated pasture for livestock. The second agroclimatic zone is located in the middle and upper part of the valley; it is characterized by steep plots, most of which are not irrigated. Agricultural activities are commercial and for self-subsistence. This is the most important area of the valley in terms of production of tubers and cereals - mostly potatoes and barley; it is occupied mainly by peasant communities. The upper part of the valley was historically owned by haciendas (large farm estates) devoted to sheep-raising. These haciendas were expropriated in the 1970s under the agrarian reform and the land was occupied by cooperative farms. Most of these cooperatives have been parcelized during the last decade.

The study area was chosen to include several districts of Huancayo province where over 37 000 commercial farmers and peasants own 308 000 ha. As shown in Table 3, most farms have less than 5 ha and 42 percent have less than 0.5 ha each. Holdings are highly fragmented: fewer than 5 percent of farmers in Huancayo have only one plot, whereas 67 percent have five or more plots. Average plot size is 6 ha, but of these only 1 ha is agricultural land. One third of agricultural land is irrigated.

TABLE 3
Farm and land distribution by land size (%): Huancayo

Source: 1994 Peruvian Agricultural Census.

The city of Huancayo, founded in 1572, is located in the centre of the corridor that connects the towns of Huancavelica and parts of Apurimas and Ayacucho. It is thus an obligatory stop for all agricultural production sent to market in Lima, and for all goods entering these departments from the capital and the rest of the country. This has led to Huancayo’s position as a commercial centre, but the industrial sector - particularly the agro-industrial element - is little developed. Lack of an agro-industrial sector may be a result of proximity to Lima and its wholesale markets, or of the absence of an appropriate agricultural product. The few private and public attempts have failed or achieved limited success, such as dairy, canned and milled products. Private potato processing farms have only recently begun to take an interest in maintaining ties with local producers, a phenomenon discussed in more detail below.

Three districts in close proximity to Huancayo were chosen for study. Huayucachi is located 11 km south of Huancayo; including outlying communities, it has a population of 1 500 households. Farmers in Huayucachi produce potatoes on communal land; they have strong links with wholesalers at the La Parada market in Lima in terms of potatoes for consumption, and with potato producers on the Central Coast in terms of potato seeds. The population has grown in recent years as a result of migration as families from outlying communities sought refuge and protection from the political violence of the 1990s. This migration has furthered the urbanization of Huayucachi in terms of housing and small businesses.

Huanchac, which consists of 480 families, is also located in the floor of the Mantaro Valley, 28 km from the city of Huancayo. Because of the availability of irrigated land, agricultural production focuses on potatoes and vegetables; the microclimate and soils are good, and two potato-processing companies based in Lima have recently started working with producers from Huanchac. The areas of Huayucachi and Huanchac are characterized by good access to public services and infrastructure; the proximity of the city of Huancayo allows for commuting to employment.

TABLE 4
Land, production and market insertion: Huancayo

Ns = not significant.
Source: GRADE field work, 1996-1997.

Aramachay represents the districts located in the upper part of the Mantaro valley. Its distance from Huancayo - 64 kilometres - has resulted in sparse provision of public services and infrastructure. Agriculture, still organized along communal lines, is the most important economic activity; potatoes and barley are the principal crops; families also depend on livestock activities and seasonal agricultural wage labour. As can be seen in Table 4, producers in Aramachay have smaller plots than the other two case-study districts and much less access to irrigated land. Combined with lower yields, Aramachay farmers produce on average only a sixth of the total average production of farmers in Huayucachi and Huanchac. Most of it is sent to local markets or consumed on-farm; farmers in Huayucachi and Huanchac export almost all of their production outside the region.

Although agriculture is the core activity of Mantaro farmers, over 24 per-cent of them acknowledge undertaking non-agricultural activities on their farms; the number of such farmers increases as the size of agricultural assets diminishes. The main on-farm non-agricultural activi-ties are handicrafts and trade. Off-farm activities are also important in the Huancayo Province, especially outside the September-December planting season. The most important off-farm activities according to the 1994 Agricultural Census are listed in Table 5.

TABLE 5
Share of total off-farm activities: Huancayo

Source: 1994 Peruvian Agricultural Census.

The Aramachay district offers few off-farm employment opportunities; this, combined with the long distance from Huancayo, means that most households work and live on-farm. The opposite is true for the other two districts, where urban growth and the proximity of Huancayo provide multiple off-farm employment opportunities and the majority of households do not live on-farm. Even though agricultural production is the principal economic activity for 62 percent of households in Huayucachi, 75 percent also work off-farm. Poor families in Huayucachi work in a variety of activities, including unskilled wage employment, shoe making and making and selling vehicle parts. In Aramachay, 90 percent of households work in agriculture but only 25 percent work off-farm.

INSTITUTIONAL INNOVATION AND LINKAGES

Chincha valley

Cotton

Cotton was the motor of economic and town growth in the Chincha valley for most of the last century. In the past decade, however, the fortunes of cotton have been declining as a result of appreciation of the real exchange rate, competition from liberalized imports of textiles and cotton and increasing input costs for farmers resulting from cuts in government subsidies for inputs and credit. These factors provoked a national slide in cotton production, from 293 000 mt in 1989 to 86 200 mt in 1998 (Ministry of Agriculture, 1999).

As cotton became less profitable, non-traditional crops emerged as more profitable options. This profit difference and other institutional factors discussed below provoked a massive shift of large farmers out of cotton and into asparagus, oranges, apples, avocados and lucuma. This left room for small farmers to enter cotton production as a less profitable but less demanding crop in terms of organization and capital requirements.

There have been some recent instances of vertical integration of ginning and textile manufacturing firms, but the usual organization has been separate firms. The ginners act as intermediaries, buying raw cotton from farmers without agreeing contracts, and process the cotton into fibre and oil seeds to sell to textile firms and edible-oil factories. When large farmers produced cotton, they either sold to ginners acting as intermediaries for the textile firms or bought ginning services and sold directly to the textile firms.

When small farmers moved into cotton, they obtained credit from ginning firms or from large cotton growers and sold the raw cotton to them. When large farms shifted away from cotton, the small farmers became part of a system in which the ginners were almost the only source of credit, which was provided in-kind in seeds and chemical inputs.

From about 1995, difficulties increased for small farmers participating in the cotton subsector. Structural adjustments reduced access and increased costs for inputs and credit. The cost increases were magnified by the disappearance of the cooperatives, which had been dismantled in the 1980s. Ginners stepped in to fill the input credit gap left by government withdrawal, but at rates well above the former state-subsidized rates.

To offset rising input costs, small farmers turned to NGOs and rural and municipal savings/credit schemes. NGOs offer technical production assistance and credit at below market cost, subsidized mainly by foreign donors. These new sources of credit reduced small farmers’ dependence on the ginners’ expensive credit. The NGOs also negotiated with the ginning firms, with varied success, to increase the price for raw cotton.

The coverage of NGO schemes was partial and left out numerous smallholders. The schemes were limited to reducing credit costs but did not address numerous other problems faced by small cotton farmers, in particular lack of marketing and negotiation expertise to deal with other cotton-chain actors, expensive variable inputs and lack of organizational capital that was formerly embodied in cooperatives, which permitted economies of scale in input and credit acquisition.

These gaps in human and organizational capital and high transaction costs created a constraint and the opportunity for innovation to meet a need. The latter arose in the form of endogenous private institutional innovation, a share-tenancy arrangement remarkably similar to that described theoretically by Eswaran and Kotwal (1985). In 1999, a local large farmer established what can be called a management company. The company sells management services to small cotton farmers in return for a 25 percent share of the profits from cotton sales.

The management company requires the formation of farmer companies. The management company negotiates a contract with each farmer company that involves production and marketing actions such as timing of input use and bulk purchase of inputs, and group acquisition of bank credit with the manager’s intermediation and the farmers’ land as collateral. By the end of 1999, ten farmer companies were in operation, involving over 400 farmers.

This arrangement led to reduction of transaction costs and economies of scale in input purchase and product marketing, and to formal input market transactions. The latter allows a farmer company to abandon tax exoneration. When a farmer company uses the system of tax exoneration, it can no longer use the tax it pays on the purchase of inputs as fiscal credit. At the same time, the textile firm punishes the exonerated firm, because receipts can not be discounted according to the Peruvian value-added system when it buys its cotton.

The arrangement allows purchase of inputs in bulk, transported in trucks rented by the management company, which overcomes a physical capital constraint. This allows scale economies in input acquisition and purchase of inputs in Lima at significant savings compared to buying from local input dealers. Beyond these services, the manager is planning to help the client farmers with diversification of their product mix to reduce risk.

The arrangement has encouraged changes in the organization of the cotton subsector in Chincha. The farmers’ companies now contract ginning services and sell the ginned cotton directly to the textile firms. The companies get better prices than before their contract with the management company, which increases profit. Other cotton farmers in Chincha, numbering nearly 1 000, sell their raw cotton to ginners at a disadvantage compared to the farmers’ companies.

The present survey revealed that profitable options for small farmers have narrowed to working with an NGO or with the management company; farmers outside these arrangements are going out of business and selling their land, or are taking one of these two survival options. This appears to be mainly because even after paying the substantial management fee, profit when working with the management company is still greater than farm profits in NGO-organized schemes. The profit/cost rate is 80 percent higher in this tenancy contract as compared to working alone and about 50 percent higher than the NGO option as a result of the use of the tax credit. The NGO arrangement is being fuelled by donor funds, so it is less endogenous and probably less sustainable in the long term.

Asparagus

Many large farmers in the valley moving out of growing cotton have in recent years switched to asparagus, fuelling a production boom. This boom emerged for several reasons. Peru’s southern coast provides an exceptionally good climate for asparagus, offering producers two harvests per year. The resulting yields of 12 000 stalks/ha are far superior to the 7 000 stalks/ha registered in Spain, Peru’s main competitor and one of the foremost producers and consumers of asparagus. Labour costs in Chincha are low compared to those of Spain; irrigated farmland with registered titles is abundant. Chincha is close to the Lima market. In the past decade, there has been a virtual elimination of political violence in the region. Finally, the heritage of the long cotton boom is a large number of experienced farmers, agronomists and input and transport firms who are geared to commercial agriculture.

The relationship between asparagus agribusiness firms and the fundos is formalized in a contract. The company agrees to provide credit, technical assistance and inputs, and operates fixed pricing rules. The producers in turn promise to reserve all output for the company, following strict production guidelines, and to allow supervision by company technicians. The firms also buy asparagus from small farmers, but the latter do not enjoy the benefits of contracts and the price they receive is usually much lower as a result of quality differences. The great majority of small farmers cannot meet the requirements of quality asparagus production, because they lack managerial and technical expertise. The asparagus agro-industrial firms have a strong preference for contracting with large farmers because of monitoring costs and the capacity constraints of small farmers.

Interviews with the asparagus companies nevertheless revealed that they would like to increase production of asparagus on their own lands, obviously outside the contract system. This has recently been made possible by yet another institutional change: the law was recently overturned that had made it illegal for agro-industrial firms to own cropland and thus vertically integrate. This may subsequently undermine contract agriculture in one of the few places in Peru where it appears to have been functioning well (see Figueroa, 1996).

The advent of the asparagus agro-industry required and brought two institutional changes, the first leading to the second. The first change was in the institutions of grades and standards. The canned/bottled asparagus export market from Peru is highly demanding in terms of quality and safety standards with certification schemes by the Peruvian export association (Diaz, 1999).

The second institutional change was driven by the stringent standards and technological and capital demands involved in asparagus production. The advent of participation in asparagus export brought with it the institutional change of requirements for stringent quality and safety standards, which in turn induced further institutional change in terms of emergence of agro-industry/farm contracts to assure compliance with the standards. These contracts had not previously existed in the Chincha valley because cotton, the dominant crop of the large farmers in the past, had not been produced and sold by contract.

Direct effects

The direct employment impact of agro-industrialization includes employment in participating farms resulting from changes in product composition, technology and scale of production, and employment in agro-industrial firms. These effects are conditioned by the extent to which agro-industrial firms outsource their intermediate inputs or produce their own, for example whether textile firms use imported cotton or buy local cotton, and by the technology and scale differences implied by these alternatives. The effects of cotton and asparagus agro-industrialization in Chincha and institutional change flowing from it are hypothesized as follows.

First, the shift in the past decade by large farmers from cotton to asparagus has among other things tended to reduce the demand for farm wage employment and for direct farm labour per unit of agro-industrial output. Asparagus production is only 25 percent as labour-intensive as cotton production; asparagus agro-industry favours links with large farmers rather than small farmers who tend to have higher labour/land ratios; the labour/output ratio in the cotton agro-industry is about twice that of the asparagus agro-industry.

Second, there is a counterbalancing increase in employment from the increase in small-farmer cotton production, because cotton is more labour-intensive than subsistence maize. The increase in profitability from the institutional innovation discussed above would magnify and sustain this increase.

Third, the two effects on employment act in opposite directions: asparagus agro-industrialization and institutional change imply a drop in local employment, whereas there is an increase in the case of cotton. A rough calculation suggests nonetheless that net aggregate effects are probably negative regarding incomes of small farmers, with a drop of about 8-9 percent; the potential drop was substantially buffered, however, by the endogenous institutional innovation in cotton. The reasoning for this is shown in the following two steps.

Based on survey information showing that the change from cotton to asparagus involved roughly 1 700 ha and on corresponding labour/output ratios, the change produced a 6.6 percent drop in agricultural employment, and a second-round effect in the industry of a 25 percent drop in employment at the agro-industry level. Overall, these figures meant an 18 percent drop in employment income due to the change.

On the other hand, the income increase from the increase in cotton-production profitability for small farmers resulting from the institutional innovation is roughly 10-15 percent. That income increase is based on the following information about net gains from working with the management company. The company charges 25 percent of farmer direct costs, which works out to approximately US$239/ha. The benefits can be described as savings of costs relative to what small farmers paid before entering into this new arrangement:

• reducing input prices - 24 percent relative to prices normally paid at the shops in Chincha for reasons discussed above - saving US$116/ha;

• reducing loan interest by going to the local bank as a group rather than to the ginning firm for credit, saving about US$90/ha;

• obtaining a better price for the cotton by now selling directly to textile companies after contracting ginning services, generating gains of about US$150/ha;

• company status allows a change in tax status, allowing them to benefit fully from the value-added tax system in Peru, which allows savings of US$180/ha.

The net effect is that the small farmers spend US$239/ha for the management service, but save US$536/ha with new system, thus netting US$297/ha with the new system.

Indirect effects

The indirect employment effects include employment from net output changes in businesses in production-linkages forward and backward from farms and agro-industrial firms, and from consumption expenditure linkages from incomes earned in farms and agro-industrial firms. Table 6 shows the use of inputs on asparagus on large farms and on cotton on small farms. While small cotton farmers are much more commercialized and technologically equipped than subsistence maize or potato farmers in Peru, there are nevertheless substantial differences in the technologies and acquisition practices of small and large farmers in Chincha; their impacts on the local economy are therefore different.

First, nearly all farms use fertilizers, but the use rate per ha is much higher on asparagus than cotton. All farms use herbicides, insecticides and fungicides. The difference for the local economy is that asparagus farmers buy chemical inputs from input firms in Lima, whereas cotton farmers buy from a merchant in Chincha; this has started to change, however, because the farmer companies can buy in bulk from Lima. Why the difference? Large farmers have the asset base to serve as collateral and the management capacity to make contacts and to rent vehicles to go to Lima, where they get lower prices for larger lots. By comparison, farmers who have to buy from Chincha input firms are forced to buy at higher mark-ups because of lack of competition.

TABLE 6
Input use in fundos (asparagus) and parceleros (cotton): Chincha

Source: GRADE field work, 1996-1997.

Second, asparagus and cotton producers produce hybrids. Comparing seed and seedling use rates is not comparing like with like; it should be noted, however, that the purchase rate is the same at 75 percent, so these small farmers are relatively technically advanced compared to subsistence farmers in other parts of Peru. Both are bought locally.

Third, both kinds of farmers use mechanics in Chincha, but for tools, machines and spare parts large farmers tend to go to stores in Lima and the small farmers to stores in Chincha or to cottage-manufacture workshops. It was found that the institutional change in the cotton subsector reinforces this tendency, because the cotton farmer companies prefer to buy in Lima. This may be changing in the medium term, however, as the recent follow-up informal survey shows that local Chincha firms are trying to compete with Lima firms and offering similar bulk deals.

A check for the difference in technology between the two kinds of farmer shows that there are roughly similar spatial acquisition patterns for inputs. The smaller/poorer farmers buy inputs locally and thus benefit the local economy through upstream production linkages, at least to local commerce. The medium/large farmers buy inputs in Lima and thus obtain better-known brands, lower unit prices, greater information from the dealer, better product quality and quality guarantees, greater diversity of product, larger lots and perhaps more up-to-date equipment. But they gain all this while bypassing the local economy.

Fourth, asparagus firms have an internalization policy and so do part of their own transport and repairs, thus reducing linkages to the local economy. This is mainly because they perceived high monitoring cost in key parts of their production process.

Fifth, although the data were not avaliable for evaluating consumption expenditure effects specifically in the households surveyed as part of the agro-industrialization study, it was possible to use data relating to the study area from the 1997 Living Standard Measurement Study (LSMS) survey. The data show that the richer the rural household, the higher the share of non-farm expenditures in the total, as expected from Engel’s law. The 20 percent of poorest households have a share of non-food expenditures of 51 percent; the richest 20 percent have a share of 58 percent. The share of processed-food items in total food expenditure is, for the same area, 17 percent in the 20 percent of poorest households and 27 percent in the richest 20 percent. Most non-farm products are purchased in intermediate cities and consist of modern manufactured goods. Hence, as richer households are earning the asparagus profits, expenditure effects tend to benefit the intermediate city and Lima rather than the rural areas. There is historical evidence of such an effect: the growth of the city of Chincha was linked to the previous agro-industrial boom in cotton as large farmers benefited; development is now further increased by the boom in asparagus.

Mantaro valley

Potatoes

Three types of linkages between potato producers and the market were identified.

1. Retail and wholesale markets for potatoes for consumption. First, potatoes for consumption are the principal crop for most producers in the Mantaro valley. As can be seen in Table 4, producers located on the valley floor in Huayacuchi and Huanchac sell to wholesale markets in Lima. Producers located in the higher parts of the valley instead sell at local markets to intermediaries who channel the production to wholesale markets.

2. Coastal producers of potato seeds for sale. An estimated 14 500 mt of potatoes for seed are produced in the Sierra for farmers located in the central and south coast, supplying seed for over 8 000 ha. This part of the Sierra is known as a traditional producer of quality seed, so coastal producers and intermediaries establish long-term relationships with the seed producers, an arrangement which often includes provision of credit.

3. Potato processing companies. This link has developed relatively recently in the valley as a result of growth in demand for processed potatoes in Lima and the rest of the country. The most important company is Savoy Brands Peru SA, a transnational in which 85 percent of all sales of processed potato products are concentrated. Until 1994, supply for this company was obtained through intermediaries, among whom were the wholesalers of La Parada market. Since then a direct relationship has been established with producers.

From August to November, the company typically derives its supply of the Tomasa Condemayta variety from medium-sized producers in the coastal valleys near Lima. This variety has the best traits for processing potato chips - low sugar and water con-tent and appropriate uniform size. Since the traditional potato varieties in the Sierra do not have these qualities, for the other months of the year the company must stock up on coastal potatoes and store them in rented refrigerated warehouses. Even so, supply is not sufficient and the company often im-ports the Diacol Capiro variety from Colombia.

TABLE 7
Ties with coastal producers for the sale of potato seeds: Huayucachi

Source: GRADE field work, 1996-1997.

In this situation, the company established a relationship in 1995 with Mantaro Valley producers for supply during critical months. The Mantaro valley was chosen because of its proximity to Lima and its good transport links. The end of political violence in the area and most importantly the managerial experience developed by Valley potato producers were contributory factors. The company selected 50 producers who control over 1 000 ha in production, with average yields above 20 mt/ha. Most of the seed producers are professional technicians with years of experience in seed production. The principal characteristics of these producers are shown in Table 7.

The company provides the seed producers with seedlings, with the objective of producing tuberculillos prebasicos of the Diacol Capiro variety in greenhouses. The company guarantees the purchase of a specified quantity of tuberculillos at a fixed price and quality. The company then passes the tuberculillos to specific-consumption potato producers in the valley, to whom the company guarantees the purchase of output at a fixed price. For their part, seed and consumption producers give exclusive rights to the company for their output, and allow company technicians to inspect production processes.

Direct effects

The direct effect on employment of the advent of institutional change in potato contracts is more straightforward than in the Chincha case. The new variety produced in Mantaro has tended to reduce the demand for farm wage employment. First, the new variety is less labour intensive than traditional Sierra varieties; second, potato agro-industry favours large farmers, who tend to have lower labour/land ratios. There is no counterbalancing increase in employment in the small-farm sector, because innovation is limited exclusively to the technically most advanced and experienced potato farmers. Little positive direct effect in employment is foreseen, and it may turn out to be negative.

Indirect effects

As in the Chincha case, the indirect employment effects include employment from net output changes in businesses in terms of production-linkages forward and backward from farms and agro-industrial firms, and from consumption expenditure linkages from incomes earned in farms and agro-industrial firms. Table 8 shows the use of inputs on farms in the three districts. Many differences emerge, which have important implications for the existence and strength of local linkages.

First, 90 percent of farmers in Huayucachi and Huanchac use fertilizer, a higher share than the 60 percent in Aramachay, and much more intensively. Almost all producers use herbicides, insecticides or fungicides. As with Chincha, the difference to the local economy is that farmers from Huayucachi and Huanchac tend to buy their inputs from commercial establishments in Huancayo or Lima; Aramachay farmers buy theirs from local merchants. As with Chincha, the larger Huayucachi and Huanchac farmers have the asset base to serve as collateral and the management capacity to make contacts and rent vehicles to go to Lima, where they pay lower prices by buying in larger lots. By comparison, Aramachay farmers who have to buy from local input firms are forced to buy at higher mark-ups because of lack of competition.

Second, all farms in Huayucachi and Huanchac use hybrids, mostly purchased from valley seed producers. The institutional innovation thus takes advantage of a long relationship between local potato farmers and seed producers. Most Aramachay farmers use hybrids as well, but rely much more on their own production of seeds.

TABLE 8
Local linkages in potato production: Huancayo

Source: GRADE field work, 1996-1997.

Third, all farms tend to use local mechanics, but Huayucachi and Huanchac farmers rely more on Lima businesses for tools, machines and spare parts; Aramachay farmers depend almost exclusively on local purchases. Farms in Huayucachi and Huanchac prefer well known brands for tools; farms in Aramachay tend to use locally produced tools from cottage-manufacture workshops.

Thus two segmented markets are evident, with correspondingly different impact on the local economy. Aramachay farmers, who are generally smaller and poorer, buy inputs locally and thus benefit the local economy through upstream production linkages. The Huayucachi and Huanchac farmers, who are larger and wealthier, buy inputs in Lima and thus obtain more known brands, pay lower unit prices, receive more information from dealers and have better product quality and quality guarantees, greater diversity of product, larger lots and perhaps more up-to-date equipment. But they gain all this while bypassing the local economy. Institutional innovation has had little indirect impact on the local economy.

Finally, a consumption impact similar to Chincha should be expected. Only wealthier farms are benefiting from the institutional innovation in Mantaro valley; these households will spend a higher share of income on non-food expenditures, which are purchased in intermediate cities and consist of modern manufactured goods. Because richer households are earning the profits from processed potatoes, expenditure effects tend to benefit the intermediate city and Lima rather than local rural areas.

CONCLUSION AND LESSONS

Chincha

In contrast to other valleys, the significant presence of modern farming equipment in Chincha guaranteed to industrial firms the ability to implement productive technologies and ensured the availability of high-quality raw materials. Factors contributing to the success of linkages between agro-industry and local asparagus growers include sufficient agricultural knowledge and management capacity of the majority of landowners, availability of modern technical equipment and availability of arable land - on average over 50 ha per holding - most of which is equipped with irrigation systems.

These agreements are established through contracts in which the local producers obtain exceptional market conditions including rent contracts, credit and technical assistance, resulting in reports of higher earnings with respect to those in cotton production. Nonetheless, in this type of arrangement between agro-industry and local growers, the small landowner does not receive the direct benefits that are generated. In the end, they benefit solely as providers of labour during harvest seasons.

As in other areas of the country, such as San Lorenzo with its mango producers, the asparagus industry in the Chincha valley continues to buy up local land and to opt for vertical integration. This process has not yet ended, but it is possible to affirm that agro-industrialists have opted to ensure access to quality raw materials for two reasons. First, many asparagus plantations have already entered a phase of diminishing yields, and their infrastructure needs renovation; this also presents an opportunity to renew contracts with landowners under a new arrangement. And until recently, outside businesses could not buy out local landowners; under new legislation it is now possible. Second, there is a need for additional areas for cultivation beyond the land in the hands of middle-sized landowners. Given this shortage, sufficient contracted lands are no longer guaranteed to agribusinesses, which are therefore now interested in buying up land. It is clear that new legislation has offered large firms greater bargaining power in their negotiations with middle-sized landowners, but the situation has worsened from the viewpoint of local landowners. In order to assure a continuation of the beneficial contracts they receive from agribusinesses, middle-sized asparagus producers must now be more efficient in their operations, because firms in the asparagus industry now have a reasonable indicator of the cost of producing 1 kg of asparagus in their own fields to compare with the prices of local producers.

In Chincha, the abandonment of cotton production on middle-sized and large land holdings in favour of asparagus crops has consequently meant that small landowners are the principal cotton producers in the valley. A recent phenomenon in this valley has been the recent emergence of contractual arrangements between a small private consulting firm and an important group of producers, with the objective of augmenting the participatory role of local growers in the commercial aspects of cotton production. This arrangement has brought significant economic benefits to small cotton producers.

With this method of operations, the management company has identified a role that is profitable for them and that generates a circle of benefits, allowing local economic agents to interact with greater ease and thereby increase overall utility. Under this arrangement, private banks are more willing to offer loans to producers on the basis of the reputation, experience and knowledge of the owners of the consulting agency supporting them - provided, of course, that each small farmer complies with the bank’s minimum requirements in terms of property titles, collateral and down payments. By connecting themselves with the matrix of commercial activity of input producers located in Lima, they can acquire significant price discounts for large-volume purchases. By gaining direct access to the textile companies, they receive higher prices with respect to those previously received from intermediaries. In other words, this arrangement generates economies of scale in production that individual farmers are unable to achieve alone, thereby providing economic gains for local growers as well as the private management consultants.

Mantaro valley

Processing potato-based snack foods is one of the most dynamic and important areas of activity in the market for processed potatoes, in which important changes are occurring as a result of private producers’ efforts to improve their market position and increase sales. The most important firms involved in this line of products are multinational affiliates. Until three years ago, the system used by these firms to obtain supplies of potatoes used intermediaries; it is now based in a direct relationship between multinationals and groups of producers that varies according to seasons of production.

In the Mantaro valley, contracts require farmers to have sufficient quality land, irrigation, human capital and managerial skills to meet demanding production and marketing schedules, to sell only to the contracting firm and to submit to that firm’s technical supervision. Only large farmers in the Mantaro valley can meet such requirements. The rewards, through the contracts, are technical assistance, credit for land rental and input acquisition, quality seedlings supplied by the company at agreed prices, and profitability and risk reduction at a fixed price adjusted between contracts depending on international prices.

The valley’s proximity to Lima, the present environment of security and the transport facility provided by some of the country’s best highways make this valley a desirable location for acquiring agricultural inputs. The presence of leading producers with significant managerial capacity is of great importance. The majority of corresponding seed growers are professional technicians with years of experience in seed production who use areas of cultivation located in agro-ecological zones that are appropriate for seed development. Similarly, the selected potato growers are producers specialized in this crop, whose production has been traditionally destined for the wholesale market of La Parada.

Lessons

It is clear that access to public goods and services coupled with the appropriate amount of private assets, especially education and managerial ability, can dramatically improve the quality of farm/off-farm linkages. In this respect, improvement of public roads in a country like Peru is a significant element in reducing marketing costs. Improvements in land titling and registration is important as a means to increase the chances of obtaining credit.

In each of the two cases analysed, however, these elements have been necessary but clearly not sufficient to improve linkages between small and medium-sized farms and agro-industry. Managerial abilities are a crucial element to developing successful linkages; this paper has focused on the phenomenon of endogenous institutional innovation in two areas of Peru involving traditional crops - cotton and potatoes - and a non-traditional crop - asparagus. These changes have been induced by changes in the general institutional context such as the emergence of demanding quality and safety standards in agro-export markets and domestic agroprocessing markets, the policy and market context in terms of the withdrawal of government support to input and credit markets, and the factor-distribution context - the scarcity of management and technical and marketing expertise among small farmers.

The institutional innovations that emerged in the zone were twofold. The first included contracts between agro-industrial firms and large farmers; the contracts were introduced by the firms to assure timely delivery and compliance with strict requirements required by the new and demanding quality and safety standards for agro-export of processed asparagus and processed potatoes. The second included management services exchanged for labour supervision and land collateral in share-tenancy contracts between a management company and farmer companies of small cotton farmers.

The importance of these institutional changes is twofold. First, they were induced institutional innovations driven by the requirements of agro-industrialization itself; second, they had ambiguous employment and income impacts that tended to be negative. In Chincha, on the one hand, the emergence of asparagus and firm/farm contracts reduced employment through exclusion of small farms and shifts to capital-intensive crops. On the other hand, the reinforcement of smallholder cotton and the emergence of farmer companies increased smallholder employment and incomes. The institutional innovation allowed them to reduce risk and increase profits, and thus access some of the benefits of agro-industrialization and globalization. Processing firm/farm contracts are common in Peru, as is the presence of NGOs bringing subsidized credit, but the private management firm innovation is rare and new in Peru, and apparently new in the region. Policymakers and NGOs have recently discovered that this innovation is taking place; they are asking hard questions about whether the innovation can and will be diffused. The interest in the private for-profit institutional change is sharpened by growing doubts about how economically sustainable and widespread NGO help can be for small farmers in maintaining their participation in income-enhancing agro-industrialization. With changes in land laws and markets, the fluidity of the situation is apparent; agro-industrial firms are even starting to ask themselves whether contracts with large farms are necessary or the best option.

Two policy issues emerge. First, can policymakers do anything to facilitate the emergence of such private institutions, which are profitable for suppliers and customers but do not jeopardize their inherent strength - their endogeneity? Second, can policymakers undertake complementary actions to facilitate agro-industrialization in demanding profitable sectors? There may be a role for the government. It may be a direct role, or it may involve temporarily subsidizing or providing facilities for private management, accounting and technical training for farmers in zones already undergoing agro-industrialization or that have potential for it. The objective would be to improve the conditions for the emergence and low-transaction-cost functioning of such private institutions among management-service providers and the farmer companies. These policies should be aimed at reducing transaction costs for the emergence and development of managerial services that will be demanded only if they achieve a superior input mix through resource pooling in the face of a moral hazard problem.

On the other hand, the larger farmers, who have better land and access to education and water, were ready to profit from the shift from cotton to the more profitable asparagus in Chincha, and to processed potatoes in Mantaro. Among the small farmers in Chincha, those with titled land were better positioned to join the lucrative farmer companies. It is thus important that there be renewed attention to building the private and public asset base of small farmers in these areas. That is crucial to their participating in agro-industrialization rather than being excluded from it. The World Bank (1998) discusses examples of these agro-industry facilitation actions in other zones of Peru, including provision or facilitation of the private development of physical capital such as wells and nurseries, managerial capital such as technical expertise and management experience, infrastructure such as roads to growing markets and input sources and essential public services such as resolution of problems related to plant and animal health or the availability of registered land titles.

Finally, at least basic preliminary evidence has been presented that the direct employment effects of agro-industrialization in the Chincha valley are ambiguous and tend toward income concentration. The same process is probable in the Mantaro valley. It seems clear, however, that the indirect spin-off effects of the process point in the direction of concentration of gains among farmers who are already better off: the larger farm households that dominate the off-farm manufacturing and service sectors in rural areas and local intermediate centres. Upstream production linkages are correspondingly weak.

There is a need to recognize that these impacts have been concentrated and to seek a broadening of participation by poorer groups. This would result in more equitable development, and smaller producers would tend to have stronger local production and consumption linkages. Training, technical assistance and credit provision for small and medium-sized enterprise development would be a step toward this, especially in subsectors such as transport and equipment repair and manufacture that are identifiable as profitable spin-offs from the agro-industrial economy.

BIBLIOGRAPHY

Diaz, A. 1999. La calidad en el comercio internacional de alimentos. Lima, Peru, PROMPEX (Comision para la promocion de exportaciones).

Eswaran, M. & Kotwal, A. 1985. A theory of contractual structure in agriculture. The American Economic Review, 75(3), 352-367.

Figueroa, A. 1996. Pequeña agricultura y agro-industria en el Perú. Economía, XIX: 37-38.

Ministry of Agriculture. 1999. Producción agrícola 1998. Lima, Oficina de Información Agraria.

World Bank. 1998. Peru: an agricultural development strategy. Washington, DC. (Gray Cover document.)