Introduction

The public agricultural extension services have been questioned since several years ago, due to their poor results and high costs. The greatest criticism to the public extension systems are linked to their own efficiency, motivated by state bureaucracy and the lack of motivation of public officials; in addition, the use of participative approaches is rarely well-adjusted to the organizational competencies of the public sector or to the individual strategies of the public administration agents (^{Kidd et al., 2000}).

A 2001 review financed by the Food and Agriculture Organization of the United Nations (FAO), for example, characterizes extension services in the entire developing world as faulty, dying and in disarray (Rivera et al., 2001, citado por ^{Feder et al., 2010}).

Thus, new proposals have been configured for the improvement of extension services, based on the participation of the private sector, stemming from the hypothesis that the possibility of establishing public-private agreements to provide them, as well as the use of public funds, although managed by a private system, would increase the efficiency of the system and the coverage to reach all groups of farmers (^{Dethier and Effenberger, 2012}).

In some places of Europe, even the name of extension service has been changed for that of pluralist consultancies. Behind this semantic change, which can be found in many countries such as France, England, Holland and Germany, the growing role of private actors is shown (^{Labarthe and Laurent, 2013}). To be efficient, the providers of extension services must collaborate with businesses, banks, non-governmental organizations, and other service providers (^{McMahon and Valdés, 2011}).

However, it must be noted that this new approach constitutes a radical change and that, therefore, its application generates certain difficulties. The step from a prolonged tradition of public services -according to which these services have been received without any cost to the producers, and the technicians have performed their functions without developing an entrepreneurial attitude- to a new situation in which producers and extension workers must manage in a service market, require strong incentives for their attainment (^{Mora, 2002}).

This agrees with the need for farmers to internalize their work as a business, in which consultancy and innovation have a cost. The commercial farmers will require training in entrepreneurial management and consultancies that will help them climb in the value chain, while small-scale farmers will require technique, both as market and financial information, and as assistance to go from subsistence production to a commercial type of enterprise (^{Pye-Smith, 2012}).

A first effect of the privatization of services is observed on the flow of information since the flow tends to decrease when changing from a public to a private good. If a farmer has to pay the extension service for the knowledge he has attained, he will be less inclined to share the knowledge with other farmers than when it was freely available (^{Van den Ban and Samanta, 2006}).

On the other hand, privatization affects the content of the services that are made available, for it will be increasingly more identified with the needs of medium-scale and large-scale farms which will be the ones to pay for the services more easily. If the current trend continues, it is likely that the extension service is increasingly less relevant for small-scale farms (^{Labarthe and Laurent, 2013}).

Another aspect to highlight in the new approaches to extension is the suggestion of a more decentralized extension work and closer to the land. The centralized extension services have always shown a difficulty with regards to their pertinence, due to their poor capacity of contextualizing the local problematic, particularly of small-scale farmers; on the other hand, the supervision of services under this system has not been possible, so it is considered that the more local the service is set out, the more effective and controllable it will be.

However, the effectiveness of the decentralization also depends on its approaches and objectives. There are reports about decentralization of extension services carried out in China, indicating that they have generated for extension technicians to spend much more time in administrative negotiations than in effective extension, because the mandates of local governments are, above all, economic growth and social stability (^{Hu et al., 2012}).

If we analyze the path of agriculture in Latin America we can see that every country has undergone similar experiences, in a general sense. Beyond their singularities, in Latin America and the Caribbean there are common features that come from the same agrarian tradition: structural heterogeneity, incomplete technological modernization, low population density, great availability of natural resources, institutional weakness, community tradition, low endowment of social capital, and high inequality in land distribution, among others (^{Sotomayor et al., 2011}).

In the early 1980s, Latin American countries were experiencing an important crisis provoked by the policy of import substitution and price regulation. Therefore, with some years of difference between each country, they had to implement structural adjustment measures. These are reflected in the reduction of the size of the State and of their mechanisms of intervention in the economy; in the weakness of the institutions that provide public services, as well as in the deregulation of markets (^{Birbaumer, 2011}).

All these changes directly affected the public agricultural extension, which is presented as an expense without great results, and the reduction of services by the State is proposed. During years a reduction of the public staff devoted to extension is observed, leaving a void that was covered by the non-governmental organizations financed by international cooperation.

With the new century and the positive perspectives of agriculture in Latin America the interest for extension services resurges. For example, the Global Development Report of 2008 emphasizes agricultural extension as an important intervention for the development of countries in two aspects: 1) to potentiate the growth of the agricultural sector, in light of the increase in demand and the pressures of the offer; and 2) to promote sustainable development (Raabe, 2008; cited by ^{Cuevas et al., 2012}).

^{Ho and Montero (2014)} indicate that the new initiatives of the extension services in Latin America combine the following schemes: i) the development of a market of services within the private sphere, in sectors of large-scale and medium-scale commercial agriculture; ii) a modality with a certain level of subsidy according to which the State transfers resources to private companies and non-governmental organizations for the provision of services to small-scale commercial agriculture, within special programs; and iii) a modality for subsistence agriculture within the framework of public programs for rural development or programs for the fight against poverty.

These modalities suggested are the ones that make ^{Alemany (2008}) say that privatization of rural extension has its conceptual bases in the view that “viable” agriculture is already industrialized and integrated in a subordinate manner to the agrifood chains, while the “non-viable” or the “residue resistant to modernization” is a problem that should be tackled as a social problematic, not a productive one and of growth.

The current designs and policies of extension also emphasize decentralization in Peru, Chile, Brazil, and México, transferring funding and decisions to the regional or local level, allowing for the process of decision making to be closer to the users. However, it is not clear that this process will be sufficiently accompanied by the strengthening of the capacities at these levels which allow ensuring the quality and sustainability of extension work (^{Aguirre, 2012}).

As we can see, there is still much to evaluate regarding these changes in extension services in Latin America to verify the problematic that its application entails. In this regard, ^{Swanson (2010}) indicates that the experience in Latin America and in other regions proves that replacing these public extension systems with private sector companies or with non-governmental organizations will probably cause another series of problems and limitations that could reduce even further the success of these alternative methods.

^{Hellin (2012}) indicates in addition that in most cases, the private sector has been proven incapable of replacing the services that the public sector formerly performed, due to the high transaction costs, the dispersion of the clientele, and the low (or non-existent) benefits, situation that becomes increasingly more evident in face of the small-scale producers.

The agrarian extension services in Peru have followed processes parallel to those that took place in all of Latin America (^{Ho and Montero, 2014}), since they were affected by similar policies. The agricultural policies in Peru were directed at reducing the subsidies, the promotion of free trade and the participation of the private sector in research and information provision (^{Ortiz, 2006}); this motivated for non-governmental organizations to assume an important role in the offer of extension services, activity that was lessened as International Cooperation funds decreased.

All these changes evidently did not obey only an external context that drove this change, but also the recognition of a very serious problem in the services offered by the State. According to ^{Carrasco (1990)}, in 1990 the situation of agricultural extension was the following: 203 rural development centers in the country, 1301 people devoted to extension, with very little equipment for their mobility, who had not developed activities in the field since years ago, and only in some localities had the support of producers for occasional actions. In general, the services were paralyzed and the staff impoverished, demoralized and outdated.

After the application of the adjustment measures, the situation changed radically. For example, the government program of agricultural extension led by what is now the National Agrarian Innovation Institute (Instituto Nacional de Innovación Agraria, INIA) employed 1400 agents in 1986, but the staff was reduced to less than 100 agents in 1992 (^{Hellin and Dixon, 2008}).

On the other hand, large agricultural investments are promoted to favor the development of agroexports. Thus, the interest in small-scale farmers by the State was reduced in favor of a greater interest in the expansion of the investment of large private capitals in agriculture (^{Remy and De los Ríos, 2012}). According to ^{Eguren (2006)}, peasants who maintain subsistence agriculture have been marginalized from all ambition of development and they have been made into the object of the so-called “social programs”.

This situation caused for non-governmental organizations to attempt to cover the void left by the State through projects financed by international cooperation. According to ^{Ramírez and Roe (2007)}, non-governmental organizations promoted countless consortiums, alliances and bilateral agreements to be supported in the implementation of activities of extension and technological diffusion. However, although many non-governmental organizations made participative suggestions, for ^{Wiener (2010)}, their approach continued to be offering.

Since 2001 the decentralization policies in Peru allowed for regional and local governments to have the possibility of promoting agricultural development; on the other hand, the national government generated a series of programs that have sought to create a market of extension services, with the most important one being the Program for Technological Innovation and Competitiveness Promotion in Peru’s Agriculture (Programa de Fomento de la Innovación Tecnológica y la Competitividad en la Agricultura del Perú, INCAGRO).

For the ^{World Bank (2012)}, INCAGRO seems to have been a good investment and a successful project, although equality continues to be a preoccupation: most beneficiaries of this program were medium-scale and large-scale farmers, instead of the most disadvantaged producers, including women., Some changes were implemented in its second phase to counteract this situation, but it is indicated that the results have not been clear.

Later new national programs have emerged in Peru which have had the same approach, so it becomes quite important to understand whether these actually reach the small-scale farmers (which constitute the most important reality of agriculture in Peru and in many countries of Latin America), or, in any case, to understand which are the characteristics of those producers who are reached by these services, in order to promote the necessary changes to broaden the access.

In this sense, the case of Peru, which has undergone both processes (privatization and decentralization) in the last 25 years, is an interesting reality to research their effects on the provision of extension services, primarily at the level of access and the typology of producers who are reached by such services.

The objective of this study is to determine, based on the analysis of data from the National Agricultural and Livestock Census carried out in Peru (INI, 2013), the level of coverage of extension services in Peru and their main providers, as well as to identify the principal factors that influence the access to extension services, from the analysis of the profile of the producers who receive these services.

Materials and Methods

Information referring to the National Agricultural and Livestock Census carried out in Peru during 2012 was taken. The information was cataloged to determine the coverage of the extension services (specified as technical assistance, training or entrepreneurial consultancy in the census results), and the entities that performed these services; this coverage is analyzed in function of the agricultural surface of the producer and the type of entity that provides the service.

On the other hand, of all the information about the producers contained in the census, the most important predicting variables were selected in function of the relationship there is between extension work and some socioeconomic characteristics of the producers. The ones selected were: size of agricultural/livestock unit, agricultural surface, age and sex of the producer, legal condition, associationism, level of literacy, level of education, use of agricultural machinery, access to credit, distance from the district’s capital, access to social programs, perception of sufficiency of their level of income, among others.

In order to determine the characteristics of the groups of producers who receive extension services, the algorithm of the CHAID (Chi-squared Automatic Interaction Detection) classification tree was applied, which is a method of multivariate analysis which through a technique of segmentation analysis of variables (explained and explicative) allows establishing the dependence relationships between them (^{Escobar, 2007}); response trees based on the relative frequencies in their relationships were formed with these variables. As restriction to the system, it was established that the groups have at least 50% of producers who manage to gain access to services of agricultural extension. For the processing of data, the statistical software IBM Statistics, version 20, was used.

After obtaining the results for the extension services in general, the same analysis was performed for the services of agricultural extension offered by each purveyor institution, whether public or private, with the aim of determining if any of them had a characteristic target population or was focused on some typology of the agriculture and livestock producer.

Results and Discussion

The coverage of extension services in Peru is shown in Table 1, in function of the size of the agricultural/livestock producer. In it, it can be seen that the coverage for the whole country is 10.2% of the total producers, with a decreasing trend in function of the smaller size of the agricultural and livestock unit, surface; this would be indicating initially that changes in the approach of the services have not managed to have efficacy with regards to the provision of extension services for small-scale farmers, particularly for those with agricultural surface smaller than a hectare.

Although this value is low, it is important to indicate that in a similar study carried out in México, the level of coverage of technical assistance and training was only 3% of a total of more than four million productive units analyzed (^{Cuevas et al., 2012}), with no similar data available for other countries in Latin America, in order to be able to generalize the situation of poor access to extension services.

In Table 2 the extension services are shown in function of the offering entity, in function of training, technical assistance and consultancies performed. Results show that the services have been decentralized, with the responsibility for these falling on the local and regional governments, a quite different situation from the one observed in the 1993 census, where the extension services of the regional and local governments were non-existent.

This higher participation of local governments could be the opportunity to foster a more flexible and contextualized agricultural extension system. Decentralized extension services improve accountability to the local users and ease for clients the “purchase” of research services and of products that better respond to their needs (^{World Bank, 2012}). However, for the work at this level to be effective, the operation of local technicians should be strengthened.

It is important to point out that under the sphere of the private sector, the extension services offered by companies, individuals, non-governmental organizations, and producers’ associations are included, which have been covering the services that have been set aside by the public sector and which constitute almost 40% of the services offer.

The possibility for there to be a private offer of extension services is important, for it is constituted with a more flexible actor, which does not mean that the regional or local governments should cease to offer them. Rather than opt for one sector or another, a unified approach should be applied, which can take advantage of the synergies that can be obtained objectively by establishing work relations between each other (^{Sotomayor et al., 2011}).

Concerning the services of the National Government, one of the greatest difficulties observed is that an important percentage is carried out through special projects, that is, they do not constitute a stable service. On the other hand, it has been seen that the offer of extension services by the universities is inexistent.

The results of the profile analysis of the producers who do receive extension services are presented in Table 3. It can be indicated that the variables that define the profile of the producers who receive extension services in Peru are: agricultural surface, associationism, and credit management. The other variables studied (age and sex of the producer, legal condition, level of illiteracy, level of education, use of agricultural machinery, distance to the capital from the district, access to social programs, perception of the sufficiency of their level of income, among others) have not shown to have a significant influence on the level of coverage of the extension, under the restrictions of the analysis performed.

If we analyze the index of response accumulated from the trial, it will be seen that the group of producers who gain access more easily to the extension services are those that have an agricultural surface of 5 to 500 hectares and of 10 000 hectares of more, who are associated and who can negotiate credits.

All of these characteristics are quite connected to those of a commercial farmer, showing with clarity that the approach developed is not effective to respond to the demand for extension services of small-scale agriculture, which, as we have pointed out before, constitutes the majority group in Peru’s agriculture. It is important to also point out the importance of associativity in the access to extension services, so it is fundamental to promote it.

The results obtained also allow us to corroborate the need to make the extension models widespread focused on small-scale farmers; it is important to implement practices that maximize the inclusion of those who are normally at a disadvantage, such as the poor, women, and small-scale farmers, and who maximize the positive impact in their livelihoods (^{Rajalahti et al., 2008}).

These models should include different actors, with the aim of responding to the complexity of their systems in a more efficient way than the models based on the approaches linked to commercial agriculture. For example, the approaches based on public funding that involve the local governments, the private sector, non-governmental organizations, and producers’ organizations in the provision of extension services can be more relevant for subsistence farmers, while various forms of private co-funding can be appropriate for commercial agriculture (^{World Bank, 2012}).

Other authors state the same need, indicating that the formation of public-private alliances is necessary to guarantee for agricultural extension to reach all the farmers, not only large-scale and commercial producers (^{Al-Sharafat et al., 2012}).

When evaluating the producers’ profile in function of the institution from which they have received extension services, it is observed that there is not a specific profile, which indicates that the institutions do not have a differentiated action on the territory, situation that reduces the efficiency of the system. This brings up the need of forming networks between institutions, so that their articulation can allow a greater coverage and efficiency in the use of the resources.

As can be seen, these results show the need to establish diversified approaches in order to reach the different typologies of producers; as ^{Aguirre (2012)} indicates the differences between countries and situations imply that unique extension models developed under different conditions cannot be implemented, even within the same country.

Conclusions

The results of the study show that the changes brought about in the extension approaches in Peru have not managed to improve access to these at the level of small-scale producers, however this verifies that the participation of local governments and the private sector in these activities is rather important. On the other hand, the relationship there is between the coverage of extension services and factors like the productive size, associationism, and access to credits, clearly shows that the new approaches manage to reach primarily the commercial producers who are associated, which proves the importance of the promotion of agrarian associativity and of seeking new approaches to reach small-scale producers.

The implementation of a decentralized and articulated innovation system becomes necessary, with different approaches in relation to the different realities and profiles of producers, a system that strengthens the efforts of local and regional governments, and the private sector.