In the starting blocks for smart agriculture: The internet as a source of knowledge in transitional agriculture

Abstract

The work described here has sought to define the role of the Internet in knowledge acquisition among Polish farmers, as well as the diversity characterising their professional activity conducted online. Relevant discussion is in this way broadened to reflect the conditioning underpinning smart agriculture, most especially in the context of states emerging from a period of economic transition. Particular attention is here paid to the factor of choice of source of information assisting with the running of a farm. Analyses relating to this matter are founded upon questionnaires supplied by almost 2500 farmers. The results show that the Internet does not constitute the most important information source for Polish farmers, though there is a close link between use of the Internet and their basic social characteristics, as also associated with structural features of Polish agriculture. On that basis, it can be considered that Polish farming still finds itself at the preliminary phase of entry into smart agriculture. The Polish case shows that we cannot assume that there is a readiness for smart farming in all places.

Keywords:

• Internet
• Knowledge
• Economic transition
• Agriculture
• Poland

1 Introduction

Smart agriculture is constituted by farms and farmers using advances in information and communication technologies to react to challenges In this case, social, economic and technological change in rural areas is very much conditioned by the development of smart farming. Thus, as new challenges are faced, rural communities and the rural economy (including farming) ought to take the smart development model into account more and more (Naldi et al., 2015). When we speak of any economic sector being” smart”, account is taken of the fact that a territorial (local or regional) system comprises its entire array of interconnected components. This is particularly true of agriculture, given the way this is entirely rooted in the community operating in the specific area. That in turn denotes influence exerted by community resources of knowledge, means of functioning and approaches to economic activity (Woods, 2005; Price and Evans, 2009). As territorial components comprise the economy, the environment, management and residents (Kumar and Dahiya, 2017), a basis for these to come together as a cohesive whole can be provided by ICT ¹ .

The bases upon which ICT is used, as well what determines effectiveness of that use, are very much constituted by issues relating to the human condition – for instance level of education (van Deursen, van Dijk, 2011), but also local resources such as networks of acquaintanceship and support (Selwyn et al., 2005; Courtois and Verdegem, 2016) and actual characteristics of farms (Czapiewski et al., 2012). This in essence leaves “smart people” as elements crucial to each and every smart area (Kumar and Dahiya, 2017). Beyond that, the basis for the utilisation of all kinds of smart solutions arrived at by the people will comprise aspects associated with knowledge transfer, where this is taken to mean adoption as much as transmission (Davenport and Prusak, 1998). In essence then, the matters of key importance to smart areas and smart agriculture are fast and direct access to information, and hence knowledge, and then the ability for these to be absorbed, as deemed to be a function of individualised “absorption capacity” at the level of the given receiver (Zahra and George, 2002), and as coupled with effectiveness of application in practice.

It should be stressed that smart farming, treated as farms and farmers interlinked via ICT, extends beyond the notion of precision agriculture (e.g. Sundmaeker et al., 2016), given that the latter is taken to mean the collection and processing of data from monitoring, as linked up with different treatments of the land, means of cultivation and ways of raising livestock, in order that rational decisionmaking can be engaged in to improve production from both the qualitative and quantitative points of view (McBratney et al., 2005). With smart farming, in contrast, the ICT helps to bring together the knowledge resources of farmers, with multiplier effects ensuing, and new modes of operation emerging as a result.

The main objective of the research ² reported here was to define the role the Internet plays in knowledge acquisition among Polish farmers and provide insights into the diversity of professional activity conducted by Polish farmers online. To this end, the role of the Internet in the running of a farm was considered in relation to such features as age and education of the farmer, production profile and economic status as self-assessed.

Analysis is based on material that encompassed more than 2400 survey questionnaires. They consist of questions regarding those social and economic characteristics of farmers and their farms proving most important from the point of view of the development of digital competences. The latter are treated as a foundation upon which smart solutions in the process of farm management gain implementation. In effect, then, an attempt was made to determine possibilities for farming to transition to the stage of smart development, albeit with an indication also given of the most important factors either advantageous to this process of transition, or else acting to hamper it.

Since the late 1980s and early 1990s, agriculture in Poland has undergone two major revolutions needing to be regarded as major shifts. The first of these went hand in hand with the post-communist transition, the second with accession to the European Union. Both shifts imposed organisational, production-related and technological change upon a Polish agriculture that continued to be dominated by a predominantly family-farming-centred model (Gorlach et al., 1994; Fedyszak-Radziejowska, 2010).

This article is organised in such a way that a literature-based matters associated with smart farming is first given, prior to a shift of attention to the link between the Internet and knowledge relevant to agricultural activity. Research methodologies are discussed in another part, before results are presented. As typically, these results support engagement in a more general discussion, as well as the formulation of certain conclusions.

2 Smart agriculture

Solutions to be found on the Internet constitute the “nervous system” underpinning smart areas and smart agriculture, given the responsibility for the flow of information and the capacity for fast action. The role with respect to smart farming will be exactly the same as with the upcoming Industry 4.0 (Lasi et al., 2014) – also known as the Fourth Industrial Revolution ³ - and the smart factories that are its key components (Wang et al., 2016). That leaves smart farming definable (of course in general terms) as that kind of farming whose deployment of ICT-based solutions ensures better and more effective monitoring and management of a farm, and (yet more generally) of farming activity (e.g. Wolfert et al., 2017). A key instrument in this case is the so-called “Internet of Things” (IoT) (Zhao et al., 2010) or “Web of Things” (WoT), in which more attention is actually paid to the application layer than to the technical one (Guinard et al., 2010) ⁴ . The actual use of such instruments in agriculture remains fragmentary, with true integration lacking, and arising solutions adopted often still being in the experimental phase (Verdouw et al., 2016).

There are further elements, proper to contemporary ICT, which are also proving to have growing significance for agriculture, i.e. GPS (Schönfeld et al., 2018), cloud computing (TongKe, 2013; Channe et al., 2015), edge and fog computing (Ferrández-Pastor et al., 2018) and Big Data analyses (Lokers et al., 2016). New technologies and requirements with respect to knowledge thus serve to move farming’s organisational culture away from experience-based management and towards data-based management (Butler and Holloway, 2016). As Lokers et al. (2016) remarked, data (and more precisely Big Data) constitute a raw material. Deriving from various sensors, including those involving people, these are being transformed into knowledge, with action then made possible, and decisionmaking offered support (Barnaghi et al., 2013).

Reports on smart farming make it clear that social aspects associated with users are most often marginalised. Poppe et al. (2015) for example emphasise that certain farms and areas will not be able to compete in the domain of smart farming if the basic infrastructure is not made available. This is obvious insofar as the development of IoT and other technologies assigns a bare-minimum role to human beings where information exchange and the management of selected processes are concerned. And this is not least a reflection of the way devices communicate autonomously with one another and are themselves involved in information exchange.

That said, a focus on the technological aspects alone would seem inadequate. Walter et al. (2017) indicate how new technologies still look very costly from the perspective of a small farm. Put that together with real-life limitations on both knowledge and skills, and essential limitations on the development of smart farming may be being imposed, in developing countries in particular. The self-same issue is pointed to by Fleming et al. (2018), who stress that, while all have theoretical chances of benefiting, real benefits from, for example, Big Data only really accrue to a few farmers running large farms.

Control over decisionmaking represents a further key problem. The so-called Second Machine Age (after Brynjolfsson and Macafee, 2014) is now being entered by agriculture, as in fact by the entire economy, and this will entail clever machines being designed to make better decisions than farmers. This is then a source of fresh smart-farming challenges, not least farmers’ loss of a key role in decisionmaking in the face of a greater influence of algorithms devised by scientists and/or those who produce farm equipment. This in some sense leaves processes taking place on a farmer’s own farm out of his/her control (Jeanneaux, 2017), and again demands a view that social aspects of the transition from conventional farming to smart farming are as important as technological ones.

An essential contribution to the state of the art is here made through the incorporation of social conditioning into the study of implementation processes where smart farming solutions are concerned. Analyses carried out are of key value where a very large number of farms are sampled, and both structural differentiation and diversity of location are then depicted effectively. A significant novel feature of the approach being espoused is thus the perceived need to consider the perspective on smart farming manifested by the immediate social actors in the discussion, i.e. the farmers themselves.

In addition, it is possible to note this kind of gap in even comprehensive studies concerned with countries (mainly CEECs) in which profound processes of agricultural transformation are taking place. There is thus no doubt that the aspect of the use of Internet in the pursuit of farming activity is just one of the elements coming together to represent smart farming solutions, and even it seems not to have had due weight assigned to it.

3 Theory: knowledge – Internet – agriculture

In farming, as in remaining sectors of economic activity, knowledge is a key factor allowing effectiveness to be maintained and increased (see Floriańczyk et al., 2012). The very first studies on the modernisation of agriculture demonstrated that farmers’ knowledge acquisition in respect of available production technologies was a necessity if economic development of farms was to increase (Wilcox, 1943). Levels of knowledge and education then started to be listed among variables explaining differences in productivity among farms and the differentiation of this phenomenon across space. While some models treated technology and knowledge as external factors Eicher and Staatz, 1998, one of the key economic concepts (the model of induced development in farming presented by Hayami and Ruttan (1985)) deemed the respective variables endogenous.

Davenport and Prusak (1998, p. 5) emphasised that knowledge is” a fluid mix of framed experience, values, contextual information, and expert insight that provides a framework for evaluating and incorporating new experiences and information. It originates and is applied in the minds of knowers”. Information and knowledge are thus treated as goods that satisfy elementary human need. Yet, as distinct from other goods, these ones are “volatile” resources (and so can be forgotten or deformed, for example). At the same time, information and knowledge are endowed with other specific and valuable attributes, in that: a) knowledge is not used up in the production process, and can be made subject to multiplication; b) a person transmitting knowledge does not lose it, but continues to have it at his/her disposal; c) knowledge can be acquired with relative ease (given sufficient motivation and conditions), and can then be at the unique disposal of the person who has acquired it (Drucker, 1994; Nonaka and Takeuchi, 1995).

Knowledge is characterised by lack of homogeneity (Bathelt et al., 2004). Polanyi (1966) introduced a classification of knowledge into the codified or formal (also called explicit) and the informal or hidden (i.e. tacit). Codified knowledge may readily be represented by symbols, so it is usually transmitted easily (often cost-free), potentially over long distances, owing to information technologies. Tacit knowledge is not articulated, and is rooted (Maggioni and Uberti, 2009). Informal knowledge is also strongly associated with a place, results from the context and specificity of an area, and originates from a variety of sources, i.e. the science, tradition, culture and economy of that area (Storper and Venables, 2004; Hilpert, 2006).

Consideration of the context and rooted nature of knowledge demands that reference be made to a concept complementary to that of human capital, i.e. to social capital (Coleman, 1988; Putnam, 2000; Cote, 2001). The general conceptualisation of the latter relates to a capacity and skill to achieve cooperation between people within a group or organisation, with a view to common interests being pursued. Reference is thus made to the basic nature of social interactions, as this facilitates understanding of behaviour, in this case decisions regarding the choice of sources of knowledge (as inter alia influenced by peers). As Aldrich (2012) notes, dense social networks and tight bonds with relatives and neighbours differ from weaker ones in providing for readier access to the information needed to operate and coordinate systems of information exchange and knowledge acquisition, as well as for facilitated decisionmaking (Granovetter, 1973). Social capital needs to be taken account of as knowledge transfer is considered. The significance will differ in line with the form this takes, i.e. “bonding” – involving ties between people in similar situations (e.g. family members, close friends or neighbours), “bridging” – involving relationships between friends of friends, or “linking” – involving relationships between individuals and institutions (Putnam, 2000; Woolcock, 2001; Aldrich, 2012).

The Internet and knowledge are interrelated in a bi-directional fashion. On the one hand, the ownership of adequate resources of knowledge allows for the use of ICT, and increases the effectiveness of that use (Czapiewski et al., 2012). On the other hand, the Internet opens up better access to information, with this in turn generating a potential for knowledge to be increased, while also allowing for knowledge acquisition and sharing (Baumüller, 2017). Furthermore, the diffusion of ICT has markedly lowered the costs of producing and disseminating information. Consequently, increases in the volume of information and in numbers of available information sources are both observed (Flanagin and Metzger, 2008). These are all matters of fundamental importance, in agriculture in particular, given the lack of mobility or low level of mobility that in turn reflect strong attachment to a given place (Fujita et al., 1999). Even in its intensive form, agriculture is always a surface-based function, and, in line with the specificity of this form of economic activity, the mobility of the farming population is also strongly limited.

Given the point-wise character and concentration of traditional sources of knowledge in the largest urban centres, farmers have been thought to experience impeded access (Czapiewski and Janc, 2011). Equally, the instantaneous nature of appearances on the Internet rapidly came to be perceived as a means by which the resistance arising out of physical distance might be overcome (Gannon, 2008). Utilisation of the Web as an essential source of information thus entails a “shrinking” of physical distance with respect to the sources of knowledge, and this ought to denote better transfer of knowledge in agriculture, and a curbing of conditions disadvantageous to productive activity.

Specifically, the Internet allows for the exchange and sharing of knowledge (Grimshaw, 2011), e.g. the current situation on agricultural markets, weather conditions for farmers, and government and local-government programmes; while also allowing intermediation to be circumvented as new solutions and technologies of production are sought (Floriańczyk et al., 2012). Therefore, thanks to the Internet, a farmer can acquire knowledge and broaden it, make contact with other producers, promote his/her own products and services, order necessary means of production, and discharge administrative duties (Heilig, 2003; Akca et al., 2007).

Galloway et al. (2011) further indicate that, among the primary benefits accruing from the use of the Internet in farming activity (not least effectiveness of transactions, interaction with customers, and improved functioning of the logistic chain), there is the very important matter of networking, as connected with the sharing of knowledge. In that regard, use of the Internet has a key role to play in reducing information asymmetry otherwise likely to arise (Jeffcoat et al., 2012). In particular this implies equalisation of chances when it comes to agents’ functioning, irrespective of their physical locations and opportunities to access sources of information “on the ground”. And, as Marra et al. (2010) noted, better access to information entails increased knowledge on innovations capable of being introduced on farms, with uncertainty as to the potential benefits of these being introduced reduced at the same time.

At this juncture it is appropriate to ponder whether resort to the Web genuinely signifies the possibility of all kinds of knowledge being acquired. And, while this is certainly true with respect to codified knowledge, a far more problematical matter is tacit knowledge, bearing in mind the spatial context associated with that, and the significance of proximity. While Feng et al. (2005) suggest a contribution of the Internet in transferring some tacit-knowledge resources (i.e. the so-called intermediate tacit knowledge); and while Torre (2008) emphasises that permanent proximity is not a precondition for the knowledge transfer process relevant to this (given that it may also be temporary or transient, for example relating to meetings at conferences, workshops, short-duration internships and exchange programmes), there remain questions regarding the transmission of the pool of experience, acquired practice, etc., in their entirety.

Worth considering alongside the positive influence of the Internet on knowledge-acquisition processes in regard to the running of a farm is the possibility that the results of information dissemination might actually be negative. For some farmers, the Internet may seem to pose a threat to traditional, inter-generational channels of knowledge transfer (Wójcik et al., 2019), as well as to jobs (Pantea et al., 2014). Equally, as use of the Internet requires the necessary skills, a lack thereof may only deepen the disparities already noted between farmers. This is especially important in the face of limited WWW resources in” more minor” languages (Kale et al., 2015), and the access to some resources being in consequence limited further – to just some groups of farmers. On the other hand, an orientation towards” international knowledge” denotes a lack of rootedness of information obtained in any local area or on any local market (Aker et al., 2016). Yet a further key challenge relating to farmers’ use of the Internet pertains to the locating of trustworthy and reliable information, as opposed to fake news (Lubell and McRoberts, 2018).

4 Study area and methodology

If smart farming is to operate effectively, not just on selected large farms, it is necessary for farmers to have the knowledge on the concept as such, the conditions underpinning its implementation, and the functioning of particular component technologies. A farmer should then possess certain minimum skills when it comes to data (e.g. in quantitative and qualitative statistical analysis, error identification and the management of (e.g. Big Data) databases). This will allow conclusions to be drawn from analysis, with the result that production processes are actually improved (van Huylenbroeck and Durand, 2003).

All of the above denotes a starting point to be found among such fundamental questions as: how do the farmers use the Internet?; what information do they look for?; and what are their priorities as they run their farm? Such issues relating to knowledge and basic skills with the use of ICT look particularly important in regions where agriculture is based on small-scale farms, with crops grown and livestock raised on small pieces of land lacking advanced and expensive technologies (Kirsten and van Zyl, 1998), and on which a single person is usually responsible for the majority (if not all) of activities relating to management and production.

On a European scale, Poland can be seen as a country of average-level development when it comes to the bases upon which the information society functions. As Cruz-Jesus et al. (2016) made clear, Poland is close to the EU average for the adoption of ICT by individuals, while being in a clearly worse-off position for eLearning, cross-border eCom, as well as civil-society participation. Where access to the Internet is concerned, 2017 data revealed that Poland resembled certain other CEECs in having under 90% of its households with fixed broadband availability. This is a relatively low value for Europe, given the numerous countries in which the parameter is now at 100%. In rural areas, Poland is again below the EU average, with a value for fixed broadband coverage just exceeding 80% (Broadband Coverage in Europe, 2017, 2018).

Poland is an interesting country in which to study the conditions underpinning and effects arising from transformations in agriculture (and rural areas more broadly) given its status among CEECs transitioning both out of communism and into the EU. This transition process in farming and the rural economy as a whole is still ongoing, not least in line with a still-declining role in the overall economy. At 2.2%, Poland’s 2016 share of GDP accounted for by agriculture is comparable with figures in the wider region (Table 1). And, while the share of total employment accounted for by agriculture as of 2017 was still high, at around 10%, the corresponding figure was at 18% when Polish membership in the EU commenced 15 years ago. Part of the change that denotes reflects modernisation across the sector, and that has been accompanied more recently by improved revenues that may now assure farmers of average wages comparable with those in other sectors. This is in turn linked with both professional advancement and innovation, as well as strategic openness to change and a willingness to learn constantly (see, e.g., Markuszewska, 2015; Fałkowski et al., 2017).

Table 1 A basic characterisation of agriculture in selected CEECs.

Country	Population ^a	Average farm area ^b	Contribution of agriculture to Gross Domestic Product ^c	Very small farms ^d	Family farms ^e	Employment in agriculture ^f	Farmers with full agricultural training ^g
Poland	38.0	10.2	2.2	64.8	99.3	10.1	27.4
Bulgaria	7.1	22.0	3.2	81.0	95.4	17.5	6.5
Czech Rep	10.6	130.2	0.8	31.7	86.0	2.7	38.7
Hungary	9.8	10.9	2.6	83.2	97.5	5.7	4.4
Romania	19.6	3.6	4.1	94.6	99.3	23.0	9.1
Slovakia	5.4	73.6	0.7	68.3	87.0	2.1	9.2
EU 28 average	511.5	16.6	1.2	67.6	94.7	4.2	9.1

Source: Eurostat.

a (2017).

b (ha) (2016).

c (2017) (%).

d (with < EUR 8 000 of standard output) (%) (2016).

e with > 50% of regular labour from family members) (%) (2016).

f (%)(2016).

g (%)(2016).

However, further consideration of Poland’s ratios regarding generation of GDP and share of employment suggests that the process of transition in agriculture still has some way to go. Furthermore, Polish agriculture remains very differentiated in structural terms. Thus, more than half of all farms in Poland cover less than 5 ha, while more than 20% of all farmland is in the hands of just the 1% of largest farms (Charakterystyka gospodarstw rolnych w 2016 roku, 2017Charakterystyka gospodarstw rolnych w 2016 roku, 2017). That just serves to underline a basic feature of the deepening disparities now observable between farms which have embarked upon modernisation and those that have adopted a persistence (subsistence) strategy, inevitably denoting a gradual loss of economic functions (Wójcik et al., 2019).

Empirical material for the study reported here was acquired by way of a questionnaire-based survey carried out among farmers in 2014. The questionnaire involved an indication of the most important sources of knowledge, an assessment of the degree to which information is used in running a farm, as well as insight into the means of acquisition of knowledge relating to production and the sale of produce. Socio-demographic characteristics of respondents and the economic profile of their farms were also taken account of. On the basis of there being 1.42 M farms in Poland in 2014, and in line with a 95% confidence level and 2% margin of error, the sample size needed for the research was set at 2397 at minimum (Brannen, 1992; Rogerson, 2001; Hesse-Biber, 2010).

The research began with the development of a ranking of rural municipalities (at the LAU2 level ⁵ ) in regard to the significance of the farming function. The aim here was to confine analysis to those areas in which agriculture was a genuinely significant profile for inhabitants’ activity. A further categorisation involved a breakdown of farms in all 16 Polish regions by area (into categories of less than 10 ha, 10–15 ha and more than 15 ha). It was then on the basis of this ranking that a selection of 67 municipalities (located in all regions) was made, with these by definition being areas of diversified farm-size structure in which the farming function was of genuine significance.

The distribution of questionnaires among farmers took place via all the primary schools located in the municipalities under study. In each case, school authorities were asked to hand out questionnaires to all pupils whose parents (or grandparents) engage in farming activity. Ultimately, 2411 responses (fully filled-out questionnaires) were received, and from almost all of the selected municipalities located in all Poland’s regions ⁶ .

While the method applied does not allow all farmers to be reached, the lack of access to other effective distribution channels (e.g. involving microdata addresses, including e-mail addresses) makes this the best way of ensuring that a large population is studied. This methodology has been applied repeatedly in Poland (e.g. Komornicki et al., 2013), and is subject to confirmation that results obtained in other ways (e.g. via telephone conversations or the Internet) are comparable. The major downside is nevertheless that information is obtained from a rather closely-defined socio-demographic group (consisting of still-mobile people who are parents, or at least who live alongside children).

Among respondents, 66% were male, while the average age was 42 years. Farmers with university-level education accounted for 8% of the total (those with agricultural education for 3%) ⁷ . That left 42% of the sample with secondary education (with secondary agricultural education accounting for 22%). Overall, more than half of all respondents lacked formal agricultural education. On average, at the moment they filled in the questionnaire, respondents had been managing their farms for more than 18 years. This meant that they had very important farming experience at their disposal, as well as comprehensive knowledge of procedures related to farming activity. The average acreage of the farms respondents managed was 17.4 ha, denoting a slight over-representation of the larger (over 15-hectare) farms in comparison with the typical structure known for Poland, with some attendant under-representation of farms in the smallest size categories (below 5 ha). In this regard it should be noted that the last two decades have seen areas of farms in Poland unchanged in half of all cases, while increasing in area in as many as 45% of cases (Table 2).

Table 2 Key characteristics of the farmers studied.

Variable	group	frequency (n)*	percentage
Sex	Male	1583	66.8
	Female	787	33.2
Age	up to 29	37	1.6
	30-39	902	38.5
	40-49	950	40.5
	50-59	373	15.9
	above 59	81	3.5
Education	primary	182	7.5
	at trade schools	1007	41.8
	secondary general	180	7.5
	secondary vocational	809	33.6
	higher	186	7.7
Agricultural education	none	999	41.4
	agricultural course	275	11.4
	vocational	479	19.9
	secondary	498	20.7
	higher	631	2.6
Self-assessment of farm	developing	481	20.0
	stable	1302	54.0
	lacking development prospects	553	22.9
Income from agriculture	above 80%	865	35.9
	60-80%	287	11.9
	40-60%	353	14.6
	20-40%	414	17.2
	below 20%	372	15.4

Source: author’s own research.

* the totals for answers in different categories differ on account of the lack of responses to certain questions.

The study offers confirmation of the progressing development of the disparities among Polish farms already referred to. In the case of roughly 1/3 of surveyed farms the income from agriculture constitutes the main source of upkeep. At the same time, these are the farms selling almost their entire production externally (on the basis of a modernisation strategy). That leaves 1/3 of all farms in which the opposite situation (involving the persistence strategy) is to be observed. The phenomenon is made yet more visible by analysis of self-assessment of farmers in relation to the economics of their own farms, with only 20% of respondents treating these as “developing”.

5 Results

Worth noting at the outset is the limited inclusion of small villages of fewer than 100 inhabitants, given the real situation where Internet coverage and access in Poland are concerned (Janc, 2017). This is despite a definite closing of the gap between towns and cities on the one hand and rural areas (i.e. areas formally enjoying rural status) on the other (Fig. 1). That still means – on the whole – that any failures to use the Internet in rural areas are less and less associated with purely infrastructural or technological factors. This is also true for broadband, an important element in capacity to introduce smart farming effectively.

Fig. 1 Access to the Internet in Poland in the years 2004–2016 ¹⁰ (A – overall, B – broadband).

Source: author’s own elaboration based on data from Poland’s Central Statistical Office (now Statistics Poland).

Of all professional groups in Poland, farmers constitute the one (other than pensioners) making the least-frequent use of the Internet. According to data from the Central Statistical Office, in 2017 regular use was made of the Internet by some 50% of farmers (persons aged 16–74 enjoying the status of paid employees working a farm they owned or leased). This is an indication of a significant increase above the level noted in 2013 (30%), but both figures are put in perspective by the comparable figure for persons working in other sectors of the economy in 2017, which was 84% (Information Society in Poland. Results of Statistical Surveys in Years 2013-2017, 2017Information Society in Poland, 2017Information Society in Poland, 2017Information Society in Poland. Results of Statistical Surveys in Years 2013-2017, 2017Information Society in Poland, 2017Information Society in Poland. Results of Statistical Surveys in Years 2013-2017, 2017Information Society in Poland, 2017Information Society in Poland. Results of Statistical Surveys in Years 2013-2017, 2017).

Farmers are not taking full advantage of the possibilities online services represent, in either their private or their professional lives. Only 14% of farmers access the Internet via a smartphone, compared to 47% for remaining employed persons ⁸ . Likewise, where 2% of farmers store data in the Cloud, the figure for all others in work is 19%. Further comparisons of the same kind relate to the use of social media – 48% compared with 64%, use of online banking – 28% vs. 63%, and online shopping – 58% vs. 76% (Information Society in Poland. Results of Statistical Surveys in Years 2013-2017, 2017Information Society in Poland, 2017Information Society in Poland, 2017Information Society in Poland. Results of Statistical Surveys in Years 2013-2017, 2017Information Society in Poland, 2017Information Society in Poland. Results of Statistical Surveys in Years 2013-2017, 2017Information Society in Poland, 2017Information Society in Poland. Results of Statistical Surveys in Years 2013-2017, 2017).

It is also important to determine the place the Internet occupies, among other sources of the information needed to engage in farming activity. From among the 17 potential sources of information detailed in the questionnaire ⁹ , farmers most often indicated TV/radio (58% at least once a month), then knowledge present in the family (38%) and the professional press (37%). Meanwhile, the web-based (online) services were indicated as an “often used” source by 34% of respondents. All of these sources represent a constant presence in the surroundings of farmers, and access to them is easy. Several times a year farmers make use of sources located farther away (both physically and socially), including specialised shops (46%), neighbours (44%) and producers (43%). Farmers overall rarely use (while 45% of respondents never use) specialised fairs and exchanges as occasions at which to gain information. There is also a virtual absence of NGO operations as sources of information, with more than 80% of respondents not availing of this kind of help. A similar situation applies to the branch organisations (eschewed by 72%). It must also be added that quite a large share (35%) of respondents have never used the Internet for such information-related purposes.

Questions allowing the significance of particular sources to be assessed reveal the major importance most frequently attached to the family (by 30%), followed by radio and TV broadcasts (29%), and then by press and the Internet (21% each). Equally, where the latter category was concerned, as many as 28% of respondents stated simultaneously that this source lacked significance where increasing effectiveness of the farm they managed was concerned.

It can be concluded from this summary that, while the Internet does play an important role from the point of view of frequency of use as a necessary source of information, a considerable share of all farmers see it as not bringing any essential value, and do not therefore make true use of it. In this sense, the Internet loses out to the “traditional” sources of information – whether acquired from the social environment in which farmers operate, or associated with older technologies.

Both the frequency of use of the Internet as a source of knowledge, and the perception of its significance, were found to be higher where the level of general and professional education of farmers was higher, and where the share of output sold externally was higher. Similarly, farmers assessing their farms as “developing” (in the context of economic prospects) make use of the Internet one-and-a-half times as often, and evaluate its utility more positively by 30%, than do those farmers not seeing their farms as having prospects for development.

For the farmers involved in the study, the functions most important to their professional activity are those associated with passive use of the Internet, i.e. with information acquisition. This is referred to in the five categories indicated most frequently (all referred to by more than 28% of respondents) as key from the perspective of the functioning of a farm (Fig. 2). The categories also offer a distinct perception of the conditions in which farming activity is pursued, in relation to external subsidies, technologies and means of production. The active forms of Web use are used to only a relatively limited extent, though most can find direct expression in increased effectiveness and increased revenue-generating possibilities for a given farm. Only 16% of respondents used online banking, and even less so the functionalities allowing resources of explicit and tacit knowledge to be increased via acquisition of practical skills and exchange of information with other farmers (participation in web fora, online training and courses was found to apply to only just over 3% of respondents). The use of the Internet as a medium providing for the sale/purchase of goods also plays only a marginal role.

Fig. 2 Activities engaged in through Internet use.

Source: author’s own research.

The patterns of use of the Internet for farming-related purposes are seen to depend on basic characteristics of farmers and their farms. Table 3 details the use of particular functionalities in relation to the general level of education of responding farmers. As remaining analyses imply, this particular characteristic of farmers is crucial in terms of the choice of knowledge sources and use of the Web. The better-educated farmers are found to take more frequent advantage of virtually every kind of functionality, and in particular of those contributing to flexibility of the production process.

Table 3 Use of particular functionalities as related to general levels of education of farmers (percentage values).

Level of education	information on EU funds	information on the possibility of selling produce	information on new technologies	information on the purchase of means of production	online courses	surveying agricultural websites	online banking	purchase of means of production	advertising own produce	selling own produce	participation in web-based fora	managing own profile on social media
Tertiary (n = 168)	66.7	31.0	57.1	43.5	2.4	53.0	41.1	32.7	7.1	13.7	10.7	0.0
General secondary (n = 155)	57.4	40.6	52.9	38.7	5.2	47.1	17.4	11.6	1.3	8.4	5.8	0.0
Vocational secondary (n = 688)	52.0	40.0	50.9	34.6	4.5	42.0	21.9	16.0	4.5	8.9	4.9	1.2
Trade school (n = 742)	51.1	34.2	45.4	37.9	4.0	37.2	15.2	13.2	3.1	7.3	3.0	0.7
Primary (n = 113)	53.1	32.7	30.1	27.4	3.5	37.2	15.0	10.6	2.7	7.1	1.8	0.9

Source: author’s own research.

Thus, for instance, use is made of online banking by 41% of those who use the Internet at all and have tertiary education, as opposed to 15% of those with primary education. Likewise, online purchase of means of production is a matter for 33% with tertiary education and just 11% of those only educated to primary level. Better-educated farmers are also more likely to sell their produce with the help of the Internet.

The above forms of Internet use allow for the reduction of the negative influence of physical distance on engagement in key activities connected with farm management, and also allow for better time management. Likewise, when it comes to efforts aimed primarily at obtaining information on various subjects, better-educated respondents generally take more frequent advantage of the Internet, though here differences are slight.

It is important to note that “active presence on web-based fora associated with farming” is yet a further domain of better-educated farmers. This kind of activity is important insofar as it implies involvement in the transfer of tacit knowledge – sharing and exchange of experience and information with other people. The pursuit of farming business requires, on the one hand, a knowledge of processes and undertaken actions, and, on the other an active attitude to the search for new resources of knowledge. It should also be noted how lack of use of the Internet for the purposes of farm management relates closely to social features within the studied population – as 38% of farmers with primary education did not use the Internet for this purpose, compared with only 10% of those educated to higher level.

The results presented make very clear the difference between farmers of differing levels of education when it comes to use of the Internet to meet needs as regards the running of a farm. In the context of professional decisionmaking – farmers’ own assessments of their farms, monitored operations, better acquaintanceship with the complexities of the farm environment, and the introduction of new, more-sophisticated solutions into the farming process, it is each time the better-educated farmers that fare better.

Beyond the issue of whether Internet functionalities are taken advantage of at all in relation to farm management, a further key matter concerns the number of such functionalities farmers use. The most typical situations involve respondents using one such functionality (in 24% of case) or at most two (28%). Only 13% of respondents take advantage of five or more functionalities. And in this context, the evident determinants of more-versatile use of the Internet are: (1) general educational level of farmers – given that 40% of those with primary education make use of just one functionality, compared with just 13% of those with tertiary education, while 5% of respondents with primary education make use of five and more functionalities, compared with 30% where tertiary education has been completed; (2) the fact of farmers having had a professional agricultural education; (3) a larger share of overall household income deriving from farming; (4) a more-pronounced orientation towards the sale of agricultural produce; (5) a better economic situation in general (as regards the extent to which the farm can be considered to be developing); and (6) an increase in farm area achieved in the recent period.

The trend outlined above is further confirmed by reference to average values for numbers of functionalities respondents use, as set against selected social characteristics of farmers and their farms (Fig. 3). Clearly, the factor differentiating the population under study most markedly is level of education, be that general and/or professional. A particularly visible “transition” is between the highest educational levels – general tertiary (3.6 functionalities used on average) and agricultural tertiary (3.9) – and the ones just below that, with general secondary education on 2.8 and secondary agricultural education on 2.9. Similarly distinct is the strong age-dependence, with the youngest respondents making use of 30% more web functionalities on average than their oldest counterparts.

Fig. 3 Average numbers of functionalities of the Internet used in relation to selected characteristics of farmers and their farms.

Source: author’s own research.

Remaining characteristics also relate to the number of online functionalities made use of, especially to the relationship between how long the Internet has been present in the given household and the respective level of activity. A longer period of presence of the Internet is seen to denote more-versatile use. However, the questionnaire-based study implies that the majority of farmers only started to use the Internet post-2005. Up to that time, only 16% of respondents’ households had Internet access at all, while by 2008 that figure had already exceeded 50%, and by 2010 it was close to 80%. Such results illustrate clearly how socio-demographic factors do much more to differentiate the population of farmers clearly, from the point of view of behaviour used to acquire knowledge via the Internet, than do features of the farm these people run.

6 Discussion

Results obtained on the basis of questionnaire-based research among Polish farmers point to several key features associated with the use of the Internet as a source of information serving the pursuit of business activity. However, conclusions arising on the basis of these need to bear certain methodological limitations in mind. The subjects of study were a particular group of farmers, and the work cannot be said to offer a true reflection of the farming population as a whole, given a (school-distribution) methodology that ensures a sample younger than the overall population and better-educated, as well as the inclusion of farms that are somewhat larger than average, and somewhat more commercial (given the presence of farmers who basically live in a household at least including a young generation). Another limitation of study is fact that estimates of sampling and non-sampling error/biases were not included.

Beyond that, it can be noted that the Internet does not constitute Polish farmers’ most important source of information, as it still lags behind traditional sources present in the social environment for much longer, and most especially sources of a personal nature present within family or among friends and neighbours. A clear significance of other farmers as sources of information is not confined to Poland, having for example been noted among New Zealand’s sheep and beef farmers (Corner-Thomas et al., 2017), or the farmers of Ohio (Diekmann et al., 2009) and Illinois (Villamil et al., 2012). This finding can be linked to the significance of social factors in the transfer of knowledge, and, consequently to preferences where sources of information are concerned.

As Roux et al. (2006) note, the diffusion of new knowledge is possible provided there are no social and cultural divisions between authors and users. Social factors also relate to the notion of context as a factor shaping the behaviour underpinning acquisition of information. Patterns of behaviour and information needs are different, depending on the context (McCreadie and Rice, 1999; Johnson, 2009). In the cases of family and neighbours, a common context involving lack of socially separating distance is usually ensured. The Internet does not provide such a guarantee, especially to those generations of farmers for whom it has not constituted an everyday instrument of communication or information acquisition. As we set our results against the background of knowledge-related issues, we note that Polish farmers prefer those sources of knowledge associated more fully with a greater capacity to transmit tacit knowledge. However, once use of the Internet has been commenced with, behaviour of users becomes oriented towards the acquisition of codified knowledge – mainly comprising pieces of information of various kinds.

Unlike bonding (exclusive), the bridging (inclusive) type of social capital is a vector for the development of new ideas, values and perspectives. According to Granovetter (1983) weakly-tied clusters (capable of being identified with bridging social capital) have more opportunities for diverse knowledge- and information-sharing. On the other hand, when it comes to farmers lacking opportunities to use formal sources of information (via linking social capital), these people are often seen to rely on informal sources of information, with agricultural knowledge transferring via farmers’ social interactions (Conley and Udry, 2001; Saint Ville et al., 2016; Pratiwi and Suzuki, 2017). The same conclusions can be drawn from the analysis of material obtained in the study of Polish farmers. First of all, contacts and direct talks with the nearest family, neighbours or representatives of various institutions remain the most important sources of knowledge (see also Wójcik et al., 2019). Secondly, if farmers already use the Internet, they are usually passive, do not actively participate in internet forums and specialist groups on social media, and therefore do not shape any form of social capital.

The Internet also needs to be treated (after Kavanaugh et al., 2005) as a vector that can integrate innovation in agriculture. By broadening and strengthening networks of linkages within agricultural communities, the Internet provides for the sharing of ideas and for a mobilisation of knowledge in the interests of innovative activity (Kaushik et al., 2018). The results presented here show clearly how the Internet does not play a major role for Polish farmers, when it comes to use being made of bridging social capital. The Internet is far more a source of information than it is a medium via which interactions can be established and maintained.

Our observed, very-high share of people making no use whatever of the potential in farm management the Internet has to offer suggests the formation of “two-speed” farmer populations, from this point of view at least. Such processes are also visible in countries better developed in socio-economic terms. Farrington et al. (2015) indicate that the inhabitants of peripheral rural areas in the United Kingdom differ from inhabitants of other areas in enjoying more limited chances of becoming more-advanced Web users who take advantage of a broader spectrum of Internet-based solutions. This also means lower chances of becoming so-called “next-generation users”, i.e. those using various devices to access the Web, also from a variety of locations (Blank and Dutton, 2014). This kind of process had been diagnosed well before, and was termed the digital vicious circle (Warren, 2007). New Internet solutions and forms of use are adopted more easily by the more-experienced, and this may lead to a fundamental worsening of disparities in capacities for development between different groups. Salemink et al. (2017) summarizing a massive body of scientific literature about ICT use on rural areas, stated that in significant amount of papers scholars consider socioeconomic context, level of education, skills, attitudes as a important factors of digital inequalities. The work presented here supports the contention that, in countries in transition whose farming is differentiated strongly in terms of farm area, market orientation, and the socio-demographic characteristics of farmers (Charakterystyka gospodarstw rolnych w 2016 roku, 2017Charakterystyka gospodarstw rolnych w 2016 roku, 2017), a considerable part of this population has only weak bases upon which smart farming solutions might be introduced.

The use of the Internet by farmers in Poland is primarily passive, with the main aim being acquisition of information. Active forms of Web use requiring interaction on the part of the user are much less common. Our analysis further justifies a contention that farmers studied have Web-oriented skills that are still at an early stage of development. Resources of digital competence are thus limited, and reasons for this situation should be looked for in: (1) a low level of knowledge and skill when it comes to operation of the devices making use of the Internet possible – with the effect that farmers are prevented from taking active advantage of possibilities the Internet offers; and (2) a failure of farmers to see much benefit for themselves out of using the Internet to buy or sell agricultural produce – with the effect that little or no use of such functionalities is made.

Noting the experiences of farmers in Western Europe, as well as the growth of e-services (including e-commerce) among all users in both Poland and other CEECs (see, e.g., Janc, 2017), it should be supposed that the first aforementioned factor plays a leading role; and that is all the more the case – as Seri et al. (2014) anyway indicated – given the way that adequate infrastructural advancement, and in particular access to broadband Internet, is such an important factor where the use of e-services is concerned. However, as such technological disparities between town and countryside noted previously in Poland are diminishing steadily (Fig. 1), increasing importance needs to be attributed to “soft” factors like competence and knowledge.

Analysis likewise indicates how strongly related use of the Internet is to such basic social characteristics of farmers as level of education and age, as well as key features of farms themselves. Confirmation of the results of analyses of the adoption of solutions from the domain of precision farming is thus offered. Work on farmer behaviour in Germany and Denmark by Tamirat et al. (2018) also stresses the importance of participation at workshops and exhibitions, given the way this increases farmers’ propensity to introduce modern solutions on their farms. However, the example of Poland shows that farmers do not have to have these kinds of sources of information, as they rather concentrate on those available in the immediate environment, indeed virtually “without leaving the farm” at all.

A will to introduce new solutions is also found to depend on numbers of information sources (Lambert et al., 2015), as well as use of a smartphone with Internet access (Castle et al., 2016). Such issues should be looked at from the point of view of the number of Web functionalities farmers actually use. As our research results show, aspects like farmers’ levels of education, ages and market orientations combine with the length of the period over which the Internet has been used to influence the number of Internet functionalities resorted to. As Park (2017) noted, technology forms an environment in which experience is an important resource, influencing use of new devices and services. This is one aspect making up the so-called digital capital, as constituted by a set of characteristics that influence people’s interaction with the Internet, and the way use is made of it. This resource thus has major capacity to bring benefits and increase profits from farm management, especially where functioning in the reality of smart rural development is concerned.

Investigations among Hungarian farmers first demonstrated how significant an issue farm size could be. A key conclusion concerned the lack of suitability of introducing advanced solutions on smaller farms, above all given the high investment costs denoted (Takácsné György et al., 2018). Our study offers up similar conclusions, while it is likewise significant that younger farmers elsewhere prove more willing to introduce newer solutions (Barnes et al., 2019), as the present study supports this assertion as well.

Pointing to the main challenges “digital farmers” face, Cho (2018) notes the strong link between the model of smart agriculture and knowledge resources. Upskilling among farmers is thus a must, as is awareness of constant change (of conditions) and an adequate level of digital literacy. In this context, our conclusions regarding features of farmers and their farms reveal a significant social barrier to the creation of bases for smart agriculture. The very technology involved, let alone its use in increasing productivity, demands that a social foundation take shape, with farmers in this way endowed with adequate digital capital. In general, farmers in Poland make insufficient use of the Internet as a source of knowledge and support regarding production. Limited digital competences and largely undiversified forms of Internet use form a weak link in undertakings aiming to put in place a smart agriculture that can actually function.

For such reasons, Poland resembles other countries in remaining on the starting block where the smart revolution in agriculture is concerned. Furthermore, the socio-demographic characteristics of Polish farmers link up with Internet use in a manner that suggests only limited possibilities for agriculture to transition in the direction of a smart phase. Large market-oriented farms, managed by well-educated farmers who function easily in an environment filled with new Internet-based technologies have a chance to move effectively into the application of solutions from the domain of smart agriculture. In contrast, the majority of Polish farms burdened by such disadvantages as small area, limited market orientation and limited possibilities for investment in advanced solutions, as combined with disadvantageous social characteristics, will have clearly-limited possibilities for transformation within the framework of the smart-agriculture paradigm.

7 Conclusions

The development of Polish agriculture through adoption of the smart farming model is conditioned economically and technologically (on the one hand), but also socially (on the other). Polish farming continues with its process of transition, with changes taking place needing to be perceived, not only from the structural-transformation point of view, but also increasingly in relation to institutions and technologies. Indeed, the technological revolution Polish agriculture faces is different in nature from those of the last 30 years involving systemic transformation away from communism, and then (in fact almost simultaneously) European Union accession. This is because the new kind of change is more conditioned socially.

The relevant technological challenges Poland faces can be considered in the very same context. New technologies and arising digitally-based innovations provoke apprehension and fear, as once did the newly-encountered economic and institutional conditions. That may mean technology being perceived as threatening by farmers, as pressure exerted from outside and working to erode a feeling of common interest based around collective behaviour. Thus far constituting one of Poland’s best-integrated interest groups, farmers may feel that new technological requirements work to undermine the previous feeling and institutions associated with common knowledge acquisition very much based on local ties of blood and neighbourliness. Furthermore, the way new technologies are dealt with proves a highly individualistic matter, via which the linkages taking shape tend to bypass hitherto-existing structures. All of this means that a challenge to the introduction of the smart model into farming lies in the development of a new knowledge culture founded upon common experience of digital change and its absorption in local environments.

Given that the main factors influencing the intensity and effectiveness of Internet use are relevant skills and competences (as quantified in particular in line with the level of education), it is necessary to recommend that greater emphasis be placed on education. It is in this way that conditions will take shape for people to build up their own digital environments, with fundamental capacities allowing for use of the functionalities the Web offers.

In contrast, a lack of adequate knowledge and skills regarding use of the Internet acts to narrow the scope of opportunities for farm development. In the face of that, comprehensive solutions by which a properly directed and specialised educational offer and assistance are put in place (as opposed to just a series of independent, single-shot initiatives) are needed to encourage increased use of the Web by farmers, ensuring its treatment and perception among them as an agricultural instrument of fundamental significance, or in fact as nothing less than a means of production. In contrast, EU-funded undertakings pursued in the years 2007–2013 by Poland in particular, and the CEECs in general, sought primarily to support technological infrastructural development (Czapiewski et al., 2012).

Even more fundamentally, a targeting of the competences of farmers and efforts made to increase their digital capital cannot operate in isolation of efforts striving to achieve further essential changes in farming structure. In this regard, the pursuit of principles linked with the implementation of smart rural development is crucially hampered by the fact that too large a share of all farms in Poland are small, and feature only very limited market orientation.

Finally, beyond the Polish context, we can state that intensifying the Internet use in countries with a high share of family farms requires attention to social and cultural factors. In this case, it is important to achieve a balance between technological progress (digital competence) and the social context of such innovations (e.g. level of education, social competences). The results of recent studies have shown that a major role in the process by which farmers’ knowledge takes shape is played by microsocial conditioning (Wójcik et al., 2019). This is confirmed by the observations that Andersson et al. (2016), among others, present; as well as those of Koster et al. (2014), who argue that local conditions is of key significance to the transfer of knowledge (see Nooteboom, 2000). Thus, we should pay more attention to issues related to neighbourhood information sharing practices between farmers – how they obtain information, what is exchanged.

Acknowledgments

We would like to thank the two anonymous Reviewers and Dr. Emma Jakku whose comments/suggestions helped improve and clarify this manuscript.

Notes

1 According to World Bank, ICT (Information & Communications Technologies) consists of hardware, software, networks, and media for the collection, storage, processing, transmission, and presentation of information (Information and Communication Technologies. A World Bank Group Strategy, 2002). In this article, ICT refers to the computer and internet connections used in the handling and communication of information.

2 Research conducted within the framework of Scientific Project 2011/01/D/HS4/03295, Models of knowledge transfer in agriculture and its influence on agricultural productivity – spatial analysis, as financed by the Polish National Science Centre.

3 After Renjen (2018), Industry 4.0 could be described briefly as analytics, artificial intelligence and IoT bringing together physical and digital technologies with a view to creating digital enterprises that are interconnected and therefore capable of more-informed decisionmaking.

4 In the case of the Web of Things, devices communicate with one another via WWW protocols, such that applications and services can be put in place. Application of existing WoT standards facilitates integration between many different pieces of equipment and installations, as well as their servicing at user level by way of online applications, as in the case of remote home-security system monitoring and administration, and remote home electricity grid usage monitoring.

5 The lower LAU level (Local Administrative Unit 2, formerly NUTS level 5) consists of municipalities or equivalent units in the 28 EU Member States.

6 For such a number of responses to be received, it was necessary to send out 15,312 questionnaires in total. In fact, a response rate of 15.7% would need to be regarded as relatively high, bearing in mind that a large proportion of questionnaires reached country-dwellers not even involved in farming. Furthermore, several schools consenting to participate at the outset ultimately went on to withdraw their cooperation.

7 All figures represent absolute values (raw sample percentages).

8 In the research cited, farmers were identified as one of the categories of people in work. The category of other employed persons thus relates to all of those in non-agricultural professions.

9 All categories were elaborated a priori, by way of an analysis of the subject literature.

10 The higher the value of the measure calculated in this way, the smaller the gap separating rural areas from towns (the perfect situation is when the value equals 1 – meaning the same level in all areas).