The Importance of Social Norm on Adopting Sustainable Digital Fertilisation Methods

Abstract

Precision farming systems promise a highly efficient resource use reducing cost for farmers and contributing to the preservation of the environment. A major obstacle, however, in such systems dissemination is the reluctant adoption by farmers. Prior work is suggesting that mainly knowledge or missing resources for investment are barriers, while social norms are rarely addressed for precision farming systems. We adopt the reasoned action approach including behavioural, social, and control aspects to analyse intentions and actual use of sustainable digital fertilisation methods. Based on a German sample of farmers, we find that social norm is the major predictor, while behavioural and control aspects surprisingly are not relevant at all. The results contribute to the understanding of what drives farmers in adopting precision farming systems on a theoretical basis and highlight the importance of considering social norms in increasing adoption.

Keywords

organizational behaviour environmental attitudes and behaviour structural equation modelling environmental management systems clean technologies

Introduction

The human population is expected to reach 11.2 billion in this century (United Nations, 2017), creating an urgent need to increase food production without transgressing planetary boundaries. Focus on this challenge has been triggered by rising societal awareness, as predominantly observed in the global North, of the need to ensure more sustainable land use and food production (e.g., Popp et al., 2014). Such sustainability implies producing more commodities from less land and water, which in turn requires increasing yields and sustainability of the production system. Framed in the concept of sustainable intensification of agricultural land use, improved interaction with the eco-system and increased resource use efficiency, including with respect to fertilizers, have been proposed, and a large variety of measures exists (Pretty et al., 2018).¹ Precision farming, and especially variable-rate input techniques with machine support, as considered in the present study, have been discussed as a key factor (Dicks et al., 2019; Lal, 2016; Leaver, 2011).

Digital support systems offer farmers assistance in complex cropland management, and thereby the possibility for crop- and field-specific demand-actuated input choices, such as nitrogen fertilizer—the subject of our study. By taking the field heterogeneity of soils into account, site-specific nutrient management can increase input efficiency (e.g., Gebbers & Adamchuk, 2010). Environmental harm from intensification can thus be mitigated; for instance, reduced fertilization on marginal sites offers better protection of groundwater resources by reducing the risk of nutrient leaching (Schieffer & Dillon, 2015; Wolfert et al., 2017).

Despite the advantages they offer, site-specific input systems seem to be rather unequally distributed, with comparably low adoption rates. For Germany, the context of our study, fewer than 20% of farms have been estimated to use these systems (Aubert et al., 2012; van der Wal, 2019). The use of optimal site-specific nitrogen management is a complex decision not only from a plant nutritional viewpoint (e.g., Kindred et al., 2015) but also from the farmer’s perspective (Jensen et al., 2012). Debated entry barriers include unclear cost savings (D’Antoni et al., 2012); doubts and perceived risk regarding economic gains related to high investment and learning cost (Barnes et al., 2019a; Pannell, 2006; A. Rogers et al., 2016); a lack of cooperation possibilities (e.g., Kutter et al., 2011) and farmers’ strong beliefs in their own knowledge of field topology (e.g., Barnes et al., 2019b).

Taking into account the fact that site-specific field management represents a paradigm shift in crop management practices (Lindblom et al., 2017), the adoption decision of farmers is itself complex, and has not, thus far, been sufficiently discussed (Pathak et al., 2019). The question of how to overcome adoption barriers to more sustainable farmland management by altering farming practices has long been debated, including in the field of organization and behavioural management. For instance, biased information and misperceptions of climate change risk could form barriers to change current farming practices (Arbuckle et al., 2015). This includes reducing nitrogen application, which has even been perceived as a production risk (Stuart et al., 2012). Thereby, a growing strand of literature emphasizes the role of information and knowledge generation (e.g., Smith et al., 2018), as well as social norms (Burton, 2004; Kuhfuss et al., 2016; Le Coent et al., 2019) for farmers’ adoption of pro-environmental behaviour, including the adoption of digital technologies. Recent reviews, however, have agreed that the role of social norms as a behavioural driver is thus far underresearched (Carroll & Groarke, 2019; Chabé-Ferret et al., 2019; Dessart et al., 2019; Palm-Forster et al., 2019), and clear causal interpretations of widely used demographic factors to explain behavioural patterns are lacking (e.g., Burton, 2014). We aim to help close this gap, and argue that overcoming adoption barriers for sustainable farming technologies requires understanding the influence of social norms on farmers’ behaviour.

We base our analysis on the reasoned action approach (RAA), a causal model proposing that adoption behaviour is determined by behavioural, normative and control beliefs. Using a theory-informed structural equation model, we identify relevant antecedents for adoption decisions among the beliefs leading to attitudes, perceived norms and behavioural control. To our knowledge, we are the first to apply the RAA to complex adoption behaviour for sustainable smart farming techniques. We focus on one study region, Germany, to keep the cultural and legal context constant (cf. Bartkowski & Bartke, 2018). Germany offers an interesting case to investigate variable-rate nitrogen technology (VRNT) adoption. The renewal proposal of the in 2017 renewed Nitrate Directive as a response to the dispute with the European Commission triggered ongoing debates between farmers and policy makers on how to best implement reduced N application (and other nutrients).

Our results point to others’ opinions and influence on adoption behaviour for VRNTs, which highlights the importance of social norms for changing farmers’ crop management behaviour; this represents a major contribution of this study. We include four different social reference groups within normative beliefs, where we find crop advisors’ influence to be negatively related to adoption, neighbouring farmers, other experts and users, and experience to be positively related to these aspects. The results have direct implications. To foster the uptake of VRNTs and other smart input technologies on a larger scale, support for exchange among users and potential users seems to be effective. Our findings thus contribute to the understanding of required organizational actions, such as fostering information exchange and demonstrating new, more sustainable, site-specific cropland management based on digital technologies, for reducing adoption barriers.

The remainder of this article is organized as follows. In Section 2, we detail the current debate on adoption barriers to sustainable digital farming techniques, followed by outlining the theoretical background of the RAA and the hypotheses in Section 3. We describe the sampling, the survey and the respective structural modelling approach in Section 4 and present the results in Section 5. The discussion and respective implications follow in Section 6, and in Section 7 we conclude.

Information Technology (IT)-Based Crop Nutrient Management as an Enabler for Sustainable Farming

In order to achieve a sustainably secure agri-food sector, smart mixes between traditional and organic crop farming have been proposed, where traditional farming systems need to become more sustainable by including new technologies (Meemken & Qaim, 2018). Thereby, precision farming technologies denote the most promising concepts (Dicks et al., 2019), making large-scale adoption unavoidable (Finger et al., 2019; Weltin et al., 2018).

Use of such farming technologies enables more precise field management, thereby reducing environmental harm from intensification and offering economic and environmental benefits (Weersink et al., 2018). The demand-actuated application of inputs offers the potential to reduce the ecological footprint of agricultural production by means of lower fertilizer use (Walter et al., 2017)—for instance, by reducing greenhouse gas emissions and energy from reduced fertilizer production (Snyder et al., 2009), or by saving fossil fuels from reduced driving (Jensen et al., 2012). The latter also provides benefits for soil life by means of reduced risk of soil compaction (Haneklaus & Schnug, 2002). In addition, site-specific fertilization can reduce harm by reducing the risk of nitrate leaching by reduced fertilizer on marginal sites, even in regions with restrictions on nitrogen application levels and problematic leaching (Bongiovanni & Lowenberg-Deboer, 2004; Rose & Chilvers, 2018).

Productivity can also be increased due to savings not only on fertilizers, herbicides, and other pesticides but also on other production inputs, such as labour and fuel (Balafoutis et al., 2017; Jensen et al., 2012). Likewise, cost savings can be achieved via machine support in the form of VRNTs; these are used to guide farmers’ nutrient decisions, and reduce documentation obligations and help meet standards (Busse et al., 2014; Link et al., 2006).

Understanding plants’ nutrient demand and the respective optimum are complex tasks because they depend on many variables, such as nutrient sources beyond fertilizer, soil content, and moisture or weather conditions (e.g., Kindred et al., 2015). While ideas of site-specific fertilization are not new, it is machine support that offers farms assistance with these complex nutrient management decisions. Such technology may set the base for specifically managing crops on a minute scale, which creates another potential for environmental benefits while reducing environmental harm (Weersink et al., 2018). For instance, biodiversity and species protection goals can be better integrated into the crop management process, as can agri-environmental policy measures (Schieffer & Dillon, 2015). Such elements require zones without any fertilization (and spraying) at all, and can therefore lead to a new type of farming if they are adopted on a broader scale (Lindblom et al., 2017; Walter et al., 2017). In fact, this may represent innovative, more sustainable land management and has the potential to change landscape configuration, in turn offering biodiversity gains. Therefore, adopting such technologies constitutes a paradigm shift in crop management practices, making the adoption decision complex (Lindblom et al., 2017). Implementing VRNTs represents an innovation behaviour, as opposed to investment in a technical tool to improve current systems and behavioural patterns, which could also be motivated by seeking more environmentally friendly production technology (Barnes et al., 2019a). Contributing to the complexity, rather simple solutions instead of simple implementations of the decision support systems have been offered (Colaço & Bramley, 2018; Rossi et al., 2014). Often based on field trials, yield and nutrient demand predictions might not represent optimal levels based on both the farm economics and environmental view (Bramley et al., 2013).

Following Nyborg et al. (2016), we argue that upscaling may only work if this type of technology-based field management becomes typical; this highlights a need for social norm change. Within this study, we evaluate the role of norms in adoption decisions for more sustainable digital farming techniques. We note that few studies have investigated farmers’ adoption of digital farming techniques from a behavioural perspective. Based on the technology acceptance model (TAM), an adaptation of the RAA for pure technology adoption behaviour (Davis, 1989), studies have identified the causal relations between farmers’ attitudes toward precision farming adoption and respective intention to adopt (Adrian et al., 2005; Flett et al., 2004; Folorunso & Ogunseye, 2008; Rezaei-Moghaddam & Salehi, 2010). With the exception of Aubert et al. (2012), however, these studies lack profound discussions of whether precision farming comprises a new technology or a sustainability-increasing farming technique, or both from the decision perspective. In addition, despite its noted importance in the adoption behaviour of more sustainable farming technologies/techniques, the role of social norms remains underresearched (Dessart et al., 2019).

This gap seems surprising in light of research highlighting the relevance of peer groups in stimulating the spread of decision support tools among farmers (Rose et al., 2016), and the general meaning of social norms in decision-making (Bicchieri, 2008). Furthermore, substantial support exists for the adoption of more sustainable behaviours and resource conservation farming practices. For instance, consumer support rates for animal welfare-labelled meat may depend on others’ decisions (Uehleke & Hüttel, 2019), as do farmers’ decisions on soil management (e.g., Braito et al., 2019; Prager & Curfs, 2016; Price & Leviston, 2014) and water management (Lynne et al., 1995); grassland use (Borges et al., 2014; Lynne et al., 1995); and adoption of agri-environmental schemes (Kuhfuss et al., 2016; Le Coent et al., 2019; van Dijk et al., 2016). In addition, perceptions of social pressure may be relevant for farmers’ land management decisions (Borges & Lansink, 2016; Mills et al., 2017).

Theoretical Approach

In order to explain the intention to adopt digital variable-rate fertilization technologies, we take a psychological perspective and refer to the RAA as the underlying theory. This approach allows us to take into account the role of social norms in decision-making regarding the (potential) adoption of new digital farming technologies. The RAA is rooted in the widely accepted theory of reasoned action (Ajzen & Cote, 2008; Ajzen & Fishbein, 1973; Fishbein, 1967; Fishbein & Ajzen, 1975) in the psychological field (e.g., Armitage & Conner, 2001), focusing on the explanation of individual behaviour. According to the RAA, individual behaviour is based on behavioural intentions, which are influenced by (a) an individual’s attitude toward the behaviour, represented by the opinion the individual has regarding the behaviour; (b) perceived norms, represented by the influence of other individuals; and (c) perceived behavioural control, represented by the possibility that the individual may influence the behaviour (Fishbein & Ajzen, 2010).

Attitude toward the behaviour refers to the individual’s positive or negative feelings about performing that target behaviour. It is determined through an assessment of the individual’s beliefs regarding the characteristics and attributes associated with the behaviour (or the inherent object, respectively). The overall attitude is influenced by the individual consequences of the behaviour, and desirability assessments of these consequences. Perceived norms refer to the individual’s perception of whether other people (who are most important to him or her) think that he or she should perform the behaviour in question. Underlying normative beliefs refer to relevant individuals or groups who support or oppose a given behaviour. Perceived norms can thus be expressed as the sum of the individual perception and motivation assessments for all relevant referents (Ajzen & Fishbein, 1973; Fishbein, 1967; Fishbein & Ajzen, 1975). Perceived behavioural control refers to the individual’s perception of whether he or she is capable of, or has control over, performing the behaviour in question. The underlying control beliefs refer to situational or personal factors that an individual deems important to control regarding the behaviour. The more positive the attitude toward a certain behaviour, the perceived norms and the perceived behavioural control, the more likely it is that the individual will have the intention to perform the behaviour. In summary, performing a certain behaviour entails a process of comparing and selecting among the attitudes, perceived norms and perceived behavioural controls associated with each of the alternative behaviours in the choice set (Sheppard et al., 1988).

Following the postulation of Fishbein and Ajzen (2010) that the RAA has to be adapted to the respective context, we transfer the approach to the context of using digital fertilization methods. The RAA serves as a theoretical model that allows us to explain the predictive validity for the adoption of digital fertilization methods.

First, an individual’s value perception of using VRNT methods is captured by behavioural beliefs. This includes the aspect of increasing productivity, since specific nitrogen application could lead to savings in nitrogen, other inputs and efforts to fulfil documentation obligations. Anticipated productivity gains may lead to an expectation of increasing not only the economic outcome of crop production but also the ecological performance of the farm (Finger et al., 2019). Likewise, the attribute of potential cost reduction can be associated with using a VRNT, as it facilitates nutrient management by taking soil and water heterogeneity into account. Hence, such technology can contribute to avoiding overfertilization and help in meeting the standards of the Nitrogen Directive, thus leading to expectations of reduced monitoring costs (Busse et al., 2014). This also includes aspects of sustainability gains that can be achieved by adopting VRNTs; for instance, by easing the integration of water protection zones in cropland management, or buffer strips without fertilization (Schieffer & Dillon, 2015).

These beliefs then lead to an individual’s attitude (positive or negative feelings) toward VRNTs. Positive feelings in our case refer to whether individuals feel the method is simple to use, are satisfied with its use (as shown to be relevant in other contexts combining technological and methodological aspects; see, e.g., Han et al., 2004) and are satisfied with the new, more sustainable fertilization concept (Kuhfuss et al., 2016). An individual having a more positive attitude is then expected to have a higher intention to use digital fertilization methods (Fishbein & Ajzen, 2010; Sheppard et al., 1988). This results in two hypotheses:

Hypothesis 1: The higher the behavioural beliefs regarding digital fertilization methods, the more positive the attitude of an individual regarding the methods.

Hypothesis 2: The more positive the attitude of an individual regarding digital fertilization methods, the higher the intention to use these methods.

Second, consideration of the opinion of other relevant individuals is captured by normative beliefs. Relevant individuals stem from the professional context, such as colleagues in similar functions (Fishbein & Ajzen, 1975). In the case of VRNTs, such reference persons for farmers are crop advisors, neighbouring farmers, domain experts and other users of digital fertilization methods (Aubert et al., 2012; Marra et al., 2003; McBride & Daberkow, 2003). The normative beliefs relating to these reference groups then result in perceived pressure or motivation to use VRNTs (subjective norms). This reflects whether an individual thinks that these reference persons support or urge the usage of variable-rate fertilization methods. The normative influence regarding VRNTs is rooted in farmers seeking exchange with others with the aim of gaining experience and opinions, since the introduction of VRNTs requires major investments and fundamental changes in the way fertilization is conducted (Kuhfuss et al., 2016; Rose et al., 2016). According to the RAA, the more positively this support is perceived in relation to norms, the higher the intention to use VRNTs. This is reflected by the next set of hypotheses:

Hypothesis 3: The higher the normative beliefs regarding digital fertilization technology, the more positive the perceived norms of an individual regarding the methods.

Hypothesis 4: The more positive the perceived norms of an individual regarding digital fertilization technologies, the higher the intention to use these methods.

Third, factors that impede or facilitate the likelihood of an individual using digital fertilization methods are captured by control beliefs. In the context of precision and variable-rate farming technologies, such factors are easy access to methods (e.g., Hudson & Hite, 2003), and the costs of using the methods (Marra et al., 2010). Both are important factors in whether the individual belief is that one is able to adopt the method. The control beliefs then lead to perceived behavioural control, which reflects whether an individual perceives that he or she has the new technology under control. That is, when easy access and cost-effectiveness are provided, variable-rate fertilization methods will likely be taken up as farmers need to be able to establish such a shift in their fertilization management procedures (Lindblom et al., 2017). Finally, higher perceived behavioural control is positively related to intention to use (Fishbein & Ajzen, 2010; Sheppard et al., 1988). Hence, the penultimate set of hypotheses is derived:

Hypothesis 5: The higher the control beliefs regarding digital fertilization methods, the more positive the perceived behavioural control of an individual regarding the methods.

Hypothesis 6: The more positive the perceived behavioural control of an individual regarding digital fertilization methods, the higher the intention to use these methods.

Finally, our analysis focuses on actual behaviour to close the typical intention-behaviour gap. This gap also holds true for digital fertilization methods: intentions do not always lead to behaviour, while there is a general positive relationship between intention and behaviour (Fishbein & Ajzen, 2010). Intention to use and usage of site-specific digital fertilization methods follow a temporal pattern; that is, it is not a mutually exclusive, single yes/no decision, but rather the frequency of usage is relevant (true adoption). This is represented by the last hypothesis:

Hypothesis 7: The higher the intention to use digital fertilization methods, the higher the actual use of these methods.

We summarize the research model in Figure 1, including the hypotheses.

Figure 1.

Research model.

Material and Method

Questionnaire

Within the research model (see Figure 1), we base all measures on the RAA, using 7-point Likert-type scales for each item (Fishbein & Ajzen, 2010). The questions were developed following Fishbein and Ajzen (2010); however, the model provides the causal foundation and the questionnaire requires contextual adjustments. In making these adjustments, we stayed close to the original framework and followed the idea of Fishbein and Ajzen (2010) that “it is important to realize that there is no single reasoned-action questionnaire. Each investigation requires construction of a suitable questionnaire” (p. 456). We therefore incorporated the aforementioned dimensions relevant for hypothesis development as items for the context of site-specific digital fertilization methods.

The main aspects of behavioural beliefs are productivity and costs, as well as easy access and cost-effectiveness regarding control beliefs as outlined in Section 2. Regarding normative beliefs, we refer to reference groups of crop advisors, neighbouring farmers, domain experts, and other users of digital fertilization methods. We chose not to include the identification items from the sample questionnaire proposed by Fishbein and Ajzen (2010) as this was targeted to individuals in a more private context, such as exercising or smoking. The identification with others in this context seems straightforward, as others were referred to as other individuals who could be grouped into a reference unit. In our case, moving the context to individuals in a specific business context, others can also be people who work in an area that is related yet different to our respondents’ focus. We were confronted with a mix of other farmers who are similar and experts to whom the items do not apply. Since we would have measured the items for some of the reference individuals and not others we would have created inconsistencies in the analysis, and thus left these items out.

Each belief is measured on a scale and multiplied by the indicated weighting of its relative importance (Fishbein & Ajzen, 2010). This procedure is stated representatively for beliefs in the normative dimension: “Normative belief strength and motivation to comply were multiplied, and the resulting products were correlated with reported mountain climbing frequency” (p. 209).

Regarding attitude, we focus on items pertaining to simplicity and satisfaction, which can be seen as judgemental items in the sense of the original questionnaire. Our items of perceived norms and perceived behavioural control (using the direct reflective form) follow the sample survey closely. Intentions to use VRNT with machine support and current usage behaviour are both measured with their frequency of occurrence.

For brevity, we present full details of the questionnaire in the appendix Table A1. In addition, we gathered the control variables of farmers’ gender, age and farm size, which are included as controls for every variable in the model.

Sample

We conducted a survey among farmers in Germany to analyse the influence of antecedents from a psychological perspective on the adoption of site-specific nitrogen fertilization methods (Fall, 2018). These VRNTs may assist in meeting thresholds according to the EU’s Nitrate Directive, which aims to ensure minimum standards are met to preserve the environment (Barnes et al., 2011; Busse et al., 2014), and offer cost savings due to nutrient management on a finer scale, including fuel savings even in regions with restrictions in nitrogen application levels (Bongiovanni & Lowenberg-Deboer, 2004). Against the backdrop of the German debate at the time, the survey was carried out with reference to the renewal of the Nitrate Directive. We believe that farmers are aware of fertilization issues, including alternative fertilization techniques. Thus, we treat potential bias regarding the understanding of the topic by respondents as comparably low but note that we cannot fully rule out social desirability bias in the answers, though this bias is expected to be much lower compared with other crop production inputs, such as pesticides.

The target group comprises current users and nonusers among German nonorganic farmers, representing a total of 266,700 farms (BMEL, 2017). We gathered a convenience sample based on personal contacts, advertising the questionnaire in relevant domain-specific journals, mailing lists from associations, and online platforms. We note the potential selection bias due to digital media preferences and interest in the topics of nitrogen fertilization, machine support, and precision farming. Another potential bias may arise from local preferences; for instance, because of the authors’ affiliations, and sending out the questionnaire such that individuals from western Germany might have been more attracted to respond.

In total, 330 participants clicked on the link registered by the online system containing the survey. Of these participants, 186 started to answer the survey, with 109 finishing it. Eleven participants from this set had to be removed because there were too many missing values regarding relevant variables. This resulted in a final set of 98 participants; for a probability level of about 0.05, we note a survey error of about 10%. The survey consisted of 65 participants who have never used VRNTs and 33 users who have used site-specific digital fertilization before; the survey error for this fraction is about 7.87%. The majority are male (91.8%; 6.1% are female and two participants did not indicate their gender). The convenience sample is not representative of regional distribution, age, education and operating size; based on the logins, we can infer that a large fraction of participants responded from a western German location (approx. 80% and approx. 70% from Lower Saxony and North-Rhine Westphalia, respectively, though this does not necessarily mean that the farms are located in this region). The remaining participants responded from locations in eastern Germany and neighbouring regions (e.g., Austria). Another potential explanation for why farmers from western Germany were more attracted to the survey is that very large farms, typically organized as enterprises (rather than family farms), are likely to be located in Northeastern Germany.² The use of machine support offers obvious financial advantages, and in this region generally higher adoption rates are reported (Giesler, 2018), which may have made the survey less attractive to these individuals.

The average age of the participants is 41.17 years, with a range between 21 and 75 years, where 62.8% of survey participants are younger than 45 years, and 19.1% older than 55 years. According to the agricultural census 2010, 32% of farmers are younger than 45 years and 31% are older than 55 years (BMEL, 2017). Participants operate 400.73 hectares on average, ranging between 20 and 6,500 hectares. While the range is representative for Germany, the average size is clearly above the overall German average of 61 hectares, whereas the eastern German average was about 259 hectares in 2016 (BMEL, 2017).

Despite the lack of representativeness of all farms in Germany, we believe that our convenience sample represents an interesting group. Biased to highly educated younger farmers who are used to online platforms and digital methods, our sample represents farmers who show innovation-seeking behaviour in the direction of machine-supported crop management (Tamirat et al., 2018)—a group that will probably continue operating in the near future and is likely to be familiar with the agricultural production structure of Northwestern Germany, where nitrate leaching problems are most severe.

Statistical Tests

In order to analyse our research model, where we focus on contributing to the explanation of the variable “Use of digital fertilization methods,” we use the partial least squares (PLS) method. As a measure for this contribution, R², as a construct-based criterion, is of core interest (Petter, 2018); hence, we chose PLS-SEM (structural equation modelling) over covariance-based SEM (Hair et al., 2011). Against the backdrop of the small sample, we apply a bootstrapping procedure with 5,000 resamples, implemented via SmartPLS 3.2.8 (Hair et al., 2011).

Before conducting the analysis, we tested our model regarding validity and reliability using the following procedure and thresholds for reflective and formative measurement models, as described by Hair et al. (2011), Hulland (1999), and Cenfetelli and Bassellier (2009). First, in a data preparation step, missing values for six variables from 15 participants regarding normative beliefs and social norms were replaced following a typical mean value imputation procedure (Little & Rubin, 2019). Second, we tested our reflective variables “Attitude,” “Perceived norms,” and “Perceived behavioural control.” The reliability of the three variables using the concept of composite and indicator reliability shows that all values regarding composite reliability are above the threshold of 0.7 (Attitude: 0.910; Perceived norms: 0.928; Perceived behavioural control: 0.902) and indicator reliability is fulfilled, since all indicator loadings (see the appendix Table A2) are above the threshold of 0.7 (Hair et al., 2011). Convergent validity is confirmed by average variance extracted (Attitude: 0.835; Perceived norms: 0.764; Perceived behavioural control: 0.821), for which all values are above the threshold of 0.5 (Hair et al., 2011). Furthermore, we tested discriminant validity by applying the heterotrait–monotrait ratio of correlation (Henseler et al., 2015). This criterion provides higher accuracy, in terms of detecting discriminant validity, compared with the frequently used Fornell–Larcker criterion. The values are all below the threshold of 0.9 (Henseler et al., 2015; see the appendix Table A3). In addition, we conducted a confirmatory factor analysis for the three reflective constructs, which gives support to the assignment of indicators to the variables.

Third, we tested the formative variables in our model. We checked multicollinearity among the indicators with the variance inflation factor, which should be below 5 for each indicator Hair et al. (2011); that is, 3.33 using a stricter threshold (Diamantopoulos & Siguaw, 2006). All values are well below the strict threshold (see the appendix Table A4). Furthermore, the relative and absolute importance of indicators was tested with loadings and weights; we found that the loading and weight of normative beliefs regarding crop advisors are both nonsignificant. Hence, we treat the construct and model each influencing group separately, following the recommendation of Cenfetelli and Bassellier (2009). Our model thus does not incorporate any indicator for which weight and loading are nonsignificant (see the appendix Table A5). Heterogeneity among indicators is tested by checking whether the bivariate correlations are higher between an indicator and the variable than between indicators (see the appendix Table A6). The results in this regard show that no suppressors and/or no collinear indicators can be identified.

Fourth, we conducted several tests to examine the quality of our structural model. We examined the standardized root mean square residual as a measure for the approximate fit of our model (Henseler et al., 2016). Our model reached .052, which is below the maximum threshold of .8. In addition, we performed a blindfolding procedure involving an omission distance of 6 to assess the predictive relevance of the model. The test revealed positive Stone–Geisser Q² values for each variable (Attitude: .036; Perceived norms: .329; Usage: .130; Intention to use: .142; Perceived behavioural control: .072), indicating a strong overall predictive power for the model (Henseler et al., 2016).

Fifth, because our study relied on responses provided by respondents at the same time (Campbell & Fiske, 1959), common method bias might have occurred. To reduce this bias in advance, we followed Kortmann’s (2015) procedure regarding anonymity, confidentiality, the placement of dependent and independent variables, and the use of different types of scales. Also drawing on Kortmann (2015), two post hoc tests were conducted for our data to assess common method bias. In a first step, we employed Harman’s (1967) single-factor test. The results show that the first factor only accounted for 27.93% (resp. 16.98% using varimax rotation) of the total variance, while 6 factors with eigenvalues greater than 1.0 accounted for 67.15% of the variance. Second, we addressed common method bias (Podsakoff et al., 2003). Our results reveal that, on average, the constructs explain 70.11% of the variance in our sample. In contrast, the method factor explains on average 5.10% of the variance, which results in a ratio of substantive variance to method variance of 13.75. Additionally, the majority of the method factor loadings are insignificant. The maximum loading of one indicator was −.263; however, this was assigned to a different substantive factor with .555. Based on our results, we conclude that common method bias is either absent or negligibly low (Kortmann, 2015).

Results

Results Regarding the Research Model

The descriptives, as well as the correlations, regarding the variables of our research model enable some high correlations to be identified (see the appendix Table A7), which is likely to occur despite the components being conceptually different (Fishbein & Ajzen, 2010). As demonstrated in Section 4, the quality of the research model is not affected according to the criteria tested. The results of our analysis regarding the research model are depicted in Figure 2.

Figure 2.

Results of the research model.

We find empirical evidence for Hypothesis 1 (β = .195*; f² = 0.041⁺), Hypothesis 3 (advisors: −0.204*; f² = 0.042⁺; experts: 0.233**; f² = 0.060⁺; farmers: 0.314***; f² = 0.131⁺; users: 0.443***; f² = 0.042*) and Hypothesis 5 (0.298***; f² = 0.099^ns), showing that the beliefs are relevant antecedents. The results show that an increase in behavioural beliefs leads to a higher positive attitude, but behavioural beliefs explain only 8% of its variance. Control beliefs have a stronger influence, explaining 14% of the variance of perceived behavioural control. In relation to the two other beliefs, normative beliefs have the highest influence and explain 50% of the variance of perceived norms.

Regarding Hypothesis 2 (0.088^ns; f² = 0.007^ns), Hypothesis 4 (0.423^ns; f² = 0.176*) and Hypothesis 6 (0.007^ns; f² = 0.000^ns), we find perceived norms to be the only predictor for intention to use (Hypothesis 4). Attitude and perceived behavioural control do not have a significant influence on the intention to use digital fertilization methods. Intention to use is thus a strong predictor for use (Hypothesis 7; 0.407***; f² = 0.199*), though we find a remaining gap between intention and behaviour.

Regarding the control variables of farmer’s gender, age, and farm size, we find the strongest and most significant negative relationships for gender (female = 1) on perceived behavioural control (−0.235, p < .01) and for usage behaviour (−0.117, p < .05). The effect of gender on perceived behavioural control is, however, negligible for our sample. We find that the intention-behaviour gap is larger for male farmers than for female farmers. Female farmers seem to be slightly more confident in transferring their intention to behaviour in this context. Neither farm size nor farmer’s age can be approved as a predictor for adoption behaviour for our sample. We find, however, farm size to be a strong, though negatively related, predictor of control beliefs (−0.189, p < .01), potentially through anticipated higher learning and investment costs. Large farm size often means a higher number of employees, who will be the end-users of VRNTs after adoption. It is likely that organizational actions associated with higher effort have to take place to train and convince employees to follow the machine recommendations; this is an important aspect, as emphasized by Lindblom et al. (2017). These findings contradict those of Kutter et al. (2011), who argued that larger farms may benefit from economies of scale, and that higher adoption rates can be estimated for larger farms (Finger et al., 2019). We note for our study that larger farms are overrepresented for the region, such that organizational burdens may be present, and economies of scale may be encountered by prevailing fragmented land use compared to other regions, where large farming structure goes along with large-scale field size.

Post Hoc Test Results Regarding Social Norms

We find normative beliefs to be distinct significant constructs. Extending the analysis of the general model via the specific indirect effects (each belief via perceived norms and via intention of usage) of each reference group within normative beliefs on usage reveals the following: crop advisors have an indirect effect of −0.035 (p < .05), experts of 0.040 (p < .05), local farmers of 0.054 (p < .05) and users of 0.077 (p < .05). Taking the mean of the respective items (without the weight used to identify what the reference groups recommend), we note that participants are indifferent regarding crop advisors’ recommendations to adopt VRNTs (M = 3.5). Likewise, respondents primarily receive recommendations for digital site-specific fertilization not from local farmers (M = 2.69), but rather from experts and other experienced users (M = 4.53 and 4.45, respectively). Generally, other farmers, as users with experience, are taken most seriously by the respondents (M = 5.29), followed by crop advisors (M = 5.13). Neighbouring farmers are given much consideration (M = 3.80), while experts are positioned in the middle (M = 4.53). In summary, the impact of peer groups with experience, who recommend using site-specific digital fertilization methods, on perceived norms is the highest, and this group is thus considered during decision making on adoption.

Discussion

The results show that social norms defined for different reference groups are the single predictor for the intention to adopt VRNTs with machine support that offer more sustainable crop management. While attitude driven by behavioural beliefs is often seen as the major predictor for intentions in general (Fishbein & Ajzen, 2010), for precision agriculture (e.g., Adrian et al., 2005) and for sustainable practices (Yeboah et al., 2015), we find attitude to be a nonsignificant predictor. This could be explained by the different decision set up faced by farmers when upgrading from machine guidance (precision farming) to VRNTs with machine support (Barnes et al., 2019a). Likewise, our results show perceived behavioural control—a known predictor in technology acceptance for system improvements (e.g., Folorunso & Ogunseye, 2008)—to be an irrelevant predictor. In fact, the new technology is not only an improvement on the old system but also offers a new crop management approach at scale, taking various sources of heterogeneity (e.g., soil, water) into account, thereby offering environmental benefits. In the context of such complex decision making, analytically related reasoning might not be possible, sufficient or helpful to lead to an opinion about the adoption of VRNT methods. Rather, social exchange with users, farmers from the region and experts is important, and these represent the most relevant reference groups for the decision to implement the new crop management technique. This result offers an explanation for the argumentation on entry barriers in the form of high investments in technical understanding and organizational burdens for precision farming techniques (e.g., Kutter et al., 2011). Likewise, often reported regional clusters of organic farming adopters, which also constitutes a different production scheme (e.g., Läpple & Kelley, 2015) suggest that exchange with others can be essential. The often reported reluctance to adopt because of a fear of technological lock-in under policy uncertainty for the livestock sector (Floridi et al., 2013) also demonstrates the complexity of adoption decisions for technologies at the fringe of a technical innovation and more sustainable practices. The difference in crop management associated with VRNT under machine support compared with the current system has even been termed a paradigm shift (Aubert et al., 2012; Lindblom et al., 2017), and could therefore be too abstract for mental processing by farmers. This calls for a more tangible representation of VRNTs’ nature and benefits, in order to realize individual advantages, by means of users with experience demonstrating practical experiences to farmers. Tangible elements in this regard include networking events, field days, presentations and provision of proofs of concept from users for users. Politics at different levels could support such events, for instance, within locally adjusted measures of the second pillar of the EU’s Common Agricultural Policy, and thereby use the opportunity to clearly communicate signals on societal expectations (Mills et al., 2017). Organized interactions between farmers have also been emphasized to engage farmers more in sustainable land management; for instance, cooperatives for upscaling nonsubsidized agri-environmental measures such as hedges (van Dijk et al., 2016). In this regard, subsidized but collective outcome-oriented agri-environmental schemes have been proposed (Burton & Schwarz, 2013). By creating common goals and increasing cooperation, Burton and Schwarz expect ecological improvements in farming systems. Given the normative influence shown in this study, it seems promising to foster the necessary social exchange to improve production systems regarding sustainability.

In the debate on adoption barriers for precision farming as a pure technology innovation, previous studies have tended to call for simpler technical solutions (Paustian & Theuvsen, 2017), benchmarking (Finger et al., 2019), or guidelines and calculators (Zarco-Tejada et al., 2014) to make farming systems comparable and thereby ease adoption. Our results thus enhance knowledge of how to best demonstrate advantages to overcome barriers. Thereby, we add to the results of Barnes et al. (2019a), who discussed exchange between farmers and stakeholders as a potential determinant of uptake. Our results further offer a theory-guided explanation for the results outlined by Tamirat et al. (2018), who found higher adoption rates among farmers who participated in workshops. Stressing the importance of the learning process, our results also augment those of Lindblom et al. (2017).

In sum, the model in which social norms are embedded explains almost 20% of the variance in usage behaviour. Theory-based evidence on the role of social norms helps in advancing the discourse on entry barriers, which remain prevalent despite the many environmental advantages of VRNTs (Finger et al., 2019; Walter et al., 2017). Given the complexity of the decision-making context, it seems that typical time constraints of farmers lead to using reference groups and, if these are trusted, following their behaviour. Farmers are not limited to their own considerations, which are more reflected in the attitude and perceived behavioural control driving the (non)adoption. Our results thus add to the discussion started by Aubert et al. (2012) on how IT-based solutions can serve as an enabler of more sustainable land management, arguing that the complexity of crop management decision making must be acknowledged.

Our results have implications for achieving a more sustainable agri-food system in Germany, and add to the debate on how to engage farmers in sustainable land management (e.g., Mills et al., 2017). The improved understanding of farmers’ behaviour regarding the adoption of sustainable technologies can represent a first step based on the fact that digital farming technologies may be perceived as a pure technology but also as a technology that enables more sustainable farming (Walter et al., 2017). We thus argue that it makes sense to foster uptake because VRNTs represent a key technology for the transformation process of crop management to a finer scale, triggering environmental benefits. According to diffusion theory, the spread of technologies can be initiated by a reference group (pioneers) from which a local cluster may emerge (E. M. Rogers, 2003). Relying on our sample, which likely overrepresents members of innovation peer groups, we suggest supporting the emergence of local clusters by means of events that offer visualization and exchange with experienced users. Once this group begins to demonstrate performance, the benefits of this new technology will be clearer, and the faster and the more widely the technology will spread (Paluck et al., 2016). Realizing the individual benefits is a precondition for adoption, where, for our sample, we find social reference groups to be the most important cohort—indeed, this is a known factor in the innovation diffusion context (cf. Nyborg et al., 2016) and in the adoption of sustainable farming practices (cf. Dessart et al., 2019). For Germany, we specifically argue that VRNTs have the potential to immediately reduce environmental harm in regions with problematic nitrate leaching, even while also offering economic benefits (Bongiovanni & Lowenberg-Deboer, 2004). Given that nitrate leaching constitutes a regional problem, mainly in western Germany, which is better represented by our sample, our findings suggest starting with demonstration events in these regions. As noted by Lindblom et al. (2017), crop managers may fail to fully exploit the potential of input savings by insufficiently following machine recommendations. In this regard, demonstration of the new system by actual users can help mitigate such behaviours and establish the new crop management as the norm.

Following the idea that digital fertilization methods may be perceived as more environmentally friendly farming practices, this attribute could be a prominent exploratory factor with regard to attitude. However, even a robustness check, where we reduced attitude to “environmentally friendly” did not change the results. This result seems somehow contrary to the implications reported in Finger et al. (2019), who suggested highlighting environmental consequences under precision farming techniques to change adoption behaviour. One explanation for this could be that, in our sample, the respective environmental contribution by technology choice is not part of the decision because farmers may not be fully aware of the environmental benefits and, as Barnes et al. (2019a) argued, precision farming may represent a technology rather than an environmentally friendly farming system. Addressing this issue could be a task for sector initiatives to improve knowledge and understanding regarding environmental effects (Kollmuss & Agyeman, 2002); however, this should be supported by institutional mechanisms to demonstrate important factors (Smith et al., 2018) and to foster management of relationships with the natural environment (Winn & Pogutz, 2013). Another explanation for the contrasting results could be that the sampled farmers simply do not show pro-environmental behaviour.

Conclusions

Our study is motivated by analysing how adopting IT-based precision farming methods can facilitate farmers’ use of sustainable farming practices. The results show a positive impact of social norms of different reference groups on usage behaviour regarding site-specific digital fertilization methods. Our findings contribute to understanding of the organizational actions of farmers, and hence the respective impact on the environment thereof, in multiple ways.

First, our results show that in the context of a complex decision-making problem, in our case demonstrated by site-specific digital fertilization management, farmers tend to partly outsource decision making regarding (non)adoption to other reference groups. We conclude that farmers’ decision on whether to adopt a technology-driven method is likely to be oriented toward reference groups with experience. In order to foster significant uptake, farmers who are able and want to tackle the complex decision problem pertaining to nitrogen (innovation leaders) should be identified and supported so that they can serve as a reference for others. This may help facilitate the emergence of local clusters, since the majority of farmers seem to base such decisions on social reference groups. This will require selective development funds instead of a scattergun approach, but also actively promoting these farmers as role models to a large number of peers.

Second, our results pertain to the adoption of a technology-driven method; that is, a combination of technique and technology that increases sustainable agricultural land use. Such variable-rate farming methods with machine support aim to optimize nutrient applications on a scale that takes various sources of heterogeneity into account. Managing the field on a finer scale minimizes not only the input but also the environmental impact, thereby generating sustainability gains. Based on our results, these adoption decisions, which have a considerable impact on farm organizational and environmental figures, are not based on pure calculations, but rather on social reference groups. Such groups contribute to forming the opinions of decision makers. This reveals the importance of better linking social reference groups with farmers; that is, lowering barriers to information and communication, for instance, via networking events.

Third, we argue that local, contextual solutions are initially needed to establish local technology clusters; in turn, these will serve as a base for upscaling sustainable farming practices. Fostering further uptake requires social norm changes (cf. Nyborg et al., 2016); respective private and public policies should therefore trigger such changes to achieve overall more sustainable land use.

Fourth, we link the adoption of digital farming methods with increasing sustainability. Our study brings the discussion on adoption barriers to the fore, and offers implications for advancing sustainable agri-food systems in Germany. Our results are theory based, whereas many other studies have based their implications on pure data work, limiting causal interpretation. We see evidence-based policy advice as a major advantage of our study, despite the limitations of our sample’s representativeness. Our implications ultimately pertain to achieving a large uptake of variable-rate nitrogen fertilization, based on which the adoption of precision farming techniques could emerge with overall environmental benefits in several directions. We see VRNTs as one key technology in precision agriculture that serves as a starting point for site-specific crop management. It offers environmental benefits at the same profitability compared with traditional fertilization methods, even in regions with restrictions on nitrogen application levels (Bongiovanni & Lowenberg-Deboer, 2004), or even increases in profitability due to savings in other inputs, such as fuel (Balafoutis et al., 2017). Given that nitrate leaching constitutes a regional problem in Germany, and has triggered regulation intensification, which in turn is perceived as an additional burden for farmers with regard to meeting these standards (recently even giving rise to protests on the street), machine support technology can help farmers in fulfilling these standards (Busse et al., 2014; Link et al., 2006). Despite these obvious advantages, adoption rates are low, showing the need to extend the discussion on adoption barriers.

Regarding the practical implications of our findings on the promotion of digital fertilization methods, farmers should be brought into contact with users who have experience in order to enhance such promotion. These users hold the most credibility for farmers, and also promote usage for others, hence influencing perceived norms more positively compared with other groups. Thus, key farmers should be selected who are in favour of using digital fertilization methods, and supported to share their experiences with other farmers. Thereby, crop advisors could play a role as network moderators to bring farmers together. In turn, crop advisors would benefit from these events to prepare for their role in continuously improving the new system. Politics at different levels could support such events, for instance within locally adjusted measures of the second pillar of the EU’s Common Agricultural Policy, and thereby use the opportunity to clearly communicate signals on societal norms and expectations (Mills et al., 2017).

This study comes with limitations to be addressed in future research. First, selection bias due to preferences for digital methods may exist. Second, bias may be inherent due to the unknown social relationships of the farmers in our sample. A snowball principle could be used in future studies, and the framework could be enhanced, to explicitly address the effects of social networks. Third, our study is limited to the use of digital fertilization. It is unclear whether farmers decide similarly for other kinds of variable-rate methods with machine support. Fourth, we argue that, once adopted, farms will follow the decision-support system’s recommendation on site-specific nitrogen quantity, though we do not have proof of this. Future studies should address this gap directly. Fifth, we focus on factors derived from the RAA and define social reference groups rather narrowly. This leaves room to tackle the societal dimension of perceived norms and take into account the societal pressure to improve the sustainability of land management that farmers increasingly face with current production systems, which likely influences adoption behaviour (Carroll & Groarke, 2019). Given that land and its use contribute to public goods such as landscape, eco-system services and groundwater, free-rider behaviour might be another issue. Therefore, one direction for research could be to include cognitive processes related to the mental models of farming systems among farmers, which are assumed to play an important role in the adoption of variable-rate farming methods. These issues could help identify factors that better explain the gap between intention and behaviour.

Footnotes

Appendix

Table A7.

Descriptive Statistics of the Overall Sample and Correlations Among Variables.

	M	SD	1	2	3	4	5	6	7	8	9	10	11
1. Use of digital fertilization methods	1.05	1.70	—	.41***	.17	−.05	.22*	.17	−.07	.13	.09	.17	−.07
2. Intention to use digital fertilization methods	3.76	1.88		—	.28**	.06	.47***	.26*	.32**	.23*	.16	.08	.16
3. Attitude	3.86	1.44			—	.27*	.46***	.41***	.33**	.36***	.49***	.14	.44***
4. Behavioural beliefs	26.46	10.93				—	.36***	.28**	.26**	.39***	.30**	.19	.41***
5. Perceived norm	3.84	1.73					—	.35***	.51***	.56***	.48***	.18	.44***
6. Normative beliefs “Crop advisors“	18.82	12.18						—	.50***	.47***	.53***	.17	.37***
7. Normative beliefs “Experts”	10.95	9.62							—	.44***	.39***	−.05	.39***
8. Normative beliefs “Farmers from the region“	21.10	11.29								—	.53***	.20*	.24*
9. Normative beliefs “Users with experience“	27.96	15.75									—	.29**	.37***
10. Perceived behavioural control	5.74	1.55										—	.27**
11. Control beliefs	20.48	11.15											—

Note. N = 98.

p < .05. **p < .01. ***p < .001 (two-tailed tests).

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Michael Leyer

Notes

Author Biographies

Silke Hüttel is full professor and leads the Production Economics Group at University of Bonn, Faculty of Agriculture. She is an agricultural economist with a quantitative methodological focus. Her research interests are in understanding farmers’ decision-making under sustainability aspects including animal welfare, efficiency and bidding behaviour in farmland markets.

Marie-Therese Leuchten was a master student at University of Bonn and associated to the Production Economics Group.

Michael Leyer is a full professor having the chair of service management at the University of Rostock as well as being adjunct professor at the Queensland University of Technology. His work is focussed on understanding decision makers’ behaviour of individuals in organisational contexts supported by digital environments.

References

Adrian

A. M.

Norwood

S. H.

Mask

P. L.

(2005). Producers’ perceptions and attitudes toward precision agriculture technologies. Computers and Electronics in Agriculture, 48(3), 256-271. https://doi.org/10.1016/j.compag.2005.04.004

Ajzen

Cote

N. G.

(2008). Attitudes and the prediction of behavior. In Frontiers of social psychology: Attitudes and attitude change (pp. 289-311). Psychology Press.

Ajzen

Fishbein

(1973). Attitudinal and normative variables as predictors of specific behavior. Journal of Personality and Social Psychology, 27(1), 41-57. https://doi.org/10.1037/h0034440

Arbuckle

J. G.

Morton

L. W.

Hobbs

(2015). Understanding farmer perspectives on climate change adaptation and mitigation: The roles of trust in sources of climate information, climate change beliefs, and perceived risk. Environment and Behavior, 47(2), 205-234. https://doi.org/10.1177/0013916513503832

Armitage

C. J.

Conner

(2001). Efficacy of the theory of planned behaviour: A meta-analytic review. British Journal of Social Psychology, 40(4), 471-499. https://doi.org/10.1348/014466601164939

Aubert

B. A.

Schroeder

Grimaudo

(2012). IT as enabler of sustainable farming: An empirical analysis of farmers’ adoption decision of precision agriculture technology. Decision Support Systems, 54(1), 510-520. https://doi.org/10.1016/j.dss.2012.07.002

Balafoutis

Beck

Fountas

Vangeyte

Wal

Soto

Gómez-Barbero

Barnes

Eory

(2017). Precision agriculture technologies positively contributing to GHG emissions mitigation, farm productivity and economics. Sustainability, 9(8), Article 1339. https://doi.org/10.3390/su9081339

Barnes

A. P.

Soto

Eory

Beck

Balafoutis

Sánchez

Vangeyte

Fountas

van der Wal

Gómez-Barbero

(2019a). Exploring the adoption of precision agricultural technologies: A cross regional study of EU farmers. Land Use Policy, 80(January), 163-174. https://doi.org/10.1016/j.landusepol.2018.10.004

Barnes

A. P.

Soto

Eory

Beck

Balafoutis

Sánchez

Vangeyte

Foundas

van der Wal

Gómez-Barbero

(2019b). Influencing factors and incentives on the intention to adopt precision agricultural technologies within arable farming systems. Environmental Science & Policy, 93(March), 66-74. https://doi.org/10.1016/j.envsci.2018.12.014

10.

Barnes

A. P.

Willock

Toma

Hall

(2011). Utilising a farmer typology to understand farmer behaviour towards water quality management: Nitrate Vulnerable Zones in Scotland. Journal of Environmental Planning and Management, 54(4), 477-494. https://doi.org/10.1080/09640568.2010.515880

11.

Bartkowski

Bartke

(2018). Leverage points for governing agricultural soils: A review of empirical studies of European farmers’ decision-making. Sustainability, 10(9), Article 3179. https://doi.org/10.3390/su10093179

12.

Bicchieri

(2008). The fragility of fairness: An experimental investigation on the conditional status of pro-social norms. Philosophical Issues, 18(1), 229-248. https://doi.org/10.1111/j.1533-6077.2008.00146.x

13.

BMEL. (2017). Daten und Fakten: Land-, Forst- und Ernährungswirtschaft mit Fischerei und Wein- und Gartenbau [Facts and figures: Agriculture, forestry and food industry with fishing and viticulture and horticulture]. https://www.bmel.de/SharedDocs/Downloads/DE/Broschueren/Daten-und-Fakten-Landwirtschaft.html

14.

Bongiovanni

Lowenberg-Deboer

(2004). Precision agriculture and sustainability. Precision Agriculture, 5(4), 359-387. https://doi.org/10.1023/B:PRAG.0000040806.39604.aa

15.

Borges

J. A.

Lansink

A. O.

(2016). Identifying psychological factors that determine cattle farmers’ intention to use improved natural grassland. Journal of Environmental Psychology, 45(March), 89-96. https://doi.org/10.1016/j.jenvp.2015.12.001

16.

Borges

J. A.

Lansink

A. O.

Ribeiro

C. M.

Lutke

(2014). Understanding farmers’ intention to adopt improved natural grassland using the theory of planned behavior. Livestock Science, 169(November), 163-174. https://doi.org/10.1016/j.livsci.2014.09.014

17.

Braito

Flint

Leonhardt

Penker

(2019). Mental models enlighten assumptions about farmers, land tenure and soil management. https://ideas.repec.org/p/ags/eaa165/288288.html

18.

Bramley

R. G. V.

Lawes

Cook

S. E.

(2013). Spatially distributed experimentation: Tools for the optimization of targeted management. In Oliver

(Ed.), Precision agriculture for sustainability and environmental protection. Routledge. https://doi.org/10.4324/9780203128329-22

19.

Burton

R. J. F.

(2004). Reconceptualising the “behavioural approach” in agricultural studies: A socio-psychological perspective. Journal of Rural Studies, 20(3), 359-371. https://doi.org/10.1016/j.jrurstud.2003.12.001

20.

Burton

R. J. F.

(2014). The influence of farmer demographic characteristics on environmental behaviour: A review. Journal of Environmental Management, 135, 19-26. https://doi.org/10.1016/j.jenvman.2013.12.005

21.

Burton

R. J. F.

Schwarz

(2013). Result-oriented agri-environmental schemes in Europe and their potential for promoting behavioural change. Land Use Policy, 30(1), 628-641. https://doi.org/10.1016/j.landusepol.2012.05.002

22.

Busse

Doernberg

Siebert

Kuntosch

Schwerdtner

König

Bokelmann

(2014). Innovation mechanisms in German precision farming. Precision Agriculture, 15(4), 403-426. https://doi.org/10.1007/s11119-013-9337-2

23.

Campbell

D. T.

Fiske

D. W.

(1959). Convergent and discriminant validation by the multitrait-multimethod matrix. Psychological Bulletin, 56(2), 81-105. https://doi.org/10.1037/h0046016

24.

Carroll

G. A.

Groarke

J. M.

(2019). The importance of the social sciences in reducing tail biting prevalence in pigs. Animals: An Open Access Journal From MDPI, 9(9). https://doi.org/10.3390/ani9090591

25.

Cenfetelli

R. T.

Bassellier

(2009). Interpretation of Formative measurement in information systems research. MIS Quarterly, 33(4), 689-707. https://doi.org/10.2307/20650323

26.

Chabé-Ferret

Le Coent

Reynaud

Subervie

Lepercq

(2019). Can we nudge farmers into saving water? Evidence from a randomised experiment. European Review of Agricultural Economics, 46(3), 393-416. https://doi.org/10.1093/erae/jbz022

27.

Colaço

A. F.

Bramley

R. G. V.

(2018). Do crop sensors promote improved nitrogen management in grain crops? Field Crops Research, 218(April), 126-140. https://doi.org/10.1016/j.fcr.2018.01.007

28.

D’Antoni

J. M.

Mishra

A. K.

Joo

(2012). Farmers’ perception of precision technology: The case of autosteer adoption by cotton farmers. Computers and Electronics in Agriculture, 87(September), 121-128. https://doi.org/10.1016/j.compag.2012.05.017

29.

Davis

F. D.

(1989). Perceived usefulness, perceived ease of use, and user acceptance of information technology. MIS Quarterly, 13(3), 319-340. https://doi.org/10.2307/249008

30.

Dessart

F. J.

Barreiro-Hurlé

van Bavel

(2019). Behavioural factors affecting the adoption of sustainable farming practices: A policy-oriented review. European Review of Agricultural Economics, 46(3), 417-471. https://doi.org/10.1093/erae/jbz019

31.

Diamantopoulos

Siguaw

J. A.

(2006). Formative versus reflective indicators in organizational measure development: A comparison and empirical illustration. British Journal of Management, 17(4), 263-282. https://doi.org/10.1111/j.1467-8551.2006.00500.x

32.

Dicks

L. V.

Rose

D. C.

Ang

Aston

Birch

A. N. E.

Boatman

Bowles

E. L.

Chadwick

Dinsdale

Durham

Elliott

Firbank

Humphreys

Jarvis

Jones

Kindred

Knight

S. M.

Lee

M. R. F.

Leifert

. . . Sutherland

W. J.

(2019). What agricultural practices are most likely to deliver “sustainable intensification” in the UK? Food and Energy Security, 8(1), Article e00148. https://doi.org/10.1002/fes3.148

33.

Finger

Swinton

S. M.

Benni

N. E.

Walter

(2019). Precision farming at the nexus of agricultural production and the environment. Annual Review of Resource Economics, 11(1). https://doi.org/10.1146/annurev-resource-100518-093929

34.

Fishbein

(1967). Attitude and the prediction of behavior. In Fishbein

(Ed.), Readings in attitude theory and measurement (pp. 477-492). Wiley.

35.

Fishbein

Ajzen

(1975). Belief, attitude, intention and behavior: An introduction to theory and research: Addison-Wesley series in social psychology. Addison-Wesley.

36.

Fishbein

Ajzen

(2010). Predicting and changing behavior: The reasoned action approach. Psychology Press.

37.

Flett

Alpass

Humphries

Massey

Morriss

Long

(2004). The technology acceptance model and use of technology in New Zealand dairy farming. Agricultural Systems, 80(2), 199-211. https://doi.org/10.1016/j.agsy.2003.08.002

38.

Floridi

Bartolini

Peerlings

Polman

Viaggi

(2013). Modelling the adoption of automatic milking systems in Noord-Holland. Bio-Based and Applied Economics, 2(1), 73-90. https://doi.org/10.13128/BAE-10882

39.

Folorunso

Ogunseye

(2008). Applying an enhanced technology acceptance model to knowledge management in agricultural extension services. Data Science Journal, 7, 31-45. https://doi.org/10.2481/dsj.7.31

40.

Gebbers

Adamchuk

V. I.

(2010). Precision agriculture and food security. Science (New York, N.Y.), 327(5967), 828-831. https://doi.org/10.1126/science.1183899

41.

Giesler

(2018). Digitalisierung in der Landwirtschaft: vom Precision Farming zum Farming 4.0 [Digitization in agriculture: From precision farming to farming 4.0] https://www.biooekonomie-bw.de/fachbeitrag/dossier/digitalisierung-in-der-landwirtschaft-vom-precision-farming-zum-farming-40

42.

Hair

J. F.

Ringle

C. M.

Sarstedt

(2011). PLS-SEM: Indeed a Silver Bullet. Journal of Marketing Theory and Practice, 19(2), 139-152. https://doi.org/10.2753/MTP1069-6679190202

43.

Han

S. H.

Kim

K. J.

Yun

M. H.

Hong

S. W.

Kim

(2004). Identifying mobile phone design features critical to user satisfaction. Human Factors and Ergonomics in Manufacturing & Service Industries, 14(1), 15-29. https://doi.org/10.1002/hfm.10051

44.

Haneklaus

Schnug

(2002). Aus dem Institut für Pflanzenernährung und Bodenkundke [An agronomic, ecological and economic assessment of site-specific fertilisation]. Landbauforschung Völkenrode, 52, 123-133. https://literatur.thuenen.de/digbib_extern/zi028255.pdf

45.

Harman

H. H.

(1967). Modern factor analysis. University of Chicago Press.

46.

Henseler

Hubona

Ray

P. A.

(2016). Using PLS path modeling in new technology research: Updated guidelines. Industrial Management & Data Systems, 116(1), 2-20. https://doi.org/10.1108/IMDS-09-2015-0382

47.

Henseler

Ringle

C. M.

Sarstedt

(2015). A new criterion for assessing discriminant validity in variance-based structural equation modeling. Journal of the Academy of Marketing Science, 43(1), 115-135. https://doi.org/10.1007/s11747-014-0403-8

48.

Hudson

Hite

(2003). Producer willingness to pay for precision application technology: Implications for government and the technology industry. Canadian Journal of Agricultural Economics/Revue Canadienne D’Agroeconomie, 51(1), 39-53. https://doi.org/10.1111/j.1744-7976.2003.tb00163.x

49.

Hulland

(1999). Use of partial least squares (PLS) in strategic management research: A review of four recent studies. Strategic Management Journal, 20(2), 195-204. https://doi.org/10.1002/%28SICI%291097-0266%28199902%2920%3A2%3C195%3A%3AAID-SMJ13%3E3.0.CO%3B2-7

50.

Jensen

H. G.

Jacobsen

L. B.

Pedersen

S. M.

Tavella

(2012). Socioeconomic impact of widespread adoption of precision farming and controlled traffic systems in Denmark. Precision Agriculture, 13(6), 661-677. https://doi.org/10.1007/s11119-012-9276-3

51.

Kindred

D. R.

Milne

A. E.

Webster

Marchant

B. P.

(2015). Exploring the spatial variation in the fertilizer-nitrogen requirement of wheat within fields. Journal of Agricultural Science, 153(1), 25-41. https://doi.org/10.1017/S0021859613000919

52.

Kollmuss

Agyeman

(2002). Mind the gap: Why do people act environmentally and what are the barriers to pro-environmental behavior? Environmental Education Research, 8(3), 239-260. https://doi.org/10.1080/13504620220145401

53.

Kortmann

(2015). The mediating role of strategic orientations on the relationship between ambidexterity-oriented decisions and innovative ambidexterity. Journal of Product Innovation Management, 32(5), 666-684. https://doi.org/10.1111/jpim.12151

54.

Kuhfuss

Préget

Thoyer

Hanley

Le Coent

Désolé

(2016). Nudges, social norms, and permanence in agri-environmental schemes. Land Economics, 92(4), 641-655. https://doi.org/10.3368/le.92.4.641

55.

Kutter

Tiemann

Siebert

Fountas

(2011). The role of communication and co-operation in the adoption of precision farming. Precision Agriculture, 12(1), 2-17. https://doi.org/10.1007/s11119-009-9150-0

56.

Lal

(2016). Feeding 11 billion on 0.5 billion hectare of area under cereal crops. Food and Energy Security, 5(4), 239-251. https://doi.org/10.1002/fes3.99

57.

Läpple

Kelley

(2015). Spatial dependence in the adoption of organic drystock farming in Ireland. European Review of Agricultural Economics, 42(2), 315-337. https://doi.org/10.1093/erae/jbu024

58.

Le Coent

Preget

Thoyer

. (2019). Do farmers follow the herd? The influence of social norms on participation in agri-environmental schemes. https://doi.org/10.22004/AG.ECON.289863

59.

Leaver

J. D.

(2011). Global food supply: A challenge for sustainable agriculture. Nutrition Bulletin, 36(4), 416-421. https://doi.org/10.1111/j.1467-3010.2011.01925.x

60.

Lindblom

Lundström

Ljung

Jonsson

(2017). Promoting sustainable intensification in precision agriculture: Review of decision support systems development and strategies. Precision Agriculture, 18(3), 309-331. https://doi.org/10.1007/s11119-016-9491-4

61.

Link

Graeff

Batchelor

W. D.

Claupein

(2006). Evaluating the economic and environmental impact of environmental compensation payment policy under uniform and variable-rate nitrogen management. Agricultural Systems, 91(1), 135-153. https://doi.org/10.1016/j.agsy.2006.02.003

62.

Little

R. J. A.

Rubin

D. B.

(2019). Statistical analysis with missing data (3rd rev. ed.). Wiley Blackwell.

63.

Lynne

G. D.

Casey

C. F.

Hodges

Rahmani

(1995). Conservation technology adoption decisions and the theory of planned behavior. Journal of Economic Psychology, 16(4), 581-598. https://doi.org/10.1016/0167-4870(95)00031-6

64.

Mahon

Crute

Simmons

Islam

M. M.

(2017). Sustainable intensification—“oxymoron” or “third-way”? A systematic review. Ecological Indicators, 74(March), 73-97. https://doi.org/10.1016/j.ecolind.2016.11.001

65.

Marra

Pannell

D. J.

Abadi Ghadim

(2003). The economics of risk, uncertainty and learning in the adoption of new agricultural technologies: Where are we on the learning curve? Agricultural Systems, 75(2-3), 215-234. https://doi.org/10.1016/S0308-521X(02)00066-5

66.

Marra

M. C.

Rejesus

R. M.

Roberts

R. K.

English

B. C.

Larson

J. A.

Larkin

S. L.

Martin

(2010). Estimating the demand and willingness-to-pay for cotton yield monitors. Precision Agriculture, 11(3), 215-238. https://doi.org/10.1007/s11119-009-9127-z

67.

McBride

Daberkow

(2003). Information and the adoption of precision farming technologies. Journal of Agribusiness, 21(1), 21-38. https://ageconsearch.umn.edu/record/14671/?ln=en

68.

Meemken

E.-M.

Qaim

(2018). Organic agriculture, food security, and the environment. Annual Review of Resource Economics, 10(1), 39-63. https://doi.org/10.1146/annurev-resource-100517-023252

69.

Mills

Gaskell

Ingram

Dwyer

Reed

Short

(2017). Engaging farmers in environmental management through a better understanding of behaviour. Agriculture and Human Values, 34(2), 283-299. https://doi.org/10.1007/s10460-016-9705-4

70.

Nyborg

Anderies

J. M.

Dannenberg

Lindahl

Schill

Schlüter

Adger

W. N.

Arrow

K. J.

Barrett

Carpenter

Chapin

F. S.

III Crépin

A.-S.

Daily

Ehrlich

Folke

Jager

Kautsky

Levin

S. A.

Madsen

O. J.

. . . de Zeeuw

(2016). Social norms as solutions. Science, 354(6308), 42-43. https://doi.org/10.1126/science.aaf8317

71.

Palm-Forster

L. H.

Ferraro

P. J.

Janusch

Vossler

C. A.

Messer

K. D.

(2019). Behavioral and experimental agri-environmental research: Methodological challenges, literature gaps, and recommendations. Environmental and Resource Economics, 73(3), 719-742. https://doi.org/10.1007/s10640-019-00342-x

72.

Paluck

E. L.

Shepherd

Aronow

P. M.

(2016). Changing climates of conflict: A social network experiment in 56 schools. Proceedings of the National Academy of Sciences of the United States of America, 113(3), 566-571. https://doi.org/10.1073/pnas.1514483113

73.

Pannell

D. J.

(2006). Flat earth economics: The far-reaching consequences of flat payoff functions in economic decision making. Review of Agricultural Economics, 28(4), 553-566. https://doi.org/10.1111/j.1467-9353.2006.00322.x

74.

Pathak

H. S.

Brown

Best

(2019). A systematic literature review of the factors affecting the precision agriculture adoption process. Precision Agriculture, 20(6), 1292-1316. https://doi.org/10.1007/s11119-019-09653-x

75.

Paustian

Theuvsen

(2017). Adoption of precision agriculture technologies by German crop farmers. Precision Agriculture, 18(5), 701-716. https://doi.org/10.1007/s11119-016-9482-5

76.

Petter

(2018). “Haters Gonna Hate”: PLS and information systems research. ACM SIGMIS Database: The DATABASE for Advances in Information Systems, 49(1), 10-13. https://doi.org/10.1145/3229335.3229337

77.

Podsakoff

P. M.

MacKenzie

S. B.

Lee

J.-Y.

Podsakoff

N. P.

(2003). Common method biases in behavioral research: A critical review of the literature and recommended remedies. Journal of Applied Psychology, 88(5), 879-903. https://doi.org/10.1037/0021-9010.88.5.879

78.

Popp

Lakner

Harangi-Rákos

Fári

(2014). The effect of bioenergy expansion: Food, energy, and environment. Renewable and Sustainable Energy Reviews, 32(April), 559-578. https://doi.org/10.1016/j.rser.2014.01.056

79.

Prager

Curfs

(2016). Using mental models to understand soil management. Soil Use and Management, 32(1), 36-44. https://doi.org/10.1111/sum.12244

80.

Pretty

Benton

T. G.

Bharucha

Z. P.

Dicks

L. V.

Flora

C. B.

Godfray

H. C. J.

Goulson

Hartley

Lampkin

Morris

Pierzynski

Vara Prasad

P. V.

Reganold

Rockström

Smith

Thorne

Wratten

(2018). Global assessment of agricultural system redesign for sustainable intensification. Nature Sustainability, 1(8), 441-446. https://doi.org/10.1038/s41893-018-0114-0

81.

Price

J. C.

Leviston

(2014). Predicting pro-environmental agricultural practices: The social, psychological and contextual influences on land management. Journal of Rural Studies, 34(April), 65-78. https://doi.org/10.1016/j.jrurstud.2013.10.001

82.

Rezaei-Moghaddam

Salehi

(2010). Agricultural specialists intention toward precision agriculture technologies: Integrating innovation characteristics to technology acceptance model. African Journal of Agricultural Research, 5(11), 1191-1199. https://doi.org/10.5897/AJAR09.506

83.

Rogers

Ancev

Whelan

(2016). Flat earth economics and site-specific crop management: How flat is flat? Precision Agriculture, 17(1), 108-120. https://doi.org/10.1007/s11119-015-9410-0

84.

Rogers

E. M.

(2003). Diffusion of innovations (5th ed., Free Press trade paperback ed.). Free Press.

85.

Rose

D. C.

Chilvers

(2018). Agriculture 4.0: Broadening responsible innovation in an era of smart farming. Frontiers in Sustainable Food Systems, 2, Article 87. https://doi.org/10.3389/fsufs.2018.00087

86.

Rose

D. C.

Sutherland

W. J.

Parker

Lobley

Winter

Morris

Twining

Ffoulkes

Amano

Dicks

L. V.

(2016). Decision support tools for agriculture: Towards effective design and delivery. Agricultural Systems, 149(November), 165-174. https://doi.org/10.1016/j.agsy.2016.09.009

87.

Rossi

Salinari

Poni

Caffi

Bettati

(2014). Addressing the implementation problem in agricultural decision support systems: The example of vite.net®. Computers and Electronics in Agriculture, 100(January), 88-99. https://doi.org/10.1016/j.compag.2013.10.011

88.

Schieffer

Dillon

(2015). The economic and environmental impacts of precision agriculture and interactions with agro-environmental policy. Precision Agriculture, 16(1), 46-61. https://doi.org/10.1007/s11119-014-9382-5

89.

Sheppard

B. H.

Hartwick

Warshaw

P. R.

(1988). The theory of reasoned action: A meta-analysis of past research with recommendations for modifications and future research. Journal of Consumer Research, 15(3), 325-343. http://www.jstor.org/stable/2489467

90.

Smith

Holmes

Paavola

(2018). Social underpinnings of ecological knowledge: Business perceptions of biodiversity as social learning. Organization & Environment, 33(2), 175-194. https://doi.org/10.1177/1086026618803723

91.

Snyder

C. S.

Bruulsema

T. W.

Jensen

T. L.

Fixen

P. E.

(2009). Review of greenhouse gas emissions from crop production systems and fertilizer management effects. Agriculture, Ecosystems & Environment, 133(3-4), 247-266. https://doi.org/10.1016/j.agee.2009.04.021

92.

Stuart

Schewe

R. L.

McDermott

(2012). Responding to climate change. Organization & Environment, 25(3), 308-327. https://doi.org/10.1177/1086026612456536

93.

Tamirat

T. W.

Pedersen

S. M.

Lind

K. M.

(2018). Farm and operator characteristics affecting adoption of precision agriculture in Denmark and Germany. Acta Agriculturae Scandinavica: Section B: Soil & Plant Science, 68(4), 349-357. https://doi.org/10.1080/09064710.2017.1402949

94.

Uehleke

Hüttel

(2019). The free-rider deficit in the demand for farm animal welfare-labelled meat. European Review of Agricultural Economics, 46(2), 291-318. https://doi.org/10.1093/erae/jby025

95.

United Nations. (2017). World Population Prospects: The 2017 Revision: Key findings and advance tables. https://reliefweb.int/report/world/world-population-prospects-2017-revision-key-findings-and-advance-tables

96.

van der Wal

. (2019, January 24). Why is adoption of precision ag so slow? Expert opinion. Future Farming, Smart Farmers. https://www.futurefarming.com/Smart-farmers/Articles/2019/1/Why-is-adoption-of-precision-ag-so-slow-385338E/

97.

van Dijk

W. F.A.

Lokhorst

A. M.

Berendse

Snoo

G. R. de

(2016). Factors underlying farmers’ intentions to perform unsubsidised agri-environmental measures. Land Use Policy, 59, 207-216. https://doi.org/10.1016/j.landusepol.2016.09.003

98.

Walter

Finger

Huber

Buchmann

(2017). Opinion: Smart farming is key to developing sustainable agriculture. Proceedings of the National Academy of Sciences of the United States of America, 114(24), 6148-6150. https://doi.org/10.1073/pnas.1707462114

99.

Weersink

Fraser

Pannell

Duncan

Rotz

(2018). Opportunities and challenges for big data in agricultural and environmental analysis. Annual Review of Resource Economics, 10(1), 19-37. https://doi.org/10.1146/annurev-resource-100516-053654

100.

Weltin

Zasada

Piorr

Debolini

Geniaux

Moreno Perez

Scherer

Marco

L. T.

Schulp

C. J. E.

(2018). Conceptualising fields of action for sustainable intensification: A systematic literature review and application to regional case studies. Agriculture, Ecosystems & Environment, 257(1), 68-80. https://doi.org/10.1016/j.agee.2018.01.023

101.

Winn

M. I.

Pogutz

(2013). Business, ecosystems, and biodiversity. Organization & Environment, 26(2), 203-229. https://doi.org/10.1177/1086026613490173

102.

Wolfert

Verdouw

Bogaardt

M.‑J

. (2017). Big Data in Smart Farming: A review. Agricultural Systems, 153(May), 69-80. https://doi.org/10.1016/j.agsy.2017.01.023

103.

Wolz

(2013). Organisation der Agrarproduktion in Ostdeutschland seit dem 2. Weltkrieg: Historische Wurzeln und aktuelle Situation [The organisation of agricultural production in East Germany since World War II: Historical roots and present situation] (IAMO Discussion Papers No. 139). Leibniz Institute of Agricultural Development in Transition Economies (IAMO) website: https://www.econstor.eu/bitstream/10419/83981/1/766253864.pdf

104.

Xie

Huang

Chen

Zhang

(2019). Prospects for agricultural sustainable intensification: A review of research. Land, 8(11), 157. https://doi.org/10.3390/land8110157

105.

Yeboah

F. K.

Lupi

Kaplowitz

M. D.

(2015). Agricultural landowners’ willingness to participate in a filter strip program for watershed protection. Land Use Policy, 49(December), 75-85. https://doi.org/10.1016/j.landusepol.2015.07.016

106.

Zarco-Tejada

Hubbard

Loudjani

(2014). Precision agriculture: An opportunity for EU farmers: Potential support with the CAP 2014-2020. https://www.europarl.europa.eu/RegData/etudes/note/join/2014/529049/IPOL-AGRI_NT%282014%29529049_EN.pdf