What Drives Landowners to Resist Selling Their Land? Insights from Ethical Capitalism and Landowners’ Perceptions

: Foreign land grabbing is acknowledged as a phenomenon that generates disempowerment and dispossession of local farmers, human rights violations. Previous studies have revealed the lack of ethical benchmarks in foreign large-scale land transactions that raise moral concerns. It is evident that when resources are scarce and people depend on them, the balance between values and interests transforms itself into a dilemma. Within this context, the aims of the paper were to bring to the fore critical reflection on a more ethical perspective of large-scale land acquisitions and to extend the scant information on what factors determine landowners not to sell their land to foreigners to limit land grabbing. This context justifies the need for a critical reflection on a more ethical perspective of large-scale land acquisitions. Therefore, two objectives were set. The first one is to document the role of ethics in large-scale land transactions. Based on the land grabbing literature, authors selected a set of eight land grabbing narratives, most often interrelated and overlapping, that pose ethical considerations. The second objective is to reveal how well a set of variables can predict the “Resistance to sell” the land to foreigners even when an attractive price is offered. As ethics is a social construct, the analysis captured the stakeholders’ perspective on land grabbing. Therefore, a questionnaire was applied to a sample of 332 Romanian landowners from twelve randomly selected counties to reveal their perceptions. Partial least squares structural equation modeling (PLS-SEM) was used to observe how well a set of seven variables could predict landowners’ “Resistance to sell” their land to foreign buyers. The use of PLS-SEM was justified by the existence of single items and the need to examine many structural model relations. Results showed that the variables with the strongest contribution to the prediction of the dependent variable are the “Probability to join an association for farmers rights defense”, the “Importance of the land price offered by the potential foreign buyer”, and the “Perceived effect of agricultural land conversion to urban land”. Raising awareness on the importance of buyer attributes, increasing people’s perception of the negative effect of agricultural land conversion to urban land, or strengthening the state’s image as a necessary actor to limit land grabbing will increase landowners’ resistance to sell their land to foreigners. Finally, it can be inferred that, within this frame of discussion, ethics should be valued as a means to create economically viable and morally justifiable solutions for foreign large-scale land transactions. from


Introduction
Land grabbing is one triggering factor of agrarian change that reshaped rural Romania, which led to the restructuring of landscapes and livelihoods with visible environmental consequences and socio-economic implications [1]. There is no official statistical information on the number of foreign investors who own agricultural land, but, according to data from various sources [2], about 30% of Romania's agricultural land has come into foreign investors' possession. Worldwide, land grabbing is triggered by a large number of drivers. Giovannetti and Ticci [3] point to biofuel-related projects that amplified international land acquisitions, showing that biofuel crops account for a share of about 54% in terms of the total number of deals in Africa. Green pretexts of areas in need of protection have contributed to land dispossession [4][5][6][7]. The distribution and accessibility of food are judged as the main drivers towards land [8,9]. Not least, developing nations in a rush to attract foreign direct investments increase the number of land grabbers [10,11]. The developed world is not bypassed by illegal land transactions commonly associated with bribes to access land for urban development [12]. Soil fertility, the permissive normative framework for foreign land acquisition, or the affordable land price are named [13,14] among the significant factors that have stimulated land grab in Romania. According to Eurostat [15], Romania has the cheapest arable land, with a hectare costing an average of EUR 1958, whereas the most expensive is in the Netherlands, at about EUR 63,000.
Land grabbing is a controversial concept [16], with no commonly accepted definition. Baker-Smith and Szocs Boruss [17] understand land grabbing as "the control (whether through ownership, lease, concession contracts, quotas, or general power) of larger than locally typical amounts of land by any person or entity (public or private, foreign or domestic) via any means ('legal' or 'illegal') for purposes of speculation, extraction, resource control or commodification at the expense of peasant farmers, agroecology, land stewardship, food sovereignty, and human rights." As can be observed, many scholars [16,17] agree that land grabbing does not necessarily imply that a transaction is illegal or that it is always a harmful process. Hall [18] warns that this concept knows vast differences in the legality, structure, and outcomes of land deals. He emphasizes domestic elites and governments' roles as partners, intermediaries, and beneficiaries of land grabbing. Therefore, a uniform land grab metanarrative is not applicable [19]. Margulis et al. highlight that land grabbing is mainly a control grabbing over resources, as it is "characterized by transnational and domestic corporate investors, governments, and local elites taking control over large quantities of land (and its minerals and water) to produce food, feed, biofuel, and other industrial commodities for the international or domestic markets" [20]. This definition also mentions that land grabbing can be both national and foreign, although foreign land grabbing is the core of this phenomenon. In the case of Romania, the buyers of large land surfaces are mostly foreigners. Land Matrix is an international database that registers land transactions over 200 hectares, and it one of the most complete and reliable data sources for large land transactions. According to Land Matrix, there are only three companies registered in Romania among the top parent companies of land investors in Romania (Maxagro SRL, ADS, and Lord Energy SRL), all the others being foreign, mostly from the EU [21]. These three companies cover together less than 6% of the total deals under contracts in Romania (over 200 ha). For this reason, the present study focuses on foreign land grabbing.
Local farmers affected by land grabbing have been widely portrayed as victims [22,23]. It is argued that foreign land grabbing that is transnational in nature enriches foreign investors [24] and generates disempowerment and dispossession of local farmers [25], arriving at the violation of human rights [26]. Land grabbing leads to the transfer of control over property and resources located on extended territories from the local level to outsiders within transactions dominated by an asymmetry of power. Thus, it is considered an economic, cultural, and ecological threat to rural communities [27]. It was estimated that as a direct result of land grabbing, 200-300 million people globally were at risk of greater food insecurity [28]. From an economic perspective, land grabbing might cause income loss of over 12 million people globally and that it had serious implications for food security, poverty levels, and urbanization [28]. In Romania, low land price, good soil quality, rural desertification, lack of adequate loan opportunities for local rural inhabitants, liberalization of land market to other EU member states, and corruption enhanced people vulnerability in front of foreign buying proposals for land and lead to the transfer of between 20% and 50% of the land to foreign property in some counties [13]. This evidence emerging from the existing land grabbing literature justifies the need for a critical reflection on a more ethical perspective of large-scale land acquisitions. Margulis [29] explained that large-scale land acquisitions are seen as unethical when they undermine sustainable development principles and when they result in constraints that force the poor to work for the wealthy. The present contribution considers that ethics should be valued as an integral part of day-to-day choices and decisions in land management [30] and that it should play its role, not as a guest but as a guide [31] regarding land grabbing-a statement further argued in detail in Section 2.
Within this frame of discussion, the paper aims to bring to the fore critical reflection on a more ethical perspective of large-scale land acquisitions and extend the scant information on what factors determine landowners not to sell their land to foreigners to limit land grabbing. Therefore, two objectives were set: (i) the first one is to document the role of ethics in large-scale land transactions; (ii) the second objective is to reveal how well a set of variables can predict Romanian landowners' resistance to sell the land to foreigners even if an attractive price is offered.
Even if the present contribution embeds a place-based analysis that reveals the local environment, it also responds to a fast-evolving European phenomenon. Like other studies, the present one discusses the land grabbing phenomenon in the light of the current social context [32,33] and the trends of rural development in European societies [34]. Nevertheless, the discussion on foreign large-scale land acquisitions was brought to the sphere of normative ethics. Thus, this paper adds to the existing land grabbing literature by revealing the effect of seven variables on landowners' resistance to sell their land to foreign buyers to limit land grabbing. Although the current research on "global land grabbing" [9,[35][36][37][38] has enriched the understanding of the roles of various actors and factors involved in land deals, no study has identified, so far, the factors that can influence the landowners' decision to refuse to sell the land even when a good price is offered. Thus, while other studies investigated the people's reasons for persistent landownership in the rural-urban fringe [39] or land attachment in general [40][41][42], this is the first attempt to reveal factors contributing to Romanian landowners' resistance to sell within the foreign land grabbing context.

Ethical Capitalism and the Need for Ethics in Large-Scale Land Transactions
Within the land grabbing equation, grabbers perceive land value much higher than the price paid for it, which is the underlying reason for land acquisitions' magnitude. As a result, in practice, foreign land investors act competitively to meet their unlimited wants with their limited means [43], put their financial benefits first and offer as little as possible to other stakeholders, like farmers and other landowners. This approach ignores the reality of value creation as a collective process and, because, most often, large-scale land transactions take place within the capitalism framework, it can be inferred that one of the greatest weaknesses of capitalism is that it confuses the price with value [44]. In the case of land, these specific conditions and factors influence the land's value, at a certain level. Place attachment, national security, and traditional preservation as creators of land value can be named here.
Lefebvre and Nicholson-Smith [45] observe that "capitalism has taken possession of the land", meaning that land is most often seen only as a concept of political economy in a worldwide context where 40% of the economic activities of many capitalist countries are connected to land [46]. Yes, the land is capital and commodity, but, as per Polany [47], it is a fictitious commodity that fulfills a social necessity. If the land is treated only as a commodity for sale on the market, our social world will be endangered rather than protecting rights related to it [48]. David Bollier [49] says the land is "embedded in social relationships and subject to the moral consideration." That is why it is time to rethink the role of land in economics and other sciences [50].
The concept "ethical capitalism" refers to acting in business [51] and because largescale land transactions are a flourishing business, the discussion about "ethical capitalism" in this context can contribute to overcoming some of the negative land grabbing consequences. Capitalism is neutral; it is neither ethical nor unethical [52]. It is about us, those who practice it, who can act either morally or immorally. Ethics is about people. Ethical initiatives are often initiated by civil society, and this kind of private value-oriented action influencing market behavior can be called ethical capitalism [24]. Moreover, ethical capitalism is about corporations' responses to ethical concerns that the legal framework does not address [51].
Ethical capitalism must lead to changed behaviors [53], but we must know these changing behaviors before changing them. Consequently, the investigation of people's perception of land grabbing is relevant because justice, in general, is constructed and negotiated through social interaction, and it is based on the compatibility of supporting values [54]. As argued by Asiama et al. [43], owners' views (perceptions, in our case) on largescale land transactions should be studied because this approach will help determine the land value that people will agree with. From a practical perspective, knowledge creation based on the investigation of the gravity of the land grabbing phenomenon, its impact on jobs, the environment, and food security can contribute to future land law amendments.
Zooming in on land rush literature, it is evident that land grabbing cannot be judged only in black or white, as some see land grabs as a significant threat to rural communities' livelihoods, while others consider these transactions as an economic opportunity. Even if intermediate views still exist (see, for example, Borras and Franco [55]), most studies tend to consider mainly the imputations of land grabbing, which conjure up a future of resource scarcity, environmental depletion, food insecurity, or infringement of local community rights. Practically, this perspective places the land grabbing debate on the "us against them" axis (where "us" represents the local farmers and "them", foreign investors) along with the emphasis of land grabbing negative consequences that outweigh the benefits. Idealistically, all participants in a land transaction would act based on ethical values. Practically, the lack of ethical benchmarks constructed a rich vein of research that reports land grabbing narratives that raise ethical concerns. Unfortunately, when resources are scarce and people rely so much on them, the balance between values and interests transforms into a real dilemma [24].
According to Damen [24], the introduction of ethical dilemmas raises, in its turn, a policy dilemma. One big question in land management decisions is how to balance economic interests and ethical values. Literature offers several answers. There are authors [72] who suggest that corporate social responsibility and a model code of conduct, promoted by international organizations, can be a solution for the conflicts arising from agrarian transformations driven by land grabbing. How the principles of such codes might work in practice is difficult to envisage. Other authors highlight problems, doubts, and concerns regarding codes of conduct and warn that they may, in fact, facilitate land grabbing because they do not address the roots of land grabbing [55]. In their opinion, the solution is a new distribution of political forces (from international to local levels) that follows a human rights-based approach [55]. Practically, international organizations (e.g., the Food and Agriculture Organization, the International Fund for Agricultural Development) cannot significantly contribute to the fight against land grabbing' negative consequences, as it is a phenomenon linked to property law and national policies [24]. Factors, such as different interpretations of the same regulatory instrument by the actors involved, the existence of legally pluralistic contexts, or the multi-facets of conflicts that cannot be solved with a single instrument alone, are blamed [73] for the failure of the international initiatives. Let us turn the attention to domestic law. It is acknowledged that there is wide variability in the legal underpinnings of customary rights and legislated processes for large-scale land acquisition [74], and legislative gaps often contribute to shaping the opportunities and risks faced by landowners.
Within the land grabbing context, normative ethics contribute to the debate on specific narratives that encompass social equity, human capabilities, or environment protection. In this analysis, ethics is valued as a social construct built by negotiating interpretations of morally desirable situations and behaviors accepted in a particular society (Romanian society, in the present case) at a specific historical moment [75]. This definition of ethics needs some more clarification. Ethics may be context-dependent [76,77], and even when individuals recognize some standard ethical norms, they can interpret and apply them differently under their values and life experiences [78]. To put the point a little bit more empirically, how can landowners make ethical decisions about selling their land? Large-scale agribusinesses are often portrayed as culpable for environmental degradation following their agricultural activity [79], and thus they infringe one of Bawden's [80] ethical canons, namely, "the principle of non-maleficence". Liao et al. [81] concluded that because of land-cover change brought by large foreign farmland transactions, amounts of carbon emissions could be emitted when transacted land is located in high carbon value forests. A rich literature highlights that land transactions accelerated deforestation [82][83][84]. Cardoso and James [85] said that very little research exists concerning farmers' or landowners' ethical frameworks and the extent to which ethics affects their decisions. Several models exist explaining ethical decision-making [85][86][87]. If we follow Jones's [87] definition of ethical decision, as "a decision that is both legal and morally acceptable to the larger community… an unethical decision is either illegal or morally unacceptable to the larger community", one could say that as long as the land grabbing negatively impact on environmental balance, displacement of small farmers, or food security, the ethical approach means the limitation of the land grabbing phenomenon. The temporary freeze on investments involving large-scale land acquisitions that are already in place (e.g., Mozambique and Cambodia) or information campaigns of local communities that empower them to give or refuse consent to a project, guarantees of project transparency so that investors can be held accountable both to affected communities and to the government, are some of the ethical solutions.
Ethics, seen as a social construct, requires that the discussion includes the stakeholders' perspectives (landowners in the present study). We asked how landowners can make "right" decisions about land transactions in a world marked by the chaos of values. Meijboom and Stafleu [88] argue that farmers have moral beliefs and values, and thus they can contribute to the public debate on relevant agricultural issues, from animal welfare, sustainable food production, to environmental protection. That is why the focus of this study is placed on landowners (who may be farmers or not) considered stakeholders in land transactions. Practically, stakeholder analysis generates knowledge about stakeholders' actions that helps us understand their behavior, intentions, and interests. According to Varvasovszky and Brugha [89], stakeholder analysis assesses the influence of these interests on decision making or implementation processes. Landowners' perceptions are worthy of investigation as any change or maintenance of the status quo in land transactions is possible only with their involvement which depends on their perceptions, knowledge, and needs [75].

Study Area
Romania ( Figure 1) is a member of the European Union, located in the southeast of Central Europe. It has an area of 238,391 km 2 that represents 4.8% of the territory of Europe and 5.4% of the area of the European Union [90], respectively. The Romanian rural area covers 207,522 km 2 , representing 87.1% of the country's surface, and concentrates 9.2 million inhabitants (46.0% of the total population) [90,91]. The share of the agricultural sector in the gross domestic product (GDP) in 2019 was 4.1%, knowing a downward trend (for example, in 2011, its contribution to the GDP was 6.25%) [92]. Romania ranks 49th in the Human Development Index [93].

Data Collection
Data were collected as follows. Twelve counties (marked in Figure 1) were selected at random, and three cities/villages were randomly selected in each county. Randomness was ensured as follows. The authors created a list with the 41 counties of Romania plus de capital and numbered them. A list with 12 random numbers was generated in Excel and the counties with those numbers were selected for the survey. The same procedure was followed to select the three cities/villages in each of the twelve counties. The lists with counties, cities, and villages are available online. In each city/village, a list with streets was generated using the information available online and one street was randomly selected. Interviewers created a list of the supermarkets, pharmacies, and markets that functioned on that street and one of them was randomly selected. If no shop existed on the selected street, a new street was picked at random. An interview was requested for each fourth person who came out of the selected shop. Between six and ten interviews were carried on in each city/village on a person who was over 18 years old and owned land. The acceptance to respond rate with complete answers was 21%. Participation in the survey was voluntary, and respondents were informed about the nature of the research.
As a consequence of the successive agricultural reforms aimed at land property restitution, urban citizens represent a significant percentage of landowners. Consequently, both urban and rural localities were considered, and landowners were the target group in this research regardless of their quality of being or not farmers. The final sample included 332 Romanian citizens.

Investigated Variables
The economic structure of agriculture, the political instability, and land market volatility, for example, show that the landowners have limited choices in their land management decisions. The perceived economic pressure is practically directly correlated with landowners participating in ethically questionable land transactions [94,95]. At the outset, it was assumed that Romanian landowners have often found themselves in an ethical dilemma to meet what seem to be conflicting needs and values. More precisely, these can be the perception of personal well-being through the money received in large scale land transactions vs. the impact of landowner's selling decision on the environment, food security, community welfare (see also [88,95,96] for the argumentation of the existing farmers' moral beliefs and values). The present study questioned the relationship between a set of seven variables and landowners' decisions of not selling the land to foreign investors. "Not sell to foreign investors, even if they offer you a good price" (briefly named "Resistance to Sell") was set as the dependent variable. Often, foreign investors offer higher prices than domestic ones, so the higher price was included in the scenario. The rest of the selected variables (  [97] tried to understand large-scale land acquisition from the systems perspective that asks questions about the land grabbing's overall scale and its main characteristics. Landowners' perceptions of a specific phenomenon, like land grabbing, depend on the existing knowledge about the investigated topic [98]. Consequently, considering the primary actors' perceptions-Romanian landowners-of the land grabbing phenomenon is critical in affecting land use behavior [99,100]. The variables were further associated with the eight land grabbing ethical narratives (detailed in Section 2). Table 1. Investigated variables (the short name used for each variable in the partial least squares structural equation modeling (PLS-SEM) analysis is written between brackets in the first column).

Investigated Variables
Association with the Identified Land Grabbing Narratives

Questionnaire Question/Statement Answer Options
Perceived gravity of foreign land grabbing over the next 10 years (independent variable) (LG gravity) "Future overall gravityland grabbing" narrative How do you assess the gravity of foreign land grabbing in the next 10 years in Romania? Probability to join an association for farmers rights defense to limit the land grabbing phenomenon (Probability to join a farmers' rights association) (independent variable) "Weak solidarity among farmers" narrative What is the probability that you join an association that defends farmers' rights to limit the phenomenon of land grabbing? 1 = Very low probability to 5 = Very high probability (reverse coded here compared to how it was presented to respondents) Probability of not selling the land to foreign investors, even if they offer a good price, to limit the land grabbing phenomenon (dependent variable) (Resistance to sell) "Xenophobia-land grabbing" narrative What is the probability that you do not sell your land to foreign investors, even if they offer you a good price, to limit the phenomenon of land grabbing? 1 = Very low probability to 5 = Very high probability (reverse coded here compared to how it was presented to respondents)

Data Analyses
Univariate analyses were made using Excel and SPSS. To discover the causal relationships between the construct, partial least squares structural equation modeling (PLS-SEM) was run. Structural equations model (SEM) based on partial least squares (PLS) were used through SmartPLS software (3.2.8). This software has a high ability to discover the causal relationships of complex models [101]. PLS-SEM was preferred because it depends less on the assumed normal distribution and, consequently, it has higher estimation accuracy, it is well-fitted in a construct with a single measurement, and it is acceptable for most research predictions [102]. Another strength of the PLS method is that a construct with only one indicator does not create a problem in the PLS method computation process [103]. Further, this software has a special ability in calculating formative and reflective measurements [101]. In other words, the software facilitates and confirms relationships, based on the relationships between research variables indicated by the researcher. In the present study, only the formative measurement was used. The "Resistance to sell" is the main construct of the model. There are direct relationships between the main construct and exogenous constructs such as the "Need for state intervention", the "Importance of buyer attributes", the "Perceived effect of agricultural land conversion to urban land", and the "Perceived effect of foreign land grabbing on the natural environment". There are also indirect relationships such as between the main construct and the "Perceived gravity of foreign land grabbing over the next 10 years" through the "Influence of the land price offered by the potential foreign buyer". Several methods can be used to draw a model and determine direct or indirect paths, such as theory, logic, previous research, qualitative research, or researcher judgment [101]. In this paper, we used previous research, logic, and researcher judgment.
Furthermore, 500 samples were used for bootstrapping method. Then, to ensure the correctness of the analysis with 300 samples, the bootstrapping method was performed again, which showed that the path coefficients and significant levels of the model are not different from each other and are valid for further interpretation of the model. This property has important implications for testing the significances of the model coefficients.

Results
The sample characteristics are presented in Table 2. Descriptive statistical tests were run in SPSS. The demographic profile of interviewed persons was described by gender, age, education, residence area (rural, and urban), and income ( Table 2). As can be seen in Table 3, the "Gravity of foreign land grabbing gravity over the next 10 years" and the "Need for state intervention to limit foreign land grabbing phenomenon" had the highest mean. This result showed that according to the Romanians, the government has a very high responsibility in limiting land grabbing. Moreover, Romanians believed that the phenomenon of land grabbing would seriously intensify in the next 10 years. Probability of not selling the land to foreign investors, even if they offer a good price, to limit the land grabbing phenomenon (1 = Very low probability; 5 = Very high probability) 332 3.01 1.192 * This variable was computed as average score of the items indicated in Table 3.

Measurement Model
To obtain valid results from the structural model of the research, first, the optimal conditions of the structural model should be provided using the measurement model. Figure 2 contains seven independent variables and one dependent variable. Figure 2 shows that out of eight variables, six variables were measured using an indicator "Perceived gravity of foreign land grabbing", "Influence of price received", "Perceived need for state intervention to limit foreign land grabbing", "Probability to join an association for farmers rights defense", "Perceived effect of foreign land grabbing on the natural environment", "Perceived effect of agricultural land conversion to urban land", and "Resistance to sell"), and one variable was measured using three indicators ("Importance of buyer attributes"). The model constructs were all measured using one indicator, except for the importance of buyer attributes. For this purpose, the model needed to assess the validity. As shown in Figure 2, the "Importance of buyer attributes" construct was a formative construct. Its convergent validity, multicollinearity between indicators, and outer weight were measured. Convergent validity (β = 0.748) and multicollinearity (ranged 1 to 1.113) were confirmed. Outer weights are shown in Table 4, and all of them are significant [101].  The general model was drawn after determining the suitability of the constructs. According to the results, Standardized Root Mean Square Residual (SRMR, which is defined as the difference between the observed correlation and the model implied correlation matrix and it is an absolute measure of model fit) was equal to 0.066, which indicated a good fit model of the research model [101].

Structural Model of Landowners' Resistance to Sell
The results of the research model were presented in Figures 2 and 3. Figure 2 showed the path coefficients between constructs, the outer weights of the indicators of the formative construct, and the R 2 of the dependent construct.  Figure 3 showed the Path coefficients and the level of significance of the paths. The significance of paths was obtained by bootstrapping. The highest level of significance and coefficient path belonged to the path of "Perceived gravity of foreign land grabbing" to "Perceived need for state intervention to limit land grabbing" and the lowest coefficient path and level of significance belonged to the path of "Perceived need for state intervention to limit land grabbing" to "Resistance to sell". Table 5 showed the total effect, direct effect, and indirect effect of constructs. According to Table 5, the effect of "Perceived effect of agricultural land conversion to urban land" had a direct and significant positive effect on resistance to sell (β = 0.144, sig. = 0.008). The "Importance of buyer attributes" had a positive and direct effect on resistance to sell (β = 0.135, sig. = 0.028).  Table  1).
The "Perceived gravity of foreign land grabbing" did not have a significant effect directly on "Resistance to sell" (β = -.007, sig. = 0.906), but through "Perceived need for state intervention to limit foreign land grabbing" (β = 0.419, sig. = 0.000) and "Importance of buyer attributes" (β = 0.405, sig. = 0.000), it was able to have a significant and positive effect on resistance to selling. Therefore, "Perceived gravity of foreign land grabbing" was able to influence resistance to sell in the presence of intermediaries such as "Perceived need for state intervention to limit foreign land grabbing" and "Importance of buyer attributes" (β = 0.092, sig. = 0.069). As a result, the higher the "Perceived gravity of foreign land grabbing", the greater the "Resistance to sell". The "Perceived effect of foreign land grabbing on the natural environment" had a positive and significant effect on "Resistance to sell" (β = 0.141, sig. = 0.006). This means that with increasing "Perceived effect of foreign land grabbing on the natural environment", "Resistance to sell" also increases.
Moreover, the "Probability to join an association for farmers rights defense" had a positive and significant effect on "Resistance to sell". The indirect route through government intervention to "Resistance to sell" did not reveal a significant effect (β = 0.0.012, sig. = 0.181). Nevertheless, the total and direct effects were highly significant (β = 0.303, sig. = 0.000). This result showed that with increasing "Probability to join an association for farmers rights defense", the tendency to "Resist to sell" increases.
"Perceived need for state intervention to limit foreign land grabbing" had a positive and significant effect on "Resistance to sell" (β = 0.105, sig. = 0.042). It can be argued that with increasing the perceived need for government intervention, "Resistance to sell" also increases.
"Influence of price" had a significant negative effect on Resistance to sell" (β = -.263, sig. = 0.000). This means that by increasing the bid price to the owners, the owner is less likely to reject the purchase request when the price matters to him/her.
According to the structural model of the research, the values of Stone-Geisser's Q 2 , R 2 , and R 2 adj were investigated ( Table 6). The results indicated that the value of Q 2 was positive for all endogenous variables and ranged from 0.119 to 0.285, which revealed the ability to predict the model accurately [104]. R 2 values also indicated that the model has an acceptable level of predictive power [101,105]. R 2 adj showed that the constructs of the research model have the ability to explain 31.7% of the variance changes of the dependent variable of the model ("Resistance to sell"). 31.7% is an acceptable value to explain variance changes.

Discussion
The study results revealed that the seven selected variables have a good capacity to explain landowners' "Resistance to sell" their land to foreign buyers. The fact the "Perceived gravity of foreign land grabbing over the next 10 years" and the "Perceived need for state intervention to limit foreign land grabbing" (Table 5) obtained the highest scores among the tested variables showed that the Romanian landowners believed that the phenomenon of land grabbing would seriously intensify in the next 10 years and that the government had a very high responsibility in limiting land grabbing. Interestingly, the results of PLS-SEM analysis indicated that the "Perceived gravity of foreign land grabbing over the next 10 years" did not have a direct effect on landowners' "Resistance to sell" their land (p = 0.896; Figure 3).
In other words, it is not enough to raise awareness on the gravity of land grabbing among landowners to determine them not to sell their land to limit foreign land grabbing. This may be caused by the fact that landowners did not consider that their contribution through "Resistance to sell" their land could significantly impact the reduction of land grabbing gravity over a time-span of 10 years. However, "Perceived gravity of foreign land grabbing" had an indirect effect on "Resistance to Sell", through "Perceived need for state intervention to limit foreign land grabbing" and the "Importance of buyer attributes", which were both significant at p-value 0.000. This means that the more serious the perceived effect of land grabbing was, the higher landowners' resistance to sell their land was but only when they saw a high need for state intervention to limit land grabbing and when buyer attributes were important to them. In other words, in the absence of the "Perceived need for state intervention to limit foreign land grabbing", the perception of "Perceived gravity of foreign land grabbing" did not influence landowners' "Resistance to sell". This result can be attributed to the fact that people decide to support land grabbing limitation by resisting selling only when they consider that state intervention is required. The mediation effect of "Perceived need for state intervention to limit foreign land grabbing" could indicate that, for interviewed landowners, the necessity of state intervention reflects the gravity of land grabbing. It can also show landowners' trust in state capacity to limit land grabbing, which is not sustained in some cases because the state, through its legal framework and institutions, is acknowledged to be a significant driving factor for foreign land grabbing [106]. As land grabbing produces changes in how, for example, the needs of a local community are addressed, how resources are exploited, or how the employment contracts are executed [107], under some circumstances, paternalist policies (where "paternalism" means any intervention of governments or other authorities in private decision-making [108]) can be justifiable where high stakes are involved. Thus, the public authorities' dilemma will be to find a rationale for a restrictive approach to the existing fragmented public intervention on large-scale land transactions, which will probably displease many foreign investors and domestic landowners alike.
The "Importance of buyer attributes" also had predictive power on the dependent variable. It was observed that people who cared who were those to whom they could sell their land would be more reluctant in selling it to foreigners. Real-life situations lived by interviewed people or their friends and relatives and cases debated mainly through mass media have induced the fear that land was bought to a dangerous extent by foreign persons and companies [71]. This may be a factor that increased landowners' attention to buyer characteristics and determined them to consider attributes such as nationality or future land destination in land transactions. This result suggests that a simple action such as disclosing more about the potential buyer (such as nationality and future land destination given by the buyer) may determine landowners interested in buyer characteristics not to sell their land. Additionally, the "Importance of buyer attributes" mediated the relationship between the "Importance of the land price offered by the potential foreign buyer" and the "Resistance to sell".
The variable "Probability to join an association for farmers rights defense" had predictive power on landowners' "Resistance to sell" their land. Additionally, this variable indirectly affects the "Perceived need for state intervention to limit foreign land grabbing". Rawls [109] states simply that cooperation allows "a better life for all than any would have if each were to live solely by his own efforts". Thus, the ethical canon of "inclusiveness" (the principle of equitability) is applicable in this context and contributes to the foundation of the "good life" [80].
The Universal Declaration of Human Rights [110] proclaims the freedom of association as a fundamental right. This right enables non-state actors' participation in economic and social policy, and the implementation of this right is a critical element of every country's rule of law system. Article 12 of the EU Charter of Fundamental Rights enshrines this right, too. Any farmers' association offers them an excellent environment to promote their rights and interests, enhance their skills, and improve their livelihood. Romanian Ministry of Agriculture and Rural Development [111] acknowledges that the difficulty for small farmers to reach the market or obtain reliable information about the market can be overcome if they understand the benefits and agree to enter into farmers' associative forms. In Romania, only one percentage (1%) of farmers are part of an associative structure, which contrasts sharply with other European Union member states [112]. In comparison, it should be noted that the European average is 34% [112].
Romanian farmers are reluctant to join an association probably because of their experience during the communist period when their land was brutally taken away and they were forced to join the agricultural production cooperatives. The high prediction power of the "Probability to join an association for farmers rights defense" conveys the idea that the best way (among the ones tested in this study) to increase landowners' determination to refuse to sell their land to foreign investors is to offer them the possibility to join a farmers' association. The challenge is to create associations that farmers will trust and join. This causal relationship may be supported by the fact that the membership gives them the feeling that their rights are better protected, and they can resist possible pressure to sell; moreover, belonging to a community with similar ideas and behaviors may strengthen their choice not to sell. Thus, joining an association can, thus, reduce power asymmetry, which shapes the actual process of land grabbing, and it tends to give landowners the appearance of passive recipients [113]. Practically, as Fafchamps [114] puts it, one function that ethics can play in large-scale land acquisitions is that it can mediate conflictual relationships between asymmetric players and guide on what behaviors are acceptable and fair.
The path coefficient between "Perceived effect of foreign land grabbing on the natural environment" and "Perceived gravity of foreign land grabbing" is positive and significant. The highest the agreement that foreign land grabbing had negative consequences on the natural environment was, the strongest the "Resistance to sell" was, demonstrating that people cared about the state of the environment and that they considered themselves willing to accept a potential monetary loss from refusing to sell the land to protect the environment. Even if there is a well-known attitude-behavior gap [115], the fact that environmental damage of land grabbing has a significant effect on people's "Resistance to sell" is a good premise for strengthening this belief in the view of consolidating their decision to preserve ownership of their land.
The path coefficient between the "Perceived effect of agricultural land conversion to urban land" on "Resistance to sell" was positive and significant. The stronger the disagreement that the conversion of the agricultural land to urban was beneficial was, the higher landowners' resistance to sell their land was. Therefore, explaining to people the effects of this conversion and raising awareness of its negative consequences is a way to stimulate them not to sell. It is acknowledged that one of the main determinants of global change with visible consequences on human vulnerability and ecosystems is land-use change [116,117]. Urbanization-related land-use changes intensify competition between different land uses and threatens environmental sustainability and food security [118]. It influences rural people's livelihoods, and the loss of cropland can endanger a country's food security when agricultural production is mainly obtained through subsistence farming [28]. Thus, it can be inferred that land conversion to urban land through large-scale land acquisitions infringes the principle of moral economy that says that all individuals should get access to the resources necessary for their subsistence [119].
The lower the "Importance of the land price offered by the potential foreign buyer" was, the stronger the "Resistance to sell" was. This result can be attributed to the fact that protecting the land from being bought by foreigners is more critical for some landowners than the money they can earn by selling it, even if the amount offered is extensive. Another possibility stems from the land property situation in Romania. Land restitution in Romania after the fall of communism led to a highly fragmented private property, with over 90% of the total numbers of farms in Romania small farms of less than 0.5 ha (although they account only for less than 30% of the utilized agricultural area in Romania) [1]. The large share of small land plots and a low market price for land in Romania may determine landowners to think that their land does not value much. Thus, not selling the land to foreigners may worth more in their minds than the small amount of money they could gain from the land transaction. This finding illustrates that markets do not always function classically, and prices can sometimes be significantly different from the fundamentals of supply and demand would suggest [120]. Without getting into a debate about the just price, more should be investigated related to it within land transactions because, the just price, as a theory of ethics in economics, attempts to set standards of fairness in transactions.

Conclusions
The present study has revealed that land is a particular asset that must not be judged only as a mere commodity because it encompasses diverse functions, such as cultural, social, and environmental, that are difficult or impossible to embed correctly in a market value. This is one of the main reasons why land grabbing is a phenomenon with complex causes and effects, often challenging to understand and foreseen, especially by the local people whose perceptions can be easily biased by investors' promises to offer jobs, develop the infrastructure, and bring wellbeing to local people.
Given their role in land transactions, landowners should be knowledgeable in this respect and capable of making informed decisions regarding the consequences of selling their land. Objective and correct information-education campaigns, open discussions within community and between communities, creation of farmers' associations are several possibilities to open the way toward access to information.
Practically, this study extends the scant information on what factors determine landowners not to sell their land to foreigners to limit land grabbing. It showed that strengthening the determinants of joining a farmers' association, raising awareness on the importance of buyer attributes, increasing people perception of the negative effect of agricultural land conversion to urban land, of the negative effect of land grabbing on the environment, decreasing the relative importance of price received for land, and strengthening the image of the state as a necessary actor to limit land grabbing will increase landowners' resistance to sell their land to foreigners.
Summing up, ethics in land transactions should embed the "moral imagination" that represents the creation of alternative options that imply consultations and negotiations able to reframe land grabbing dilemmas and create economically viable and morally justifiable solutions [121]. However, ethics is not a matter of negotiation. Things should not be considered black and white but as a matter of sound judgment and willingness to talk with other people involved [122]. The results of the present analysis testify how several factors that predict landowners' "Resistance to sell" their land to foreign investors, even if they offer an attractive price to limit the land grabbing phenomenon, can reveal ways of transforming the landowners into active actors within the efforts of reducing land grabbing. This does not mean under any circumstances that the landowners are the only stakeholders who should bear the responsibility to limit land grabbing. It means that they are active participants in these transactions and, consequently, their role must not be ignored, and they should responsibly play their part. Land grabbing is a complex phenomenon that needs to be addressed from many perspectives and by many actors that would ultimately lead to a real alternative to this type of investment [106]. Beyond the ethical concerns mentioned here and the role of landowners, further steps are needed to address better foreign large-scale land transactions, both their positive and negative outcomes.
The authors of this study acknowledge the existence of several limitations. A larger representative sample should generate more reliable information on landowners' resistance to sell their land at the country level. Another is that more variables should be added to the model to increase its prediction power, such as individual health, land quality, and family labor. Future research should focus on extending the understanding of land grabbing gravity perception by finding out what factors or phenomena make people consider land grabbing more or less severe. It is also worthy of investigating the mechanisms that might link the Corporate Social Responsibility strategies of big land investors and poverty alleviation in host countries. Further attention should be paid to the values of public institutions that are involved in land transactions. Authors acknowledge that statistical tests have a limited power to predict reality, especially in the context of the study of human behavior, which is very complex and difficult to be encoded in variables and statistical models. In this context, it would be useful to dedicate efforts to using additional methods to predict landowners' resistance to sell their land and compare results. Thus, less sophisticated statistical tests (e.g., regression) could be run; other option can be to carry on simulations in the social science lab by involving landowners in various scenarios that simulate possible land transactions and use different values of the variables (e.g., several price levels, and different buyer attributes) to observe how landowners react in each context. Despite its limitations, this study has the potential to guide decision-makers in their efforts to stimulate landowners to preserve land ownerships in front of foreign buying offers when ethical aspects of transactions, which are relevant to landowners, are infringed.  Informed Consent Statement: Informed consent was obtained from all subjects involved in the study.

Data Availability Statement:
The data presented in this study are available on request from the corresponding author.

Conflicts of Interest:
The authors declare no conflict of interest.