Impediments to efficient land reallocation in agriculture: Multi‐agent simulation model of transaction costs and farm retirement

This study examines the effect of land exchanges among farm households (FHs) on agricultural sustainability and viability using a multi‐agent simulation model. The initial values of the model are fixed so that the simulation results might reproduce the sequence of land use pattern in Japanese agriculture. The novelty of this study is three‐fold. First, it quantitatively assesses the extent to which transaction costs incurred by lessors and lessees impede the sound development of land rental markets. Second, the simulation model considers inter‐regional heterogeneity in land use patterns and the interactive behavior of FHs regarding land exchanges. Third, the model elucidates the mechanisms by which the equilibrium land rent reaches zero and farmland is abandoned when some FHs retire from farming. Farmland abandonment has become a global phenomenon that has profoundly altered the traditional agricultural landscape in various countries. The present study suggests that the development of land rental markets associated with farm retirement alone does not strengthen agricultural production capacity at the national level. Particularly, if FH retirement progresses to the point where the equilibrium land rent drops to zero, farmland abandonment occurs, which threatens agricultural sustainability and viability. This study concludes that preventing farmland abandonment is a policy issue of the highest priority to ensure global food security and the sustainable development of agriculture.


| INTRODUCTION
Currently, there are more than 608 million farms worldwide, with the majority being family farms.Although there is no universally agreed-upon definition of family farms, FAO (2014) and Lowder et al. (2016) state that family farms are those managed by an individual, group of individuals, or household whose labor is primarily supplied by family members.Lowder   et al. (2021) estimate that family farms cultivate ca.70%-80% of the world's agricultural land and produce about 80% of the world's food in terms of value, although large-scale corporate farms have increasingly played a role in food production in recent years. 1 This implies that family farming continues to be the predominant mode of global farm production.
Nevertheless, family farms face a significant disadvantage in that their business continuity relies heavily on intra-family succession and inheritance, and therefore, their managerial sustainability is severely constrained by the demographics of members (Burton &   Fischer, 2015; Eastwood et al., 2010; Jansuwan & Zander, 2021;   Mishra et al., 2010; Rosenzweig & Wolpin, 1985).Accordingly, the aging of farm managers, coupled with the chronic shortage of successors, may force them to close their farming business (Duesberg   et al., 2017; Leonard et al., 2017).In addition to these demographic transformations, the general trend of young people aspirating for non-agricultural careers and migrating out of rural areas has also contributed to the overall reduction in agricultural production.From a macroscopic perspective, economic growth driven by non-farm sectors or an increase in the opportunity cost of farm labor accelerates farm retirement, which has become increasingly visible in economically high-performing countries (van Vliet et al., 2015).It poses a serious challenge to sustainable farmland use and agricultural development.
This study explores how transaction costs and farm retirement affect farmland use and agricultural sustainability using a multi-agent simulation model.In economics, transaction costs are generally defined as the various expenses associated with frictions in the exchange process (McMillan, 2002).An example of a transaction in agriculture is the usufruct exchanges of farmland between lessors and lessees in rental markets.If the transaction costs are too high and outweigh the potential benefits of the deal, the sound development of the land rental market may be impeded (Deininger & Jin, 2005; Ito,   Bao, et al., 2016; Ito, Nishikori, et al., 2016; Key et al., 2000; Ravallion & van de Walle, 2006; Skoufias, 1995). 2 The author firmly believes that transaction costs and farm retirement are two key factors that determine the economic performance of land rental markets and agricultural sustainability.However, to the best of my knowledge, few attempts have been made to provide theoretical insights into this subject.
The simulation model used in this study is uniquely equipped to address the two important issues by incorporating three key characteristics of agricultural land use.First, given that family farming remains the predominant mode of global crop production, it is assumed that land exchanges only occur between farm households (FHs).Second, land is a physically immobile factor of production, and therefore, land rental markets are locally segmented and not integrated with neighboring markets.This situation hinders inter-regional convergence of land rent (rental price) and diversifies the land use pattern extensively.Third, retired FHs (RFHs) play a crucial role in the land rental market as potential lessors.However, if the land demand of working FHs is insufficient to absorb all land tracts supplied by RFHs, some tracts may be abandoned and left uncultivated. 3rmland abandonment associated with farm retirement has become a global phenomenon, thus significantly transforming the traditional agricultural landscape in various countries (Ito et al., 2019;   Li & Li, 2017; Plieninger et al., 2016; Terres et al., 2015; Xu   et al., 2019; Yan et al., 2016; Zavalloni et al., 2021).With the declining profitability of farming, a reduction in land use intensity has been widely observed not only in hilly and mountainous regions but also in areas with favorable agricultural conditions.The widespread occurrence of farmland abandonment implies land degradation in agricultural production and highlights the inefficiency of land use (Quaranta   et al., 2020).
The remainder of this paper is organized as follows.Section 2 presents models of land rental markets that focus on transaction costs and FH retirement.Section 3 provides the simulation results, while Section 4 discusses the policy implications.Section 5 concludes with a brief summary of the study.

| A basic model
In this section, I model the farmland rental market equilibrium based on FHs' optimization behavior, and theoretically examine the impact of transaction costs and farm retirement on land transactions.Based on the axiom of FHs' profit maximization, we can derive a land demand function as For the sake of simplicity, this study assumes the following land demand function: where da k =dp ≥ 0 and . Although S k is specified ad-hoc, it reflects the reality that FHs do not have unlimited demand for land; depending on the values of p and r, S k may become zero.The land rental market is in equilibrium when the sum of the excess demand for all FHs equals zero.Owing to the uniform distribution of a k p ð Þ, all FHs except k ¼ N=2 participate in land transactions without retiring from farming when the following equation is met: where r Ã and S Ã 0 represent the equilibrium land rent and the corresponding demand for land of FH 0, respectively.
As land is exchanged only within a specific area, the rental market is not integrated with neighboring markets, which prevents the convergence of land rent and diversifies land use patterns among local land markets (communities).To incorporate inter-regional heterogeneity into the model, I modify Equation (1) as follows: where b is a coefficient that represents community heterogeneity.To reflect ∂S k =∂X k > 0, we assume that dk=dX k > 0 and b > 0. These two assumptions suggest that households in each community are ranked by the amount of fixed capital, and communities in which a significant amount of fixed capital has been accumulated have a larger value of b.
To simplify the market equilibrium condition, this study makes two strong assumptions: a uniform distribution of a k p ð Þ and a linear demand function with respect to land rent and bk.However, the author is confident that these assumptions do not significantly affect the core conclusion of this study.
As half the FHs are lessors or lessees in a community, the rented land area (RA) is given by The equilibrium land rent and RA are determined by r Ã ¼ a þ 0:5bN À S and RA ¼ bN 2 =8, respectively.Additionally, Equation (2) can be expressed as bN À 2S ≤ 0: If the agricultural product is the numeraire (p ¼ 1), the farm revenues before and after land exchanges are, respectively, given by See Appendix A for the derivation of the farm revenue equations.
Thus, we obtain indicating that the participation of FHs in land rental markets makes potential lessors and lessees wealthier.
2.2 | Shrinkage of rental markets due to transaction costs Deininger and Jin (2005) and Key et al. (2000) have shown that an increase in transaction costs drives a wedge between potential lessors and lessees, which expands the range of producers who remain in autarky. 4The simulation model in this study assumes that FHs who participate in land exchanges incur a transaction cost per RA of T.
Therefore, the lessors' and lessees' land demand functions can be expressed as As FHs with k 1 ≤ k ≤ k 2 find it unprofitable to participate in the rental market, they cultivate only their own land.Refer to Appendix A for more details on this issue.The values of k 1 and k 2 are determined this way so that the excess demand for land equals zero.
The market equilibrium condition is given by From Equation ( 6), the RA for lessors and lessees is, respectively, given by Given the initial values of a, b, N, S, and T, we can solve the system of Equations ( 7) and ( 8) numerically and obtain the values of k 1 and k 2 , using the generalized reduced gradient non-linear method developed by Lasdon et al. (1978).

| Farm retirement and farmland abandonment
With the increase in the opportunity cost of farm labor and the decreased demand for land, FHs stop farming.This mechanism is discussed below.I assume that once FHs retire, they supply their land (S) to the local rental market and do not return to farming.
In land rental markets, RFHs (the number of which is t) and FHs numbered t through x become lessors (t < x), whereas FHs numbered x through N become lessees.Thus, the market equilibrium condition yields Furthermore, the excess demand for land of FH x equals zero (ED x ¼ 0), which yields Consequently, the total RA in this community is given by Equations (9-11).
With an increase in RFHs, t ¼ x.Substituting t ¼ x into Equation ( 10), we have When t ≥ t Ã , the land is only exchanged between RFHs and lessees whose farm numbers run from t through N. In other words, no FHs rent out a portion of their land while engaging in agriculture on the remaining land.In this scenario, the total RA is obtained by replacing x in Equation ( 11) with t, which is equal to the total land area owned by RFHs (RA R ¼ tS).
A further increase in farm retirement creates a situation in which the total excess demand for land in the market becomes negative, which reduces the equilibrium land rent to zero.By substituting r Ã R ¼ 0 into Equation ( 9), we obtain We always have t þ > t Ã .However, once r Ã R ¼ 0, the farmland supplied additionally by RFHs to the rental market is not used for cultivation and is therefore abandoned.As the total demand for land in the abandoned farmland is equal to the total land supply minus D r¼0 , which is given by Meanwhile, the total RA is represented by D r¼0 minus the land area owned by FHs continuing production, which is given by From Equations ( 14) and ( 15), we obtain AF þ RA R¼0 ¼ tS, meaning that a portion of the land supplied by RFHs cannot be used for cultivation.
Before FHs exchange their land, the total farm output in a community (G b ) is given by Meanwhile, the total output after they exchange their land (G a ) is given by As a corollary of Equation ( 5), we have This study assumes that the number of RFHs (t) is endogenously determined, based on the following equation: where R a t denotes the farm revenue of FH t, and w and L denote the opportunity cost of farming (wage rate) and labor force endowment for each FH, respectively.I assume that every FH has the same amount of labor endowment of L ¼ 1.As FHs with a number less than t have lower farm revenue than FHs with the number t, they will have already retired from farming when FHs with the number t decide to retire.This study merely hypothesizes that the rising opportunity costs of farm labor promote farm retirement, although several factors may be involved, as discussed in Section 1.
The farm revenues depend on S Ã k and r Ã R , whereas Equation (3) implies that S Ã t depends on r Ã R and the number of RFHs.Furthermore, Equation ( 9) indicates that r Ã R depends on the number of RFHs (t).It is natural to assume that FHs regard the wage rate as given, but they do not see the equilibrium land rent in the same way because it depends on their own rental behavior.Thus, R a t , r Ã R , S Ã t , and t are determined by Equations ( 3), ( 4), (9), and (18) simultaneously.Given the initial values of a, b, N, L, and S, we can numerically solve the system with the same method adopted in Section 2.2.

| Initial values
The simulation analysis in this study assumes that b ¼ 0:0005, 0:02 ½ , with 0.0005 increments between the values, implying that there are 40 communities in a county.A regulatory condition of ∂R b k =∂S ≥ 0 for all k yields a ≥ S. The initial values are fixed throughout this study, set at a ¼ 3, S ¼ 2, and N ¼ 200.The wage rate (w) is assumed to range from 4 to 9.These values were decided this way so that the simulation results might reproduce the sequence of land use pattern and land rent in Japanese agriculture.There are two reasons for focusing on Japan.First, farm retirement in the country has accelerated over the past several decades owing to the increased opportunity costs of farm labor.Second, to the best of my knowledge, Japan is the only country that officially reports detailed statistical data on agricultural land use at the national level, including the area of abandoned farmland.However, as the model does not encompass all relevant issues, a caveat must be employed before generalizing the simulation results in this study to the real world.

| Simulation results of a basic model
Table 1 presents the equilibriums of land rental markets for 40 communities with no farm retirement.The second column denotes the value of bN À 2S, indicating that Equation (2 0 ) is met for all 40 communities.The first and third columns indicate that as b increases, the equilibrium land rent also increases.The fourth and fifth columns display the values of RA and land rental ratio, respectively.As the value of b increases, so do the equilibrium land rent and the rental ratio.This result suggests that land is more actively exchanged among FHs when there is relatively high demand for land in the local market (community), leading to a higher equilibrium land rent.The land rental ratio of the community with b ¼ 0:02 is among the highest at 25%.The analysis in Appendix B proves that when there is no farm retirement and a k p ð Þ in Equation ( 1) is uniformly distributed, the maximum value of the land rental ratio is 25%.The average equilibrium rent across the country is 2.03, and the total RA is 2050 (the bottom of Table 1).Thus, the average land rental ratio, defined as the total RA divided by the total cultivated land area, equals 2050= 40SN À Á ¼ 0:128 (12.8%).
In the absence of transaction costs, FHs with 0 ≤ k < 100 become lessors, earning rental revenues, whereas FHs with 100 < k ≤ 200 become lessees, incurring rental costs.Figure 1 shows the incremented farm revenues from land exchanges for all FHs in the communities with b = 0.005, 0.01, 0.015, and 0.02. 5 There is no revenue increment for an FH with k ¼ N=2 ¼ 100 because it does not participate in the land market.Figure 1 illustrates that the incremented revenues are non-negative for all FHs, implying that land exchange among FHs makes both lessors and lessees wealthier, consistent with the theoretical prediction.In addition, the larger the inter-household difference in land demand within a community, the greater the incremented revenues, which is in line with intuitive understanding.Although it is not shown here, farm production at the aggregate (national) level in this model increases by 3.8% if all FHs except k ¼ 100 participate in the land rental market.

| Simulation results of transaction costs
Substituting it into Equation ( 7), we have The effect of transaction costs on land rentals.The equilibrium land rent for b = 0.005, 0.010, 0.015, and 0.020 equals 1.5, 2.0, 2.5, and 3.0, respectively (see Table 1).[Colour figure can be viewed at wileyonlinelibrary.com] that the equilibrium land rent is independent of the transaction costs. 6gure 2 reveals that the larger the value of T, the lower the land rental ratio, thus suggesting that the transaction costs cause shrinkage of the land rental market.In an extreme case, a transaction cost of T ¼ 2:0 causes the land rental market to become dysfunctional, and results in zero land exchange even in the community with b ¼ 0:020 (a relatively high demand for land).Although it is not shown here, the number of FHs that do not participate in the rental market increases as T rises.10), equaling 88.9.

|
Thus, x À t=88.9-48.1 = 40.8FHs rent out a portion their own land while engaging in agriculture on the rest of the land, whereas N À x=200-88.9= 111.1 FHs rent land from RFHs and other FHs.
The land supplied by RFHs and FHs amounts to 96.1 and 15.0, respectively, which equals lessees' demand for rented land.
The second and fourth columns of Table 2 reveal that t ≥ t þ is met for 0:0005 ≤ b ≤ 0:005; in this range, the equilibrium rent is zero and farmland abandonment emerges.For communities with b ≥ 0:005, we have dr Ã R =dt ≤ 0, meaning that the equilibrium land rent is decreasing with the number of RFHs.The equilibrium rent averages 1.16 (the bottom of Table 2), which is smaller than the equilibrium rent shown in Table 1.This is because RFHs as a whole supply a large tract of land to the rental market.For the communities with 0:0005 ≤ b ≤ 0:002, all FHs retire from farming, and therefore, all land in these communities is abandoned.Meanwhile, for the communities with 0:0025 ≤ b ≤ 0:005, t ≥ t þ is met, and therefore, the equilibrium rent is zero.However, the number of RFHs is less than 200, thus leading to an abandonment ratio of less than 100%.A positive farmland abandonment ratio indicates excess supply over demand for land, which naturally reduces the equilibrium land equal to zero.In other words, as long as the equilibrium rent is positive, there is no farmland abandonment in the communities.
The rental ratio in this simulation is defined as the RA divided by the cultivated land area.Unlike in Table 1, the rental ratio is negatively associated with the equilibrium rent.The last column of Table 2 denotes that the average farm size, defined as the cultivated area divided by the number of FHs, records a maximum of 3.73 for the community with b ¼ 0:005, in which the rental ratio reaches a maximum value of 46.4%.By contrast, the average farm size for the community with b ¼ 0:02 is among the lowest at 2.00, in which the rental ratio has the minimum value of 25% and there is no farm retirement.
In communities with higher b, few FH retire from farming, and therefore, the equilibrium rent is high and rental ratio is low.

| Effect of wage growth
This subsection presents the simulation results regarding the effects of wage growth on macroscopic variables.When w ¼ 4, R a t > wL for all FHs in the entire range of b.Likewise, when w = 4.5, 5.0, and 5.5, R a t > wL for the range of 0:0105 ≤ b ≤ 0:020, 0:0145 ≤ b ≤ 0:020, and 0:0175 ≤ b ≤ 0:020, respectively.Nevertheless, this study assumes that labor force does not move from the non-farm to the farm sector. Figure 4 illustrates the statistical data on farmland use and land rent in Japanese agriculture during the 1985-2015 period.In Japan, the opportunity cost of farm labor (wage rate) rose linearly 1.8 times during this period. 8Figure 4 shows that the real land rent has declined for the past several decades, reflecting a consistent decrease in nationwide demand for agricultural land relative to supply.The land rental ratio increased from 7.0% in 1985 to 25.9% in 2015.Although not illustrated here, this rise has been accompanied by land usufruct accumulation in favor of large-scale farms (Ito, Nishikori, et al., 2016).
The farmland abandonment ratio has also tended upward for the past four decades, increasing from 2.5% to 9.4% in 2015. 9Figures 3 and 4 demonstrate that the model predictions are consistent with the past reality of Japanese agriculture, both indicating a parallel rise in the ratios of land rental and farmland abandonment and a steep decline in the land rent.
Figure 5 reveals that, as the wage rate rises, the percentage of RFHs in the total FHs and the average farm size move in the same direction, implying that farm retirement is essential for increasing the average farm size.During the 1985-2015 period, the number of Japanese FHs decreased from 4.95 to 2.16 million, while the average farm size increased from 0.97 ha to 1.42 ha.Evidently, there is no serious contradiction between the simulated results and reality.We can compute the quantity of pre-and post-rental outputs in each community from Equations ( 16) and ( 17).The total outputs at the national level are obtained by aggregating them for 40 communities.
T A B L E 2 The equilibriums of land rental markets when some farm households retire (w ¼ 6).12) and ( 13), respectively.Abbreviation: RFHs, retired farm households.
Figure 5 illustrates that, with a rise in the wage rate, the ratio of postrental total output to pre-rental total output rises, suggesting an enhanced contribution of land exchanges toward farm production increases.However, both the pre-and post-rental total outputs decrease as the wage rate rises.To compensate for the loss of total farm output due to FH retirement, the demand for land of FHs that continue cultivation must be increased.It is the final issue addressed in Section 3.4.3.

| Capital reinforcement to maintain farm production quantity
FHs that increased their farm size by renting land from RFHs are likely to invest in capital equipment if the commensurate returns are expected.I consider how much b needs to be increased after a certain number of FHs retire from farming so that the total postrental farm output remains at the same level as that when there is  total farm output will be maintained, as it depends on expected returns to farm investments.

| DISCUSSION
Scholars agree that family farms have efficiency advantages over non-FH producers (NFHPs) such as agro-industrial corporations and collective farms.Large-scale NFHPs incur additional costs due to the need to supervise hired labor working in widely dispersed agricultural environments where quality of work is hard to monitor; a lack of such monitoring would undoubtedly lead to poor crop production performance (Eswaran & Kotwal, 1986; Feder, 1985; Otsuka et al., 2016).
By contrast, family members, as residual claimants on farm income, have a strong motivation to work diligently on their own farms, which provides them with a competitive edge in production efficiency (Allen & Lueck, 1998).NFHPs, which increasingly enter farm businesses as cultivators and contribute to the development of land rental markets in certain countries and regions (Giller et al., 2021).It is also important in this context to note that the largest 1% of farms in the world (those larger than 50 ha) currently operate on more than 70% of the world's farmland, and that farmland is becoming increasingly concentrated among such large farms (Lowder et al., 2021).When NFHPs and/or large-scale farms with much more capital-intensive technology than small-scale farms join land rental markets as lessees, the value of b in Equation ( 3 agriculture, claim that the model's prediction cannot be aggregated linearly when heterogeneity and interactions are significant.Other simulation models address the issues of land use dynamics, farmland preservation, and agro-ecological resource uses (Bai et al., 2015; He  et al., 2013; Schreinemachers & Berger, 2011; van Vliet et al., 2016).However, much remains to be studied regarding the effect of FH retirement on land use dynamics and agricultural production. 4As transactions costs, as opportunity costs, are difficult to measure directly, many empirical studies use a proxy variable to examine the effect of transaction costs on land transactions (Deininger & Jin, 2005;  Huy et al., 2016; Kung & Bai, 2011; Li & Ito, 2021; Ricker-Gilbert &  Chamberlin, 2018).Léger-Bosch ( 2019) is an exceptional study that estimated the transaction costs directly.
5 Equation ( 5) can be used to compute the incremented farm revenues.
6 It is due to the assumption that lessors and lessees incur the same transaction cost per rented land area. 7The simulation in this study considers uncultivated land as abandoned without distinguishing between farmland abandonment and the conversion of agricultural land to non-farm uses.
8 The wage rate is measured by the monthly per-capita real income of 30 regular establishments (Monthly Labor Survey, Ministry of Health, Labor, and Welfare) deflated by the agricultural product price index.The nominal wage rate has declined slightly since 2000, owing to prolonged Japanese economic stagnation.However, during this period, the agricultural product price declined more significantly than the nominal wage rate, which led to a linear increase in the real wage rate during the period 1985-2015.
How to cite this article: Ito, J.The land demand function given by Equation ( 3) is derived from the first-order condition of the profit maximization: ∂Q=∂S ¼ r=p, where Q denotes agricultural output.Regarding agricultural product as the numeraire (p ¼ 1), we have the following equation: When there are transaction costs involved in land exchanges, the farm revenues of the lessor and lessee are, respectively, given by represent the rental revenue for the lessors and the rental cost for the lessees, respectively.From the first-order condition of the revenue maximization, we can derive the demand function of land as Lessors (S k ≤ S) do not participate in the rental market if the following equation is met: As an FH's demand for land in this situation equals S, the equation above can be rewritten as Similarly, lessees (S k ≥ S) do not participate in the rental market if the following equation is met As an FH's demand for land in this situation equals S, the equation above can be rewritten as where k, ranging from zero to N, represents the FH number, and p, r, and X k denote the agricultural product price, land rent, and fixed capital of FH k, respectively.From the regulatory conditions of the factor demand function, we have ∂S k =∂p > 0 and ∂S k =∂r < 0. Without loss of generality, we assume that ∂S k =∂X k > 0. If all FHs own the same area of farmland, S, then the excess demand for land can be expressed as ED k ¼ S k À S; ED k < 0 for lessors and ED k > 0 for lessees.

Figure 2
Figure 2 illustrates the simulation results regarding the impact of transaction costs on the land rental ratio, where the value of coefficient b represents community heterogeneity.There is a high demand for land in a community with a higher value of b.The simulation assumes that the transaction cost per RA (T) ranges from 0 to 2.0.The results confirm the establishment of Equation (8), where the RA of lessors and lessees are equal.The equilibrium values of k 1 and k 2 are endogenously determined in the model, satisfying k 1 þ k 2 ¼ N.

F
I G U R E 1 Incremented agricultural revenues from land exchanges.[Colour figure can be viewed at wileyonlinelibrary.com]

Figure 3
Figure 3 illustrates how the average equilibrium land rent and land use patterns at the aggregate (national) level change as the wage rate increases.The equilibrium land rent consistently decreases for the entire range of the wage rate because a rise in the wage rate is accompanied by farm retirement, which increases the land supply to the rental market.The land rental ratio increases from 12.8% for w ¼ 4 to 55.3% for w ¼ 9, although the RA peaks when w ¼ 7:5 and decreases thereafter.Meanwhile, the farmland abandonment ratio remains at zero for w ¼ 4 and w ¼ 4:5, and rises gradually thereafter, reaching 39.7% for w ¼ 9. 7 Simulation results of land use and land [Colour figure can be viewed F I G U R E 4 use patterns and the land rent in Japanese agriculture.The land rent at constant prices measures the average nominal rental price deflated by the agricultural product price indexes.Source: Census of Agriculture and Forestry (MAFF), Crop Statistics (MAFF), and Survey of Farm Production Costs (MAFF).MAFF, The Ministry of Agriculture, Forestry and Fisheries.[Colour figure can be viewed at wileyonlinelibrary.com] no farm retirement (w ¼ 4:0).Equation (3) indicates that a rise in b increases the demand for land in each community, which helps to boost aggregate farm production.The number of RFHs is assumed to have been fixed at the level obtained in the analysis of Section 3.4.2.

Figure 6
Figure 6 displays the simulation results."Ratio" represents the expansion rate of b needed to sustain the total farm output compared with the base case of b ¼ 0:0005, 0:02 ½ .As the wage rate rises However, ongoing demographic transformations and changes in young peoples' career aspirations pose serious challenges to sustainable farmland use and family-based agricultural development.Furthermore, the simulation analysis in the present study indicates that FHs that continue farming cannot absorb all tracts of land supplied by RFHs and therefore fail to maintain farm production at the national level if their demand for land is limited.It raises concerns about who will step in to prevent further farmland abandonment and shrinkage in agricultural production.Although this is beyond the scope of the present study, it is worthwhile to observe the embryonic development of ) increases, which could mitigate the reduction of agricultural production, as discussed in Section 3.4.3.TheFAO (2014) stresses the significant role of family farming in ensuring global food security, alleviating rural poverty, and promoting environmental sustainability.Meanwhile, van Vliet et al. (2015) discuss the dubiousness surrounding whether a specific focus on family farming is necessary or helpful to attain the above-mentioned FAO's objectives and conclude that policies need to address the long-term economic viability of farming in all its shapes and forms.Graeub et al.   (2016) claim that the contribution of family farmers to food security and local and regional development is surprisingly poorly documented.Similarly,Galli et al. (2020) argue that the contribution of small farms to food and nutrition security in the European context is underexplored or assumed to be of limited relevance at local, regional, and global scales.As evidenced by these discussions, further research efforts are needed to examine which kinds of farm types in which regions play a leading role for efficient land use and agricultural sustainable development in the future(Garcia-Alvarez-Coque   et al., 2021; Kessler et al., 2021).5 | CONCLUSIONWell-functioning land markets allow competent farm producers to access to additional land.It not only benefits the lessors and lessees involved in land transactions but also enhances the overall efficiency of land use and agricultural productivity.The simulation results in this study lend strong support to this assertion; post-rental total output exceeds pre-rental total output, suggesting that the development of land rental markets brings a significant benefit to FHs and total agricultural production.In addition, the simulation analysis in this study elucidates that transaction costs in land rental markets adversely affect efficient land use by reducing the number of FHs who participate in the markets.Given that land reallocation among farm producers is achieved through the functioning of market mechanisms, neither governments nor external agents need to be directly involved in such transactions.However, if increased transaction costs of land exchanges outweigh the potential benefits of the deal, the sound development of the land rental market may be impeded.In such cases, Institutional support to reduce transaction costs are of particular importance to achieve efficient land reallocation and increased agricultural production.This study provides clear evidence that a certain amount of farm retirement is necessary to promote land exchanges among FHs.Without farm retirement, the supply of land necessary for the development of rental markets is limited.However, the development of rental market associated with farm retirement alone does not strengthen agricultural production capacity at the national level.When FH retirement progresses to the point where the equilibrium land rent drops to zero, farmland abandonment occurs, which can threaten agricultural sustainability and viability(Li et al., 2018).It reflects what can happen in any country where family farms are major cultivators and their demand for land is limited.Widespread farm retirement and farmland abandonment are common features in many countries and regions, which raise concerns about efficient land use and agricultural sustainability(Duesberg et al., 2017).It follows that an agricultural policy issue of the highest priority is to mobilize concerted efforts to prevent farmland abandonment for ensuring global food security and the sustainable development of agriculture.An entry of large-scale agribusiness firms into farming business as direct cultivators may lead to an increase in the total demand for land and help solve the critical issue of farmland abandonment.
(2024).Impediments to efficient land reallocation in agriculture: Multi-agent simulation model of transaction costs and farm retirement.Land Degradation & Development, 35(4), 1553-1566.https://doi.org/10.1002/ldr.5005APPENDIX A suggests that FH k does not participate in the rental market as a lessor.
Simulation results of farm retirement 3.4.1 | The model solution When w ¼ 4, there is no farm retirement, which makes the market equilibriums identical to those represented by Table 1.An increase in the wage rate above w ¼ 4 triggers farm retirement from FHs with smaller values of k.Table 2 presents the model solutions for w ¼ 6.Similar to Table 1, Table 2 also captures the regional heterogeneity of land rental markets and land use patterns well.
The second column shows the number of RFHs.The community with a larger value of b has a lower number of RFHs, suggesting that farm retirement is less likely to occur in communities with high demand for land.The bottom of the second column indicates that the number of RFHs averages 91.4 or 45.7% of the total FHs.The second and third columns show that t ≥ t Ã is met for 0:0005 ≤ b ≤ 0:015; in this range, land is exchanged only between RFHs and FHs continuing farm production.In the community with b ¼ 0:01, for example, 82.8 of 200 FHs retire from farming, and therefore, 117.2 FHs rent land from the RFHs.The RA in this community equals 82:8 Â S ¼ 165:7.Meanwhile, for the communities with b ¼ 0:018, for example, t < t Ã is met, and 48.1 of 200 FHs retire, whereas 200-48.1 = 151.9FHs continue cultivation.In this situation, x is given by Equation (