Agricultural Systems 161 (2018) 81–88

Contents lists available at ScienceDirect

Agricultural Systems
journal homepage: www.elsevier.com/locate/agsy

Assessing the impacts of land fragmentation and plot size on yields and T
costs: A translog production model and cost function approach
⁎
Hua Lua, Hualin Xiea,b, , Yafen Hea, Zhilong Wua, Xinmin Zhangc
a
Institute of Ecological Civilization, Jiangxi University of Finance and Economics, Nanchang 330013, China
b
Co-Innovation Center of Institutional Construction for Jiangxi Eco-Civilization, Jiangxi University of Finance and Economics, Nanchang 330013, China
c
Graduate School of Life and Environmental Sciences, University of Tsukuba, 1-1-1 Tennodai, Tsukuba, Ibaraki 305-8572, Japan

A R T I C L E I N F O A B S T R A C T

Keywords: More attentions should be focused on the changes in plot size of each household rather than the size of farmland
Land fragmentation in the discussions of economic problem of land fragmentation in China. This study empirically analyzes the
Economies of scale impact of land fragmentation and plot size on yields, along with average costs, using household survey data
Plot size collected from the Jiangsu province in China. A detailed and careful translog production model and cost function
Yields
are employed to understand and analyze these problems. The empirical results reveal that there are increasing
Costs
returns to scale in agricultural production. Land fragmentation reduces yields through changes in marginal
China
outputs of agricultural inputs. Especially in areas with high opportunity costs of labor, the negative impact is
more obvious. A one-unit increase in the Simpson index leads to a 39% increase in the average cost, whereas a
one-unit increase in plot size leads to an 8% decline in the average cost. Thus, moderate expansion of the size of
the plot can reduce the average cost, implying that agriculture can achieve economies of scale within each plot.
Economies of scale should be developed by keeping farm size constant, reducing the number of plots, and
expanding the size of each plot. We suggest that economies of scale can be achieved in each plot by either land
consolidation or land transfer as well as by joint farming and joint association.

1. Introduction 30.682 million ha were circulated, which was approximately one-third

Research on agricultural development in China has increasingly
focused on farm size. People generally think of the scale of agricultural
operations in terms of the farms' size, especially in the context of land
management. The Household Responsibility System (HRS) im-
plemented in China contributed to the rapid development of agriculture
by increasing farmers' incomes and narrowing the urban-rural divide.
However, its implementation also led to each farmer separately pos-
sessing numerous small plots of lands. In 2013, the farm size in China
was 0.66 ha, with an average plot size of 5.1 ha per household, each
with an area of 0.129 ha.1 For the increase in farm size, some policies
do encourage the circulation of agricultural land management rights.
Examples of these policies include the land circulation pilot in 1978; the
legal promotion of the orderly circulation of land management rights in
the central NO.1 document in 2016; and, actively encouraging farmers,
through village organizations, to voluntarily exchange land in order to
achieve contiguous land cultivation in the central NO.1 document in
2017. However, the results of these policies were not optimal. As of the
end of June 2016, of the 0.087 billion ha of contracted land, only
of the total area of contracted land. Furthermore, of the 234 million
Chinese farmers holding contracted land, nearly 170 million have not
circulated their land; only around 66 million have partially or com-
pletely circulated their land (Xie and Lu, 2017).
Land fragmentation means that a household's land resources are
divided among several spatially separated plots (Mcpherson, 1982).
However, it does not refer to a lack of economies of plot size. Although
farmers do have numerous dispersed plots, if the size of each plot ac-
counts for, say, hundreds of hectares, the economic problem of land
fragmentation would be harder to observe. With the development of
agricultural outsourcing services and other forms of scale operations,
economies of scale have moved from internal economies at the house-
hold level to external economies across households. The realization of
economies of scale is no longer entirely dependent on expanding farm
size; the importance of scale economy at the plot level has increased
further. The economic problem of land fragmentation should be studied
from the perspective of plot size rather than that of farm size.
In fact, the extant literature often focuses on the expansion of farm
size, which neglects the size of each plot. More importantly, the

⁎
Corresponding author at: Institute of Ecological Civilization, Jiangxi University of Finance and Economics, Nanchang 330013, China.
E-mail address: landuse2008@126.com (H. Xie).
1
The data came from the fixed observation points of the Ministry of Agriculture of the People's Republic of China (MOA).

https://doi.org/10.1016/j.agsy.2018.01.001
Received 24 September 2017; Received in revised form 3 December 2017; Accepted 2 January 2018
Available online 11 January 2018
0308-521X/ © 2018 Elsevier Ltd. All rights reserved.
H. Lu et al.
measurement of changes in farm size should consider two aspects. First,
the area of each plot may be constant, but the number of plots changes.
Second, the number of plots may be constant, but the area of each plot
changes. The average cost of some inputs is irrelevant to the number of
plots but is related to the area of each plot. In addition, currently, land
transfer is still inefficient—a long exchange chain results in a low
probability of matching with adjacent areas and high transaction costs.
Due to the high transaction costs and the difficulty of effective land
transfer, farm size enlargement is likely to increase the number of plots
rather than the area of each plot. Therefore, it is more reasonable to
study economies of scale by focusing on the area of each plot.
In regions with large populations, relatively less arable land, and
surplus labor, land fragmentation is a rational choice for farmers to
maximize their income (Xu et al., 2008). Varied crop planting and
harvesting seasons can effectively offset labor insufficiency during busy
seasons and surplus labor during slack seasons, which allows a flexible
allocation of agricultural inputs (Blarel et al., 1992). With problems like
labor shortages and natural disasters or droughts, planting diversity can
effectively decrease market risks, increase yields, or narrow the in-
equality of income (Falco et al., 2010; Foster and Rosenzweig, 2011; Li
and Li, 2017; Niroula and Thapa, 2005; Paul and Githinji, 2017). Ad-
ditionally, economic development or transition in many major agri-
cultural regions worldwide, especially Asia, are experiencing drastic
non-agricultural utilization (Long et al., 2009; Sreeja et al., 2015). You
(2017) analyzed the impact of economic transition on agricultural
landscape dynamics (ALD) and revealed that the magnitude of ALD is
larger in non-urban planning zones. In particular, agricultural land-
scapes are changing into fragmented, irregular, decreased, and isolated
patterns at a faster pace. The analysis also found that the efficiency of
agricultural land protection in the non-urban planning zone was much
lower than in the urban planning zone, and the absence of a land use
master plan was a critical contributor to the low efficiency of agri-
cultural land protection in the non-urban planning zone.
With the increase in non-agricultural labor supply, the negative
effects of land fragmentation on agricultural production or on the
ecology and environment have gradually become more evident (Cai,
2008). The rising labor cost led to the inevitability of mechanized al-
ternative labor, and land use changed. Su et al. (2016) revealed that
households with low agricultural labor intensities have a high prob-
ability of growing tea and mulberry plantations. Xiao et al. (2015) also
found a consistently higher probability of cash crop expansion in places
with abundant farmland, and the distance to a counter and to a pro-
vincial road were decisive determinants for farmers' choice of cash crop
plantation. Furthermore, small plots that require extensive labor but are
unsuitable for mechanical operations are abandoned or used in-
efficiently (Carter and Yao, 2002). Under these circumstances, land
fragmentation could reduce both land and labor productivity, thus af-
fecting agricultural production (Ali and Deininger, 2014; Barrett et al.,
2010; Deininger et al., 2012; Jia and Martin, 2014; Kalantari and
Abdollahzadeh, 2008; Schultz, 1953; Tan et al., 2010). Wan and Cheng
(2001), in their study on corn, late rice, and wheat, noted drops in
technical efficiency by 4%, 15%, and 17%, respectively, when the de-
gree of land fragmentation increases by one unit. Small and dispersed
plots waste more resources and time when transferred to different plots.
Furthermore, small plots are an obstacle to the application of agri-
cultural machinery and the construction of farmland infrastructure.
They can also affect productivity, decrease overall grain production
capability, and simultaneously increase input costs (Haji, 2007; Latruffe
and Piet, 2014; Rahman and Rahman, 2008; Sklenicka et al., 2014; Tan
et al., 2008). Latruffe and Piet (2014) found that land fragmentation
increased cost but decreased income, profit, and efficiency. Wei (2015)
further found that land fragmentation could influence the application of
agricultural technology, forcing farmers to disburse more costs in terms
of labor, time, and psychological costs.
To summarize, the effects of land fragmentation on agricultural
production are different under varying assumed conditions. In cases of
82
Agricultural Systems 161 (2018) 81–88
low costs, such as labor costs, land fragmentation is reasonable.
However, an increase in agricultural costs decreases agricultural profit,
forcing farmers with small plot sizes to desert their farmland. Besides,
moderate farm size has become an inevitable trend in agriculture with
rising costs and non-agricultural labor supply. However, here, moderate
farm size refers to plot size enlargement, not an increase in the number
of plots. Only a few studies measure economies of scale for each plot,
but they lack empirical analysis. Therefore, we explore the relationships
among land fragmentation, plot size, and economic benefit from the
perspective of the average area of each plot.
2. Materials and method
2.1. Research hypothesis
Land fragmentation affects yields by influencing the allocation of
other agricultural inputs. Small and dispersed plots not only increase
the time cost but also evaporate agricultural inputs during plot trans-
fers. Moreover, these plots reduce the efficiency of fixed assets, which
are indivisible in agriculture. Due to increased boundaries and ridges
between small and dispersed plots, irrigation efficiency falls and agri-
cultural operation time is wasted, leading to poor field management.
Small and dispersed plots also limit the use of machines and new
technologies. Thus, we hypothesize:
H1. Land fragmentation reduces yields by affecting the efficiency of
agricultural inputs.
The impact of land fragmentation on yields varies in different re-
gions because the opportunity cost of farm labor varies. Based on
maximizing benefits, farmers will choose to acquire more off-farm in-
come in an area with higher opportunity costs of farm labor. Small and
dispersed plots reduce farmers' enthusiasm for agriculture and increase
the probability of extensive management. To meet basic family needs,
farmers will not fully abandon their land, even with uneconomical
scales in small plots. Furthermore, farmers will neglect increasing yields
due to time and energy constraints. However, agricultural income may
be a major source of income in an area with lower and less stable op-
portunity costs of farm labor. Land fragmentation also provides more
opportunities to grow diversified crops. Thus, we hypothesize:
H2. The higher the opportunity cost of farm labor, the greater the
negative impact on yields.
Land interchange or integration may hedge the negative effects of
land fragmentation and reduce average cost. Even though the number
of plots does not change, plot size enlargement will bring economies of
scale. Before the analysis, we identify economies of and returns to scale.
Samuelson and Nordhaus (1948) defined economies of scale as in-
creased agricultural productivity or decreased average cost caused by
the same proportional increase of all agricultural inputs. Mankiw
(1998) concluded that economies of scale are defined by a long-term
decrease in average cost along with increased output. Thus, there are
similarities and differences between economies of and returns to scale.
First, the conditions of the latter are stricter compared to those of the
former. Returns to scale emphasize yield fluctuations with the same
proportional increase of all agricultural inputs, whereas economies of
scale also include non-proportional changes in inputs. Second, econo-
mies of scale that are measured from the perspective of average cost
facilitate the analysis of monetary value. In contrast, returns to scale
that are measured from the perspective of production function tech-
nology enable an analysis of physical value (Xu et al., 2011). Since the
possibility of farmers changing agricultural inputs at the same pro-
portion is low, they will rely more on economies of scale to lower cost
and improve efficiency.
Fig. 1a illustrates economies of scale. With a yield increase, the
average cost first increases and then decreases. If marginal cost
(MC) < average cost (AC), the average cost will continue to decrease,
H. Lu et al.
Fig. 1. a. AC and MC curve. b. Maxey-Silberston curve.
as the marginal cost is lower than the average cost. If MC > AC, the
average cost will continue increasing, as the marginal cost curve will
raise the average cost curve. Thus, the average cost curve will make a
“U” pattern. We can determine that Q⁎ is the minimum efficient yield
for a specific production process. According to the Maxey-Silberston
curve, the unit cost keeps decreasing with the expansion of the farm
size, but the speed of this decrease is progressively slower (see Fig. 1b).
This finding also reveals that a moderate expansion of farm size con-
tributes to reducing costs and increasing profitability.
Compared to overall agricultural production, small plots not only
use more time when a machine changes direction but also slow down
the speed and efficiency of machinery. Moreover, small plots also in-
crease the mechanical cost, and, thus, machines may be an inefficient
alternative to labor. The larger the plot, the more effective it is to use
machinery in terms of time and speed. However, if the plot size is too
large, it may exceed the capability of farmers, since the requirements
for agricultural inputs and other fixed assets will increase, which would
eventually make management difficult. The tendency of hired labor to
shirk duties will also increase the cost of supervision and management.
Therefore, there is a moderate scale in plot size that minimizes the
average cost. Thus, we hypothesize:
H3. Even if land fragmentation cannot be changed completely, plot size
enlargement decreases the average cost.
2.2. Research model
In certain existing literature, the number of plots, average plot size,
plot shape, and distance between plots have been used to identify the
extent of land fragmentation (Demetriou et al., 2013; King and Burton,
1982; Manjunatha et al., 2013). However, because of the constraints
associated with the applied and available data, few empirical studies
take all of these factors into account. In general, long line plots are
conducive to the use of fast-moving machinery, and wide line plots are
conducive to the use of wide-width machinery. The study areas are
mainly in the plains, and the plots are relatively regular. There are no
significant differences in the shape of the plots among farmers. It is also
difficult for farmers to tell exactly the shape of each plot.
The cadastral database in China currently cannot identify which
farmer the farmland belongs to because the plot boundary and land
property are not clear, and only land type can be distinguished, such as
arable land, woodland, or garden. In recent years, agricultural land
system reform and the separation of land ownership, contract rights,
and management rights in China are advancing, and the plot boundary
among farmers are being clear. The space distribution of farm plots,
land shape, and distances to home might be included in the cadastral
database in future. The aim of this paper is to analyze the economic
problem of land fragmentation from the perspective of farmer.
Therefore, data used in this paper is the collected empirical data based
on household surveys.
Additionally, due to the low efficiency of land transfer in China, it is
very difficult for farmers to gain land adjacent to their own land
through transferring one time, and they often need two or three
83
Agricultural Systems 161 (2018) 81–88
transfers to succeed. Operating multiple, separate plots of land by one
single farmer is also very common because of the implementation of the
HRS in China. It seems that the plots spread over a larger area results in
the increase in transportation length, and the expansion of farm size is
more likely due to an increase in the number of plots rather than in the
size of the plot.
Unfortunately, the household survey data we use also did not pro-
vide a variable to identify the travel time and distance between plots.
However, there is a positive relationship between the number of plots
and the travel time in the study area based on the above analysis. The
number of plots can serve as a proxy variable for the distance between
plots. Furthermore, the distance between the plots and home is gen-
erally endogenous to their production decisions. That is, farmers often
take the distance between the plots into consideration in determining
land use. Therefore, the variables of the distance between the plots and
the home may be endogenous, and the regression results are biased
when using a production function for empirical studies.
Some scholars use the Simpson index as a measurement for land
fragmentation (Chen et al., 2009; Kawasaki, 2010), and it is one of the
most known and basic indexes of fragmentation. The Simpson index
takes into account the number of plots and the area of each plot.
Kawasaki (2010) indicated that the Simpson index must have a positive
relationship with travel time or distance and, hence, can be used as a
proxy for those variables. Furthermore, it is difficult for farmers to
choose the locations and number of plots. The Simpson index and the
size of each plot are assumed to be exogenous, and it is appropriate to
use the Simpson index and plot size to study the economic problem of
land fragmentation in the study area. The values of the Simpson index
also remain identical. Therefore, taking previous studies and limited
data into consideration, we can use the Simpson index and plot size to
measure land fragmentation. The Simpson index is defined as follows:
n
⎛⎜ ∑ a 2 ⎞⎟
i
si = 1 − ⎝
i=1 ⎠
n 2
⎛⎜ ∑ a ⎞⎟
i
⎝i = 1 ⎠ (1)
where n is the number of plots, ai is the area of each plot, and si lies
between 0 and 1. si = 0 indicates that the farmer only has one plot, and
si = 1 indicates that the farmer has many plots. The larger the value of
si, the higher the degree of land fragmentation.
2.2.1. Input-output model
The analysis builds an input-output model to test whether land
fragmentation affects the efficiency of agricultural inputs and yields.
Despite the advantages of the C-D function in calculation and inter-
pretation, it imposes some strong restrictions on flexibility by assuming
that the elasticity of substitution is one among farmers. On the other
hand, the translog function, which imposes no restriction on the elas-
ticity of substitution, is very flexible. The differences in agricultural
production capability give the function a unique advantage in esti-
mating the input-output among farmers. The function can be written as
follows:
ln yh = a0 + a1 ln fc + a2 ln ccs + a3 ln lab + a4 ln area
1
+ a11 ln fc ln fc
2
1 1 1
+ a22 ln ccs ln ccs + a33 ln lab ln lab + a44 ln area ln area
2 2 2
+ a12 ln fc ln ccs + a13 ln fc ln lab + a23 ln ccs ln lab + a14 ln fc ln area
+ a24 ln area ln ccs + a34 ln area ln lab + Xi + ui (2)
where yh is the yield; fc is the value of fertilizer per mu, including
commodity and manure fertilizer sold to other farmers, which is esti-
mated by rewards; lab is the working days, including family and hired
labor; and ccs is the other inputs containing the seeds, irrigation, ma-
chinery, and other capital expenses. Area is the area, α0 is the efficiency
H. Lu et al. Agricultural Systems 161 (2018) 81–88

parameter, Xi is a vector of control variables, and ui is a random dis- technologies and reduce cost.
turbance term. Cross-sectional data are used in this study. The Ramsey regression
Land fragmentation is not in the form of specific production factors equation specification error test is conducted to determine whether the
involved in agricultural production but affects the change in the mar- model has omitted variables and to select whether the linear, double
ginal output of labor, fertilizers, and other substance capital that affects logarithmic, or semi-logarithmic model is most appropriate by using R2
agricultural yield. Hence, this analysis places Simpson index into the and F tests. The final model is as follows:
model in linear form. The function can be specified as follows:
ln aci = β0 + β1 age + β2 edu + β3 ratio + β4 ser + β5 ps
ln yh = a0 + a1 si × ln fc + a2 si × ln ccs + a3 si × ln lab
+ β6 si
+ a4 si × ln area + β7 ftl1 + β8 credit + vi (5)
1 1 1
+ a11 ln fc ln fc + a22 ln ccs ln ccs + a33 ln lab ln lab
2 2 2 where ac is the average cost per mu; ps is the average plot size; si is the
1 Simpson index, which reflects land fragmentation; credit is a dummy
+ a44 ln area ln area + a12 ln fc ln ccs + a13 ln fc ln lab
2 variable where a value of 1 indicates receiving credit and 0 indicates
+ a23 ln ccs ln lab + a14 ln fc ln area + a24 ln area ln ccs otherwise; and ser is a dummy variable where a value of 1 indicates that
+ a34 ln area ln lab + Xi + ui (3) farmers received an agricultural extension and 0 indicates otherwise.
ftl1 is the proportion of family labor to total labor, ratio is the propor-
The cost increase in farm labor already exists, especially in devel-
tion of off-farm income to total income, and age and edu represent
oped regions. The gaps in the rural economy cause the effects of land
household characteristics.
fragmentation on labor, yields, and flexible management to differ.
Therefore, we add dummy variables to reflect the differences in eco-
nomic development. With Xuzhou as the reference, the dummy variable 2.3. Research material
d1 = 1 in Yangzhou and 0 in other locations, and the dummy variable
d2 = 1 in Huai'an and 0 in other locations. To estimate the compre- 2.3.1. Study area
hensive effects on agriculture at different levels of economic develop- The Jiangsu province is located in the middle of the Yangtze River
ment, we establish the interaction between the labor and the dummy Delta (Fig. 2. It covers about 1.76 million ha with an average elevation
variables. The model is as follows: below 50 m. The northern subtropical monsoon climate dominates this
region year round, with a mean annual temperature of 17.5 °C and a
ln yh = a0 + a1 si × ln fc + a2 si × ln ccs + a3 si × ln lab mean annual rainfall of 1055 mm, all of which are beneficial for agri-
+ a4 si × ln area cultural production. Jiangsu has 4.58 million ha of arable land, with
1 1 1 0.057 ha of arable land per capita. The topography in Jiangsu is mainly
+ a11 ln fc ln fc + a22 ln ccs ln ccs + a33 ln lab ln lab
2 2 2 plains, which cover an area of more than 70,000 km2, accounting for
1 more than 70% of the province. Economic development is not balanced
+ a44 ln area ln area + a12 ln fc ln ccs + a13 ln fc ln lab
2 across the regions of Jiangsu. Southern Jiangsu Province is one of the
+ a23 ln ccs ln lab + a14 ln fc ln area + a24 ln area ln ccs fastest growing areas of eastern China's economy. With a geographic
2 proximity to the most industrialized city of Shanghai, this region is
+ a34 ln area ln lab + (si × ln lab) ∑ ρm dm + Xi + ui renowned for its export-oriented, predominantly high-tech manu-
m=1 (4)
facturing industry.
We can estimate the effects of land fragmentation on agricultural Agricultural modernization in Jiangsu is at the forefront in China,
production by using these models. By contrasting Models 1 and 2, we and the results in the province may be a reference for the whole
can test H1. When Model 2 is contrasted with Model 3, H2 is tested. country. However, the agricultural economies in the southern, central,
and northern regions of the province differ greatly in terms of devel-
2.2.2. The average cost model opment. Agriculture in southern Jiangsu is managed by collective
Due to the flexibility of the translog function, the elasticity of each management, and there is very few in operating agriculture by one
input is given by eyx = ∂ ln yh/∂ ln xi, and the returns to scale also can single farmer. Huai'an is in the central region of Jiangsu, Yangzhou is
be calculated. Simultaneously, the decrease in the average cost relies close to the southern region, and Xuzhou is in the northern region. The
more on economies of scale in agricultural production. Therefore, we circulation of agricultural land management rights is changing quickly
use a linear model to test whether a moderate expansion of plot size in the province; there is a major need for spatial structure reorganiza-
contributes to reducing the average cost. tions for agricultural modernization, and some systemic and practical
The average cost is affected not only by land characteristics such as activities are taking place.
the Simpson index and the number of plots, but also by household All of the data used in this study were collected through a ques-
characteristics, the family environment, and external market informa- tionnaire survey of households in Jiangsu conducted in 2013, in order
tion. The higher the degree of land fragmentation, the greater the dif- to ensure representativeness of the overall region, considering the level
ficulty in managing the land. The average cost will also increase, as of economic development and the actual situation of agricultural pro-
small plots limit the use of modern agricultural technology. duction in these regions. Using the method of stratified sampling, the
Additionally, the effect on agricultural efficiency is also different be- study selected Yangzhou, Huai'an, and Xuzhou as the areas of research.
tween family and hired labor. The increasing proportion of off-farm Subsequently, we selected two counties in each city, two towns in each
income implies that the time spent managing agriculture decreases, and county, and two villages in each town.
extensive management leads to an increase in average cost. Tan et al. The indicators we used in questionnaire include: (1) household
(2010) revealed that credit could reduce agricultural costs, as it eases characteristics, such as land resources, labor, household income, age,
funding constraints in planting, allowing farmers to purchase agri- education, and whether there is a village leader in family; (2) land
cultural machinery, hire labor, and manage land in a timely fashion. characteristics, such as soil quality, land slope, and land type; (3) costs
The higher the level of agricultural extension received by farmers, the and benefits of agricultural production, such as the crop varieties, areas,
more agricultural information they can obtain. With increased age, and the input and output at the plot level; and (4) agricultural market
farmers' planting experience will also increase, but their physical cap- information, containing agricultural extension, credit, and price in-
abilities will decline. Hence, the impact on agricultural cost may be formation. To ensure the quality of the data, the researchers were
mixed. Rising education will increase the probability of adapting new provided training related to the content of the questionnaire, along with

84
H. Lu et al. Agricultural Systems 161 (2018) 81–88

Fig. 2. Location of Jiangsu Province and the study area.

the meaning and purpose of each question, prior to conducting the Table 2
survey. During the survey, the researchers were required to read the Descriptive statistical of variables.
questions from the questionnaire to the respondents and record their
Variables Mean S.E
answers for each question. Then, all responses were collected for a final
check. After excluding disqualified questionnaires, 446 valid responses Yields (lny) 7.06 0.61
were obtained, implying a 96% recovery rate. Average cost (lnac) 7.12 0.896
Labor (lnlab) 3.99 1.07
Fertilizer (lnfc) 5.74 0.77
2.3.2. Variables description and statistical analysis Other capital (lnccs) 5.86 0.62
Table 1 shows the land characteristics of the study area. The average Simpson index (si) 0.57 0.24
Plot size (ps) 2.34 2.89
size of a plot is 0.16 ha, and the difference ranges from 0.03 ha to Number of Plot (npt) 3.72 1.93
2.87 ha. Approximately 73.4% of farmers own small plots, with an Soil quality (quality) 1.60 0.59
average size below 0.16 ha. The average plot size is 0.18 mu in Huai'an, Land slope (slope) 1.06 0.31
which is the largest area compared to 0.14 ha in Xuzhou and 0.16 mu in Household age (age) 58.1 10.66
Household education (edu) 6.48 3.64
Yangzhou. We observe that the larger the plot size, the lower the
Proportion of family labor to total labor (ftl1) 0.91 0.21
average cost. In some regions, fragmentation is a serious issue—the Proportion of off-farm income to total income (ratio) 0.31 0.38
average Simpson index is 0.57, and the highest is 0.9. The average Availability of credit (credit) 0.18 0.38
number of plots is 3.72 per family, with 12 plots being the biggest. The Availability of agricultural extension (ser) 0.95 0.22
proportion of families who have dispersed plots is 91.8%. This result is
Note: calculated from survey data.
due to the implementation of the HRS and the equal distribution of land
based on the land quality and the distance from home. The majority of
farmers' plots are located at different places in the village, with very few agricultural labor relies predominantly on family labor. It also reveals
within the distance of residence. that family laborers had more concerns than laborers hired in agri-
Household characteristics also influence the total cost of agri- cultural production. Thirty one percent off-farm income suggests that
cultural production. Table 2 shows the statistics of variables. The pro- family income is more dependent on farm income, possibly indicating a
portion of family labor to total labor is 91%, which indicates that low and unstable opportunity cost of farm labor and a lower tendency
to work outside the village (in the areas under study). The household
Table 1
age is an average of 58.1 years. The average number of years of edu-
The land characteristics of the sample area. cation is 6.48 years, which causes a lower probability of adopting new
agricultural technologies. As per our investigation, only 20% of
Yangzhou Huai'an Xuzhou households can obtain credit from the financial sector. Approximately
Plots 4.13 3.8 3.25
95% of households accept different types of agricultural extensions.
Simpson index 0.61 0.57 0.52 Doing so not only reduces certain transaction costs but also helps
Average plot size (ha) 0.16 0.18 0.14 farmers keep abreast of market information.
Average cost (Yuan/ha) 35391.75 15237.75 41085 We take into account the influence of land quality on yields.
Obs 154 154 158
Furthermore, the model controls for soil quality. Its value is obtained
Note: Calculated from the household survey data.
through the farmers' cognition of the land quality. The value of one

85
H. Lu et al. Agricultural Systems 161 (2018) 81–88

Table 3 Table 5
The characteristics of individual cost. Parameter estimates of the trans-log models.

Total cost Labor cost Fertilizer cost Other material Parameter Model 1 Model 2 Model 3
cost
Coef. S.E Coef. S.E. Coef. S.E.
Average cost (Yuan/ 30,663.3 15,835.95 8410.05 6417.3
ha) α0 −3.98⁎⁎ 2.02 6.09⁎⁎⁎ (0.14 6.50⁎⁎⁎ 0.18
Proportion (%) 100 0.52 0.27 0.21 α1 1.08⁎⁎⁎ 0.37 0.11 (0.15 − 0.30⁎⁎⁎ 0.16
α2 1.92⁎⁎⁎ 0.47 0.09 (0.14 0.12 0.15
Note: Calculated from survey data. α3 0.365 0.28 0.03⁎ (0.12 0.08 0.11
α4 0.88 0.54 0.46⁎⁎⁎ (0.14 0.42⁎⁎⁎ 0.15
α11 −0.10⁎⁎⁎ 0.03 − 0.10⁎⁎⁎ (0.03 − 0.05⁎ 0.03
indicates that the farmer thinks the land quality is good. A value of α22 −0.25⁎⁎⁎ 0.08 − 0.16⁎⁎⁎ (0.05 − 0.08 0.06
three indicates that the farmer thinks the land quality is very poor. α33 0.01 0.03 0.006 (0.03 0.02 0.03
Otherwise, its value is two. The average soil quality is 1.6 in the sample α44 0.1 0.1 − 0.07 (0.07 0.05 0.08
area, which is a medium level. In general, soil quality has a positive α12 −0.01 0.06 0.17⁎⁎⁎ (0.04 0.10⁎⁎ 0.04
α13 −0.1⁎⁎ 0.04 − 0.07⁎⁎⁎ (0.04 − 0.02 0.04
effect on yields. Furthermore, the analysis also controls for the land
α23 0.03 0.04 0.06 (0.04 0.008 0.04
slope in the model. A value of one indicates a land slope less than 15°. A α14 0.09 0.07 0.08 (0.06 0.08 0.07
value of two indicates a land slope between 15° and 25°. A value of α24 −0.10 0.07 0.06 (0.06 0.06 0.07
three indicates the land slope is greater than 25°. If the land is terraced, α34 0.03 0.04 0.06 (0.04 − 0.003 0.04
the value is four. Quality −0.17⁎⁎⁎ 0.05 − 0.18⁎⁎⁎ 0.05 − 0.18⁎⁎⁎ 0.05
Slope −0.307 0.20 − 0.223 0.18 − 0.227 0.34
Plot characteristics are closely related to agricultural costs. Table 3
Age −0.001 0.005 − 0.001 0.003 − 0.001 0.00
reports the statistics of the individual cost. The average cost is 30,663.3 edu 0.021 0.013 0.021 0.013 0.021 0.013
CNY per ha. The labor cost is 15,835.95 CNY per ha, which accounts for ρ1 – – – – − 0.10⁎⁎⁎ 0.02
52% of the average cost. The fertilizer cost amounts to 8410.05 CNY per ρ2 – – – – − 0.03 0.02
R2 0.6745 0.6738 0.7356
ha, which is 27% of the average cost. The cost of other materials, which
F 71.65⁎⁎⁎ 68.12⁎⁎⁎ 108.10⁎⁎⁎
accounts for 21% of the average cost, is 6417.3 CNY per ha. As is
evident, labor is the most important cost in agriculture. Notes: (1) ***, ** and * indicate significance at the 1%, 5%, and 10% level, respectively.
(2) Model 1, 2, and 3 correspond to Formulae 1, 2, and 3.

3. Results
Table 6
Output and scale elasticity evaluated at the geometric sample means.
The choice between a C-D and translog model can be tested using F-
tests, with a null of ∑αij = 0 for all i,j. The relevant F-values can be Output elasticity Scale elasticity
found in Table 4 with the corresponding degrees of freedom. Since the
null hypothesis is rejected at the 1% level of significance, the translog Land Labor Fertilizer Other capital

model is appropriate for this study. Constant returns to scale can also be Model 1 1.113 0.009 0.276 0.323 1.721
tested by F-tests. At the 1% or 5% level of significance, the null hy- Model 2 1.158 0.011 0.218 0.227 1.614
pothesis of constant returns to scale is rejected; hence, positive or ne- Model 3 1.137 0.018 0.154 0.236 1.545
gative returns to scale may appear in agricultural production.
Note: Since only the labor elasticity of Xuzhou, Yangzhou, and Huai'an can be calculated
Table 5 reports the estimated results of the translog model, and the
in Model 3, the labor elasticity in Table 6 is the geometric average of the three regions.
elasticity of scale is given by eyx = ∂ ln yh/∂ ln xi; that is, a positive
return to scale exists in agricultural production (Table 6). Cai and Li
Furthermore, other capital has a positive effect on yields. The estimate
(1990) found that the proper integration of scatterplots could produce
of the coefficient for labor is statistically significant at the 5% level of
economies of scale and also revealed that economies of scale exists in
significance, with a value of 0.009; that is, if labor increases by 1%, the
Chia's agriculture. However, Wan and Cheng (2001) revealed that there
yields will increase by 0.009%. Furthermore, rising off-farm wages also
is no significant economy of scale in agricultural production in China.
decreased the concerns of family labor. Hence, the impact of labor input
Hayami and Ruttan (1970) believed that agriculture in less developed
on the increase in agricultural production is less than that of increased
countries offered the characteristics of constant returns to scale and
fertilizer and other capital.
agriculture in developed countries had the characteristics of positive
Model 2 is the estimated result when factoring in land fragmenta-
returns to scale.
tion. The coefficient for labor is statistically significant at the 10% level
Model 1 is the estimated results without considering land frag-
of significance, and its value drops to 0.011. Hence, land fragmentation
mentation. The land elasticity of 1.113 is the largest among all inputs,
affects labor efficiency in different regions and reduces yields to a
which implies that a 1% increase in land can lead to a 1.113% increase
certain extent. The estimate of single coefficients for fertilizer is sta-
in yields. The coefficients for fertilizer and other capital are statistically
tistically insignificant at all levels of significance. However, the esti-
significant, with values of 0.276 and 0.323, respectively. A 1% increase
mates of part of the square and cross term for fertilizer and other capital
in fertilizer or other capital will result in yields increasing by 0.276%
are statistically significant, but the values drop to 0.218 and 0.227,
and 0.323%, respectively. These numbers also reveal that investment in
respectively. The value of land elasticity increased to 1.158. Comparing
fertilizer and other capital can promote an increase in yields.
the results of Models 1 and 2, land fragmentation does affect agri-
cultural production efficiency, and it further reduces yields. Therefore,
Table 4
F-values for testing C-D verse Trans-log and returns to scale.
H1 has been supported.
Model 3 estimates the results when factoring in regional differences.
H0:C-D H0:Constant returns to scale The estimated single coefficients for labor are statistically significant at
the 10% level of significance, with values of 0.031, 0.013, and 0.01 in
F-value Degrees of freedom F-value Degrees of freedom
Xuzhou, Huai'an, and Yangzhou, respectively. The reason may be due to
Crops 39.63*** (10,450) 56.57*** (1450) the difference in opportunity costs of farm labor, which has different
impacts on agriculture. The higher the opportunity cost of farm labor,
Note: *** indicates that H0 is rejected at the 1% significance level.

86
H. Lu et al.
the greater the negative impact on labor productivity. The countryside
in Yangzhou is close to the market, so its cost is higher, and social
security is more efficient. Land fragmentation promotes off-farm labor
supply, resulting in a negative impact on yields. The opportunity costs
of the farm and off-farm labor supply are relatively weak in Xuzhou,
which is also supported by the fact that the proportion of family labor
to total labor is relatively higher when compared to those of Yangzhou
and Huai'an. Thus, there is a decrease in the elasticity of land, fertilizer,
and other capital, with values of 1.137, 0.154, and 0.236, respectively.
Differences in the opportunity cost of farm labor impact agricultural
differences. Therefore, H2 has been supported.
Land quality has a strong influence on yields. The coefficients for
land quality are negative and statistically significant at the 1% level
among the models. The worse the land quality is, the greater its ne-
gative influence on yields is. Jacoby et al. (2002) also revealed that
improving land quality help reduce costs and increase yields. The
conclusion also proves that land fallow is very necessary and timely.
Land slope also have a negative influence on yields. The greater the
land slope is, the less conducive the cultivation of food crops is. It also
implies that land consolidation is conducive to improving food pro-
duction. The influences of age and education are consistent with ex-
pectations.
The input-output model reveals that land fragmentation reduces
yields by the change in the marginal output of agricultural inputs. Small
and dispersed plots not only reduce agricultural production capacity
but also increase agricultural cost. Currently, a moderate farm size puts
more emphasis on the average farm size but neglects plot size. Farm size
enlargement is more likely to be the result of plot increases. Therefore,
it may be more appropriate to discuss farm size from the perspective of
plot size. Before regressing on the average cost, we use the robust
weighted least squares estimation to address heteroscedasticity, and we
test collinearity by the variance inflation factor. We find that the values
of all variables are less than 10, which indicates that collinearity is not
present in the model. Table 7 reports the results of the average cost
model.
The coefficients for plot size are negative and statistically significant
at the 1% level, with a value of −0.08. A one-unit increase in plot size
will decrease the average cost by 8%. Due to economies of scale and the
outside effect, plot size enlargement will lead to a decrease in the
average cost. Thus, H3 has also been tested. Using survey data of rice
production in the Jintan district in the Jiangsu province, Lv et al.
(2014) also revealed that the cost decreases with the size of the plot
increase, and economies of scale can be achieved in each plot. The
conclusion is also consistent with the results of this analysis. Ad-
ditionally, the Simpson index is statistically significant at the 5% level.
We may thus conclude that a one-unit increase in the Simpson index
leads to a 39% increase in average cost.
The coefficient for credit is statistically significant at the 1% level,
but the sign contradicts the expectation. One possible reason for this is
that access to credit can ease funding constraints in planting, allowing
farmers to purchase agricultural machinery and hire labor on time.
Table 7
Parameter estimates of the average cost models.
Variables Coefficient
Household age − 0.004
Household edu − 0.01
Proportion of off-farm income to total income − 0.027
Availability of credit 0.21⁎
Availability of agricultural extension − 0.20
Plot size − 0.08⁎⁎⁎
Simpson index 0.39⁎⁎
Proportion of family labor to total labor − 0.003
Constant 8.18⁎⁎⁎
Notes: (1) ***, ** and * indicate significance at the 1%, 5%, and 10% level, respectively.
Agricultural Systems 161 (2018) 81–88
However, due to age constraints of farmers, they must inevitably rely
on traditional planting methods; hence, they lack the initiative to use
modern agricultural technologies. The negative coefficient of family
labor to total labor implies that the higher the proportion, the greater
the decrease in the average cost. Hired labor will require more ex-
tensive production methods, and they invest less in the farm, as the
inputs are not their own. The tendency to shirk and the high cost of
supervision also decrease the efficiency of inputs.
4. Conclusion and discussion
This study estimated the impacts of land fragmentation and plot size
on the efficiency of agricultural production. In studying the economic
problems of land fragmentation, we should focused more on the as-
sessment of the changes in the size of the plot for each household rather
than the size of the holdings. In this study, a translog and cost function
approach has been employed to analyze these problems by using mass
empirical data collected from surveys. The results reveal that land
fragmentation affects yields by the change in the marginal outputs of
labor, fertilizer, and other capital. In areas with a higher opportunity
cost of farm labor, the negative influence on yields is clearer. By
keeping the number of plots constant, plot size enlargement could re-
duce average cost. A moderate plot size could influence the effective
allocation of labor and other inputs. There is a similarity with the most
important effects of land consolidation process, the area structure of the
farms unchanged, the greatest changes are observed in the size of plot.
Additionally, there may be a moderate number of plots for farmers to
own, and, thus, economies of scale in each plot can be achieved in
agriculture.
The existence of economies of scale in each plot shows that land
fragmentation has a negative effect on agricultural production. This
effect is closely related to the size of each plot. Keeping the farm size
constant, reducing the number of plots, and expanding the area of each
plot to aid contiguous cultivation can also lead to economies of scale.
Therefore, agricultural scale management should be discussed not only
from the perspective of expanding the farm size but also from that of
expanding the size of the plot. Economies of scale in each plot can be
achieved by either land consolidation or transfer, or by joint farming
and joint association.
Currently, the construction of ecological civilization and the pro-
tection of landscape values are becoming increasingly important in
China. The needs with regard to ecological and environment protec-
tions and agreement-based land consolidation for the implementation
of area-based environment policies should be taken into consideration
(Zhang and Zhao, 2007; Haldrup, 2015). Additionally, Janus and Iwona
(2017) evaluated the existing funding strategies of land consolidation in
Poland and found poor effects of land swapping and re-parceling, which
are problematic given the financial scale and time spent conducting
these activities.
In the context of an accelerating process of urbanization, accel-
erating migration of the rural labor force, and rising labor costs, the
scale of operating conditions in agricultural production will be more
mature (Lu and Xie, 2018). Thus, issues of lower land transfers and land
fragmentation may be significantly resolved. The government should
S.E create more non-farm employment opportunities and provide condi-
tions for farmers exchanging land to expand their plot size. Ad-
0.004 ditionally, in order to increase the machine efficiency and strengthen
0.012
0.11
the adoption of new technologies, the government should also provide
0.113 preferential taxes, subsidies, and credit funds to encourage farmers to
0.19 voluntarily exchange land and give up equivalent barter. As per dif-
0.015 ferent planting structures, the government should both consider in-
0.19
tensive cultivation by small farmers and prevent quality reduction in
0.004
0.4 the scale management of farmers with large operations. Moreover,
farmers should have easy access to timely market information and
agricultural technology extension services.
87
H. Lu et al.
Conflicts of interest
The authors have no conflict of interest to declare.
Acknowledgements
This study was supported by the National Natural Science
Foundation of China (No. 41561040); the Key Projects of the National
Social Science Fund of China (No. 15AZD075); the Academic and
Technical Leaders Funding Program for Major Disciplines in Jiangxi
Province (No. 20172BCB22011); the Sixty-second Batch of China
Postdoctoral Science Foundation Funding (No. 2017M622097); the
Technology Foundation of Jiangxi Education Department of China (No.
KJLD14033 & No. GJJ160431), the Humanities and Social Sciences
Research Project of Jiangxi Universities in 2017 (No. JJ17205); Jiangxi
Social Science Planning Youth Doctoral Fund Project in 2017 (No.
17BJ39) and the Fok Ying-Tung Fund (No. 141084).
References
Ali, D.A., Deininger, K., 2014. Is there a farm-size productivity relationship in African
agriculture? Evidence from Rwanda. Land Econ. 91 (2), 317–343.
Barrett, C.B., Bellemare, M.F., Hou, J.Y., 2010. Reconsidering conventional explanations
of the inverse productivity-size relationship. Soc. Sci. Electronic Publishing 38 (1),
88–97.
Blarel, B., Hazell, P., Place, F., 1992. The economics of land fragmentation: evidence from
Ghana and Rwanda. World Bank Econ. Rev. 6 (2), 233–254.
Cai, F., 2008. The terminative era of unlimited supply of labor. Fin. Econ. 3, 16–17.
Cai, F., Li, Z., 1990. The existence and utilization of economies of scale in agricultural in
china. Mod. Econ. Sci. 2, 25–34.
Carter, M.R., Yao, Y., 2002. Local versus global separability in agricultural household
models: the factor price equalization effect of land transfer rights. Am. J. Agric. Econ.
84 (3), 702–715.
Chen, Z., Huffman, W.E., Rozelle, S., 2009. Farm technology and technical efficiency:
evidence from four regions in China. China Econ. Rev. 20 (2), 153–161.
Deininger, K., Savastano, S., Carletto, C., 2012. Land fragmentation, cropland abandon-
ment, and land market operation in Albania. World Dev. 40 (10), 2108–2122.
Demetriou, D., Stillwell, J., See, L., 2013. A new methodology for measuring land frag-
mentation. Comput. Environ. Urban Syst. 39 (39), 71–80.
Falco, S.D., Penov, I., Aleksiew, A., 2010. Agrobiodiversity, farm profits and land frag-
mentation: evidence from Bulgaria. Land Use Policy 27 (3), 763–771.
Haji, J., 2007. Production efficiency of smallholders' vegetable-dominated mixed farming
system in eastern Ethiopia: a non-parametric approach. J. Afr. Econ. 16 (1), 1–27.
Haldrup, N.O., 2015. Agreement based land consolidation-in perspective of new modes of
governance. Land Use Policy 46, 163–177.
Hayami, Y., Ruttan, V.W., 1970. Factor prices and technical change in agricultural de-
velopment: the United States and Japan, 1880–1990. J. Polit. Econ. 79, 1115–1141.
Jacoby, H., Li, G., Rozelle, S., 2002. Hazards of expropriation: tenure insecurity and
Investment in Rural China. Am. Econ. Rev. 92 (5), 1420–1447.
Janus, J., Iwona, M., 2017. Land consolidation-a great need to improve effectiveness. A
case study from Poland. Land Use Policy 65, 143–153.
Jia, L., Martin, P., 2014. How does land fragmentation affect off-farm labor supply: panel
data evidence from china. Agric. Econ. 45 (3), 369–380.
Kalantari, K., Abdollahzadeh, G., 2008. Factors affecting agriculture land fragmentation
in Iran: a case study of Ramjerd sub district in Fars province. Am. J. Agric. Biol. Sci. 3
(1), 358–363.
Kawasaki, K., 2010. The costs and benefits of land fragmentation of rice farms in Japan.
Aust. J. Agric. Resour. Econ. 54, 509–526.
King, R., Burton, S.P., 1982. Land fragmentation, notes on fundamental rural spatial
problem. Prog. Hum. Geogr. 6 (4), 475–494.
88
Agricultural Systems 161 (2018) 81–88
Latruffe, L., Piet, L., 2014. Does land fragmentation affect farm performance? A case study
from Brittany, France. Agric. Syst. 129, 68–80.
Li, S.F., Li, X.B., 2017. Global understanding of farmland abandonment: a review and
prospects. J. Geogr. Sci. 27 (9), 1123–1150.
Long, H., Liu, Y., Wu, X., Dong, G., 2009. Spatio-temporal dynamic patterns of farmland
and rural settlements in Su-Xi-Chang region: implications for building a new coun-
tryside in coastal China. Land Use Policy 26, 322–333.
Lu, H., Xie, H.L., 2018. Impact of changes in labor resources and transfers of land use
rights on agricultural non-point source pollution in Jiangsu Province, China. J.
Environ. Manag. 207, 134–140.
Lv, T., Ji, Y.Q., Yi, Z.Y., 2014. Economies of scale in plot size of rice production-based on
the investigation and analysis of Jintan in Changzhou, Jiangsu Province. J. Agron.
Tech. Econ. 2, 68–75.
Manjunatha, A.V., Anik, A.R., Speelman, S., 2013. Impact of land fragmentation, farm
size, land ownership and crop diversity on profit and efficiency of irrigated farms in
India. Land Use Policy 31 (31), 397–405.
Mankiw, N.G., 1998. In: Liang, X.M. (Ed.), Principles of Economics. Peking University
Press.
Mcpherson, M.F., 1982. Land fragmentation: a selected literature review. In:
Development Discussion Papers. 141. Harvard University, pp. 85.
Niroula, G.S., Thapa, G.B., 2005. Impacts and causes of land fragmentation, and lessons
learned from land consolidation in South Asia. Land Use Policy 22 (4), 358–372.
Paul, M., Githinji, M.W., 2017. Small farms, small plots: land size, fragmentation, and
productivity in Ethiopia. J. Peasant Stud. 4, 1–19.
Rahman, S., Rahman, M., 2008. Impact of land fragmentation and resource ownership on
productivity and efficiency: the case of rice producers in Bangladesh. Land Use Policy
26 (1), 95–103.
Foster, A.D., Rosenzweig, M.R., 2011. Are India Farms Too Small-Mechanization, Agency
Cost, and Farm Efficiency. Working Paper. Yale University.
Samuelson, P.A., Nordhaus, W.D., 1948. In: Yu, J. (Ed.), Microeconomics, Nineteenth
edition. Posts & Telecom Press.
Schultz, T.W., 1953. The Economic Organization of Agriculture. McGraw Hill, New York.
Sklenicka, P., Janovska, V., Salek, M., 2014. The farmland rental paradox: extreme land
ownership fragmentation as a new form of land degradation. Land Use Policy 38 (38),
587–593.
Sreeja, K.G., Madhusoodhanan, C.G., Eldho, T.I., 2015. Transforming river basins: post-
livelihood transition agricultural landscapes and implications for natural resource
governance. J. Environ. Manag. 259, 254–263.
Su, S.L., Zhou, X.C., Wan, C., Li, Y.K., Kong, W.H., 2016. Land use changes to cash crop
plantations: crop types, multilevel determinants and policy implications. Land Use
Policy 50, 379–389.
Tan, S., Heerink, N., Kruseman, G., 2008. Do fragmented landholdings have higher pro-
duction costs? Evidence from rice farmers in northeastern Jiangxi Province, China.
Econ. Rev. 19 (3), 347–358.
Tan, S., Heerink, N.B.M., Kuyvenhove, A., 2010. Impact of land fragmentation on rice
producers' technical efficiency in South-East China. NJAS Wagening. J. Life Sci. 57
(2), 117–123.
Wan, G.H., Cheng, E.J., 2001. Effects of land fragmentation and returns to scale in the
Chinese farming sector. Appl. Econ. 33 (2), 183–194.
Wei, C.L., 2015. Land fragmentation governance and rural land system reform: based on
Gulf county rural research. Beijing Soc. Sci. 5, 90–97.
Xiao, R., Su, S.L., Mai, G.C., Zhang, Z.H., Yang, C.X., 2015. Quantifying determinants of
cash crop expansion and their relative effects using logistic regression modeling and
variance partitioning. Int. J. Appl. Earth Obs. Geoinf. 34, 258–263.
Xie, H.L., Lu, H., 2017. Impact of land fragmentation and non-agricultural labor supply on
circulation of agricultural land management rights. Land Use Policy 68, 355–364.
Xu, Q., Tian, S.C., Xu, Z.G., Shao, T., 2008. Agricultural land system, land fragmentation
and farmers' income inequality. Econ. Res. J. 2, 83–92.
Xu, Q., Yin, R.L., Zhang, H., 2011. Economies of scale, returns to scale and the moderate
scale management of agricultural: based on the empirical analysis of grain production
in China. Econ. Res. J. 3, 59–71.
You, H.Y., 2017. Agricultural landscape dynamics in response to economic transition:
comparisons between different spatial planning zones in Ningbo region, China. Land
Use Policy 61, 316–328.
Zhang, Z., Zhao, W., 2007. Effects of land consolidation on ecological environment. Trans.
Chin. Soc. Agric. Eng. 23 (8), 281–285.