Socio-Economic Planning Sciences xxx (xxxx) xxx

Contents lists available at ScienceDirect

Socio-Economic Planning Sciences

journal homepage: http://www.elsevier.com/locate/seps

Invited paper

Has agricultural labor restructuring improved agricultural labor

productivity in China? A decomposition approach
Tomas Baležentis a, Tianxiang Li b, Xueli Chen c, d, *
a
Lithuanian Institute of Agrarian Economics, Lithuania
b
College of Economics and Management, Nanjing Agricultural University, China
c
Anhui University of Finance and Economics, China
d
Chinese Academy of Social Sciences, China

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

Keywords: This paper presents a framework for analyzing the changes in agricultural labor productivity with regards to the
Agricultural labor productivity structural, land intensity, and land productivity effects. This approach allows for the residual-free decomposition
Index decomposition analysis of data from different levels of aggregation. The logarithmic mean Divisia index was applied for the analysis and
LMDI
a data envelopment analysis model was constructed to identify potential gains in agricultural labor productivity
Chinese agriculture
due to the optimization of input use and output production. The proposed approach was applied to the case of
China over the period of 1997–2017. Province-level data were used to identify the major driving factors behind
agricultural labor productivity change. Land productivity change appeared to be the major source of agricultural
labor productivity gains in China. The structural change was rather negligible, suggesting that the reallocation of
the agricultural labor force did not add to the agricultural labor productivity growth in China. A frontier analysis
indicated that agricultural labor productivity could increase by some 45% on average in case full technical ef­
ficiency is achieved.

1. Introduction decomposition and time reversal,1 such as the logarithmic mean Divisia
index, are missing in the research. In this paper, we suggest that the
Growth in agricultural productivity is an important facet of eco­ changes in agricultural labor productivity be modeled by taking a
nomic development, as it allows for the saving of resource inputs and regional perspective. Specifically, we propose the decomposition of
contributes to poverty reduction [1,2,3]. By tracking agricultural labor changes in agricultural labor productivity at the national level by
productivity (which is a partial productivity indicator), one can identify considering regional dynamics. This allows for the isolation of the effects
the major trends in rural income and, eventually, the economic devel­ associated with pure change in agricultural labor productivity occurring
opment of the rural areas in general. [4] noted that different sources of at the regional (sub-national) level and structural changes at the
growth in agricultural productivity can be identified, including changes inter-regional (national) level. We follow [1] and relate agricultural
in both land productivity and land area per agricultural labor unit. labor productivity to land use. Thus, the changes in agricultural labor
Recently, [1] applied decomposition proposed by Ref. [4] to determine productivity are explained in terms of the pure change in land produc­
global agricultural productivity change. However, the proposed frame­ tivity, the change in the land endowments per agricultural labor unit
work did not account for changes in the distribution of resources across occurring at the regional level, and structural change (i.e., dynamics in
the regions. the distribution of the utilized agricultural area across regions). It should
Even though approaches such as shift-share analysis have been be noted that structural change has been neglected in previous research.
developed to trace the dynamics of labor productivity [5], there is still a In order to decompose the changes in agricultural labor productivity,
gap in the literature regarding the sources of agricultural labor pro­ we suggest that the contributions of land productivity, land intensity
ductivity. Particularly, the application of techniques allowing for perfect (per agricultural labor unit), and the spatial distribution of agricultural

* Corresponding author. Anhui University of Finance and Economics, 962 Caoshan Road, 233030 Bengbu, China.
E-mail addresses: tomas.balezentis@laei.lt (T. Baležentis), garylee0428@gmail.com (T. Li), chenxl@cass.org.cn (X. Chen).
1
Perfect decomposition indicates that the variable of interest (i.e., the change in the aggregate variable) is decomposed into explanatory factors without a residual.
Time reversal means that switching the base and current time periods does not entail changes in the results. See, e.g. Ref. [6,7], for more details.

https://doi.org/10.1016/j.seps.2020.100967
Received 25 February 2020; Received in revised form 21 September 2020; Accepted 1 November 2020
Available online 6 November 2020
0038-0121/© 2020 Elsevier Ltd. All rights reserved.

Please cite this article as: Tomas Baležentis, Socio-Economic Planning Sciences, https://doi.org/10.1016/j.seps.2020.100967
T. Baležentis et al. Socio-Economic Planning Sciences xxx (xxxx) xxx

labor force be isolated by means of the index decomposition analysis agricultural factors within the agricultural sector and between agricul­
(IDA). IDA emerged in energy economics [6,7] and has been applied in a tural and non-agricultural sectors. Capital deepening includes not only
number of other fields [8], including agricultural economics [9,10]. IDA the capital deepening within the agricultural sector (whereby the capital
relies on a multiplicative relationship among the aggregate and factor of a single agricultural labor force is constantly increasing), but also
variables, which allows for the decomposition of change in the aggre­ capital deepening outside the agricultural sector. That is to say, the
gate variable by isolating the effects of the factor variables. The capital accumulation of non-agricultural industries will promote agri­
decomposition can be implemented using different indices [6]. In this culture through increasing investments and employment-induced effects
study, we use the logarithmic mean Divisia index (LMDI), which is by attracting surplus rural labor flows to non-agricultural industries.
computationally efficient and maintains the desirable properties of the There have been different measures and methods adopted to track
decomposition (e.g. perfect decomposition and time reversal). For ap­ the dynamics in agricultural labor productivity. [26] applied DEA to
plications of the LMDI, one can consult, e.g., Ref. [7]. decompose changes in labor productivity in American farms with
China has been acknowledged as a country with agricultural labor respect to efficiency change, technical change, and factor intensity
productivity gains [1]. Indeed, China is still on its way toward estab­ (capital deepening). However, this model did not allow for the structural
lishing a modern agricultural sector, especially if one looks at this pro­ effects to be taken into account. Along similar lines, [27] proposed the
cess on a country-wide scale. Economy-wide transformations are also cereal equivalent productivity of agricultural labor indicator, which
taking place in China [11,12], and there have been attempts to analyze expresses productivity in terms of the amount of grain (tons). [28] used
the dynamics and convergence in agricultural labor productivity [13]. the ratio of agricultural output per labor unit to measure labor pro­
However, China’s low agricultural labor productivity has become the ductivity in monetary terms. European Union (EU) countries and
key weakness of its agricultural competitiveness and sustainable farming types were then compared in terms of this indicator. In another
development. Furthermore, strategies for improving China’s agricul­ study, [29] considered variation in labor productivity (measured in
tural labor productivity lack clear and consistent theory and empirical monetary terms) on the premises of the Solow/Swan model by means of
support. This indicates that an analysis of agricultural labor productivity a spatial estimator for the EU, and [30] established the production
in China is topical from both theoretical and empirical viewpoints. frontier for Nicaraguan farms and calculated the shadow price (wages)
The proposed framework is applied to the case of China to address of labor, which reflects labor productivity. These studies applied frontier
the following research questions: What are the major factors driving methods to measure deviations from the production frontier or other
agricultural labor productivity growth, and how do these factors vary representations of production technology. While these approaches allow
across time and space? In order to quantify the contributions of struc­ for observation-specific estimates of agricultural productivity change,
tural dynamics and pure change in agricultural labor productivity in the structural dynamics remained unaccounted for.
China, we consider the 31 provinces of China over the period of
1997–2017. As this period saw a shift from taxation toward subsidiza­ 2.2. Agricultural labor productivity in China
tion of the Chinese agricultural sector, the changes in agricultural labor
productivity may provide insights into the effects of these policy reforms The issue of agricultural labor productivity growth is particularly
and directions for further improvements. In addition, the situation of important for China. As the most populous country in the world, China
international agricultural markets requires a transition toward more has long been experiencing its agricultural sector as endowed with rural
effective agricultural policies in China. In these circumstances, identi­ surplus labor and a lack of land [31–33]. Such a pattern has resulted in
fying sources of change in agricultural labor productivity may shed light subdued agricultural labor productivity growth in the country. [34]
on the prospective adjustments. measured China’s agricultural labor productivity from 1952 to 2003 and
showed that since the founding of the new China, the country’s agri­
2. Literature review on labor productivity in agriculture cultural labor productivity has been rising with fluctuations. Based on
sample data from the National Bureau of Statistics for 70,000 peasant
2.1. Measuring agricultural labor productivity households, [19] found that China’s agricultural labor productivity has
been increasing at a faster rate than land productivity since 2003.
Numerous studies have shown that labor productivity growth is one Additionally, compared with other industries, labor productivity growth
of the key elements contributing to the economic growth and develop­ in China’s agricultural sector has obviously been lagging behind.
ment of most countries [14]. Labor productivity growth plays an As indicated by Ref. [35]; although China’s agricultural labor force
important role not only in transforming the mode of economic growth has shifted from rural areas to urban areas by large numbers ever since
and coordinating regional economic development, but also in deter­ the 1980s, i.e., the share of the agricultural labor force has been
mining labor wages and income distributions [15,16]. Agricultural labor declining, the expected convergence of labor productivity between the
productivity generally refers to the agricultural output produced by a agricultural and non-agricultural sectors has not occurred. In terms of
unit of agricultural labor input. It appears to be one of the most the constant prices of 1978, the labor productivity of the secondary and
important partial productivity indicators for measuring the performance tertiary industries has been significantly higher than that of the primary
of the agricultural sector in general and the degree of agricultural industry, and the absolute gap is still widening. A broader international
modernization in particular [17,18]. Thus, investigating the dynamics of comparison shows that the development of agricultural labor produc­
agricultural labor productivity is crucial to understanding the agricul­ tivity in China is also significantly lagging behind that in developed
tural development and transformation within a country or region [19, countries. According to the data provided by the World Bank in 2010,
20]. China ranked 16th out of 128 countries in terms of land productivity
Following the framework proposed by Ref. [4]; changes in agricul­ (calculated by grain output per hectare), exceeding the world average;
tural labor productivity can be attributed to changes in land productivity however, China’s labor productivity (calculated by the agricultural
and changes in the land area per labor unit. Therefore, agricultural labor added value of each agricultural worker) ranked only 103rd out of 128
productivity improvement can be secured through two main channels: countries, which was merely 50% of the world average [17]. Another
the continuous improvement of land productivity and the continuous international comparison carried out by Ref. [18] also found that there is
expansion of the scale of agricultural land operation [21–23]. [24] and still a big gap between the absolute value of agricultural labor produc­
[25] have argued that agricultural labor productivity could be improved tivity in China and OECD countries. Norway, New Zealand, Australia,
by changing land productivity and per capita land endowments, and the United States show agricultural labor productivity that is 34.6,
including structural transformation or capital deepening. The structural 28.7, 25.8, and 23.5 times higher than China’s rate, respectively. Mex­
transformation mentioned in these studies refers to the redistribution of ico’s agricultural labor productivity, which ranks lowest in OECD

2
T. Baležentis et al. Socio-Economic Planning Sciences xxx (xxxx) xxx

countries, has also reached 1.85 times that of China. Generally speaking, denoted by Xk . The quantity of agricultural labor input is measured at
∑
the fact remains that labor productivity lags behind other industries in both the national scale, L, and region-wise, Lk . Obviously, Kk=1 Lk = L.
China and has significant disadvantages in international comparisons. Thus, for a certain time period t, the following identity relating labor
This constitutes a typical feature of China’s agricultural development [8, productivity at the national level, pt , to region- and country-wide factors
23,36]. Accordingly, improvements in labor productivity have become a can be established:
task for the structural reform of the agricultural sector in China. Spe­
cifically, these findings suggest that growth in China’s agricultural labor Yt ∑ Ykt ∑ Ykt Xkt Lkt ∑
K K K
pt = = = = ykt xkt lkt (1)
productivity requires improvements to ensure the competitiveness of the Lt k=1
Lt X Lkt Lt
k=1 kt k=1
agricultural sector at both the national and international levels. At the
national level, rural-urban migration is partially affected by differences where ykt represents land productivity in region k for time period t, xkt
in labor productivity among agricultural and other sectors of the econ­ stands for agricultural land area per labor unit in region k, and lkt is the
omy. At the international level, the low levels of agricultural labor share of agricultural labor force in region k for time period t. In this
productivity indicate a lack of technological innovation. setting, ykt can be regarded as a measure of pure agricultural labor
[37] employed the stochastic frontier analysis (SFA) method to productivity, whereas xkt quantifies the land intensity effect, and lkt may
explore the determinants of China’s agricultural labor productivity capture the shift of labor structure toward specific regions with possibly
changes during 1988–2009, and found that the factor input changes, different resource endowments and agricultural labor productivity (i.e.,
technological progress, and the accumulation of human capital all pro­ structural effect). Land intensity is defined as the ratio of land area per
moted the growth of agricultural labor productivity in China’s provinces agricultural labor unit. This is the term taken into account by Ref. [4].
and regions. On the basis of estimating the agricultural capital stock at Note that its inverse is the labor intensity per land area. Also note that pt
the provincial level in China, [38] incorporated capital, labor, and land is defined in terms of output per agricultural labor unit, whereas ykt is
inputs into the agricultural production function. The results showed that defined in terms of output per land area.
the per capita capital ownership and per capita arable land area had a Agricultural labor productivity may evolve over time. Considering
positive impact on agricultural labor productivity. In another study, the two time periods 0 and T, one can define the additive decomposition
[24] decomposed the regional disparity in China’s agricultural labor of change in yt as follows:
productivity growth into its components, including technical change,
Δp = pT − p0 = Δpy + Δpx + Δpl (2)
efficiency change, and input accumulation per worker, and found that
during 1987–2005, although the growth of China’s agricultural labor where Δpy is the effect of change in land productivity (capturing the pure
productivity mainly depended on the accumulation of inputs, technical change in agricultural labor productivity), Δpx is the effect of agricul­
changes contributed more to regional disparities in agricultural pro­ tural labor intensity per unit of land area (in an inverted form), and Δpl is
ductivity growth. [39] pointed out that factor endowment differences the effect of the changes in the regional distribution of the labor force.
are still an important factor affecting agricultural labor productivity in Accordingly, the aggregate agricultural labor productivity can change
different regions of China. This suggests that agricultural subsidies may due to region-specific agricultural labor productivity change (captured
be used to improve agricultural performance through resource reallo­ by land productivity change) given the fixed land use and labor distri­
cation. Further, [40] and [41] argued that institutional arrangements, bution. For fixed agricultural labor productivity at the regional level and
land transfer, and land reallocation are also important in promoting fixed labor distribution, the aggregate agricultural labor productivity
agricultural labor productivity in China. An efficient reallocation of land may change due to land use changes (i.e., allocation of agricultural area
to existing farmers in China could have increased aggregate labor pro­ per unit of labor). For fixed regional agricultural labor productivity and
ductivity by 1.88 times during 2003–2014 [41]. While there have land use, the shifts in the regional distribution of the agricultural labor
certainly been different factors governing the growth in agricultural force may affect the level of the aggregate agricultural labor produc­
labor productivity identified in the literature, China’s overall agricul­ tivity. In reality, one may expect the three aforementioned factors to
tural labor productivity change has not yet been effectively interact, causing temporal variations in the aggregate agricultural labor
decomposed. productivity. The LMDI can be applied to quantify these effects.
In summary, earlier studies have enriched the understanding of
China’s agricultural labor productivity dynamics and growth mecha­
nism. Still, there are some deficiencies in the following aspects: First, 3.2. LMDI
structural change and other factors governing change in labor produc­
tivity have not been analyzed in a unified model. Second, the potential The changes in the aggregate variable need to be attributed to dy­
for increasing labor productivity is often neglected by focusing on namics in the explanatory variables. Accordingly, the relationships
measures of productivity change. We address these issues by proposing defined in Eqs. (1) and (2) can be operationalized by applying a number
the index decomposition analysis framework in the next section. of techniques [6]. The LMDI can be considered an effective tool for
isolating the effects of the explanatory factors by calculating the
3. Methods weighted means. In our case, the regional changes in the explanatory
variables are weighted by changes in regional contributions toward
3.1. Index decomposition analysis agricultural labor productivity:
( ) ( )
IDA is a flexible approach that can be adapted to many economic ∑K
YkT Yk0 ykT
Δpy = L , ln (3)
problems. In this case, we seek to decompose changes in a relative in­ k=1
L T L0 yk0
dicator, namely agricultural labor productivity. This implies that we
( ) ( )
must assume constant returns to scale and ignore the scale of operation ∑K
YkT Yk0 xkT
as measured by the quantity of the labor input (this is referred to as the Δpx = L , ln (4)
k=1
LT L0 xk0
activity effect in the IDA literature, e.g., [6,7]. As mentioned in the
Introduction, we propose accounting for regional and inter-regional ∑K ( ) ( )
YkT Yk0 lkT
effects. Thus, let the regions be indexed over. Further, let 0 and T Δpl = L , ln (5)
L L lk0
denote the base and current time periods, respectively. k=1 T 0

The agricultural output is measured at the regional level; thus, we ( ) ( ) ( )

YkT Yk0 Yk0
observe output levels as Yk . The land endowments for each region are where L L T , L0 = YkT
LT − L0 / ln YLkTT − ln YLk00 . Thus, the ratio of the

3
T. Baležentis et al. Socio-Economic Planning Sciences xxx (xxxx) xxx

absolute change over the logarithmic growth rate in the contribution from China’s official statistical database [44]3,4. Because Chongqing
toward labor productivity by a certain province k defined in L( ⋅) serves became a municipality in 1997, we selected 1997–2017 as the sample
as the weighting factor for growth in the explanatory factors at the period to enhance data availability and comparability.
province level. Accordingly, the relative changes in the explanatory For the analysis of labor productivity based on IDA, the gross output
factors at the province level are related to the absolute changes in along with land and labor endowments were used. The potential gains in
agricultural labor productivity at the country level. The use of logged labor productivity were assessed by embarking on the two sets of vari­
values ensures the time reversal property. The changes in agricultural ables in the DEA model. First, “unrestricted” technology consisted of the
labor productivity can be attributed to the three explanatory factors, and same variables as in the case of the IDA (gross output, land, and labor).
the results can be aggregated across time and space. As the decompo­ Second, “restricted” technology also included machinery, pesticides,
sition in Eq. (2) is that of an additive nature, the aggregation is also and fertilizer.
carried out by summing the results across time periods or regions. We further considered three groups of provinces with regard to their
importance to Chinese agriculture. The zone comprised of the major
agricultural producers included the following 10 provinces: Hebei,
3.3. Frontier approach
Heilongjiang, Jiangsu, Anhui, Shandong, Henan, Hubei, Hunan,
Guangdong, and Sichuan. The second zone consisted of 10 provinces of
Data envelopment analysis (DEA) is a frontier technique allowing for
medium agricultural importance: Inner Mongolia, Liaoning, Jilin, Zhe­
the calculation of technical efficiency [42]. Efficiency is defined as the
jiang, Fujian, Jiangxi, Guangxi, Yunnan, Shaanxi, and Xinjiang. The
distance to the production frontier. The conventional DEA model as­
third zone consisted of 11 provinces of minor agricultural importance:
sumes input- or output-orientation, which keeps the input/output mix
Beijing, Tianjin, Shanxi, Shanghai, Hainan, Chongqing, Guizhou, Tibet,
fixed during the optimization. The slacks-based DEA model allows for
Gansu, Qinghai, and Ningxia.
non-radial movement toward the production frontier (i.e., the pro­
portions among inputs or outputs are not maintained), and both inputs
5. Results
and outputs can be optimized simultaneously. In this case, we follow
[43], who presented an instance of the slacks-based model.
The Chinese agricultural sector has expanded in terms of output and,
Let there be K regions indexed over k = 1, 2,…,K. Then, each region
to a much lesser extent, agricultural area since 1997. More specifically,
has its input quantities xki and output quantities ykj , where i = 1, 2, …, m
the stochastic annual rate of growth5 in gross output was 4.7% for
is the index of inputs, and j = 1, 2, …, n is the index of outputs. Then, the 1997–2017. There are some regional differences that can be observed, as
following linear programming problem yields the projection onto the the minor and major agricultural producing provinces show average
production frontier for a certain input-output bundle (x, y) under the rates of growth of 4.5% p.a. and the medium producing provinces show
variable returns to scale [43]: an average rate of 5% p.a. Meanwhile, the agricultural area increased by
∑
m ∑
n just 0.4% p.a. Obviously, this is related to the economic transformation
maxβ = βi + γj in China, which rendered shifts in land use through urbanization. The
i=1 j=1
effects of economic transformations are even more evident in terms of
∑
K
agricultural labor, which declined by 1.5% p.a. during 1997–2017.
λk xki ≤ xi − βi ei , i = 1, 2, ..., m,
k=1
Major agricultural producing provinces were affected to the highest
extent, as suggested by the rate of growth of − 1.8% p.a. The use of other
∑
K (6)
λk ykj ≥ yj + γj ej , j = 1, 2, ..., n, inputs representing mechanization and intensification increased during
k=1 1997–2017. Further, the increasing use of machinery is especially
βi , γj ≥ 0, evident with the annual rate of growth of 4.9% (note that the latter rate
∑
K exceeds that of the total agricultural output). Table 1 presents the sto­
λk = 1, λk ≥ 0. chastic rates of growth for different inputs and outputs.
k=1 The data in Table 1 suggest that the change in agricultural labor
Note that e = (e1 , e2 , …, em , em+1 , em+2 , …, em+n ) is the vector of ones productivity has been driven by a decreasing labor input on the one side
for variables that are optimized and zeros otherwise. The optimal so­ and an increasing agricultural output on the other side. The change in
lution β* is then used to identify the projection point (x − β* e, y + β* e).
Table 1
4. Data used Dynamics in inputs and outputs for Chinese agriculture, 1997–2017 (% p.a.).
Variable Major Medium Minor China
The data covered China’s 31 mainland provinces, municipalities, and
Labor (persons) − 1.8 − 0.9 − 1.1 − 1.5
autonomous regions (hereafter, “provinces”) over the years 1997–2017.
Gross Output (yuan of 2010) 4.5 5.0 4.5 4.7
The data contained five inputs and one output. The five inputs were Total Sown Areas of Farm Crops(hectares) 0.4 0.7 0.2 0.4
labor (primary industry, 10,000 persons), machinery (total power of Agricultural Machinery (kW) 4.7 5.5 4.5 4.9
agricultural machinery, 10,000 kW or kW), land (total sown areas of Agricultural Pesticide(tons) 1.5 3.2 5.3 2.3
Fertilizers (tons) 1.9 3.5 2.1 2.4
farm crops, 1000 ha or ha), pesticides (use of agricultural pesticides,
10,000 tons), and fertilizer (volume of effective component of fertilizer, Note: stochastic annual rates of growth are provided.
10,000 tons). The output measure was the value of the gross output
value of agriculture, forestry, animal husbandry, and fishery (100
million yuan2 in 2010 prices). The input and output data were collected

3
Variable definitions change slightly over time. For example, after 2003, the
2
Yuan is the Chinese currency (Renminbi Yuan). gross output value of agriculture includes services in support of agriculture.
4
Concerns exist regarding the accuracy and reliability of China’s economic
statistics. For agricultural data, the concern is fueled in part by gaps between
official production and consumption data [51,52].
5
The stochastic rate of growth is the slope coefficient, b, from log-lin trend ln
X = a + bt.

4
T. Baležentis et al. Socio-Economic Planning Sciences xxx (xxxx) xxx

agricultural land remained rather insignificant. However, there may be Table 2

certain structural changes that cannot be observed by looking at the Dynamics in relative indicators for Chinese agriculture, 1997–2017 (% p.a.).
absolute data at the aggregate level. Therefore, we further analyzed the Variable Major Medium Minor China
dynamics in relative indicators and input allocation.
Labor productivity (yuan/worker) 6.3 5.9 5.6 6.1
The trends in the relative indicators related to labor productivity in Land productivity (yuan/ha) 4.1 4.3 4.3 4.2
the whole of China are depicted in Fig. 1. As one can note, land intensity Land intensity (ha/worker) 2.2 1.6 1.3 1.9
appears to be the most stable variable due to restrictions on land
Note: stochastic annual rates of growth are provided.
availability in China. Either way, land intensity shows an increasing
trend for 1997–2017 (an increase of 39.5%). However, land productivity
productivity are observed for Jiangsu, Chongqing, Zhejiang, Sichuan,
shows a more robust growth of 130% over the same period. Thus, the
Shaanxi, and Shandong.
growth in labor productivity was around 220% for 1997–2017.
The coefficient of variation of labor productivity indicates the degree
We also looked at the regional differences in the relative indicators
to which the Chinese provinces converge or diverge in regards to the
by considering the three groups of provinces (Table 2). The major
productivity level (Fig. 2). As one can note, the inter-provincial differ­
agricultural producing provinces show a steeper increase in land in­
ences in labor productivity had been increasing during 1997–2003 but
tensity if opposed to the medium and minor agricultural producing
have been declining ever since. In general, the coefficient of variation for
provinces. Slight differences appear in the opposite direction regarding
2017 (0.38) is much lower than that for 1997 (0.5). Therefore, the
land productivity. Regarding labor productivity, major producing
provinces have become more similar in terms of agricultural labor
provinces exhibited a higher rate of growth (6.3% p.a.) if compared to
productivity (i.e., sigma-convergence is observed).
the medium and minor producing provinces (5.9% p.a. and 6.1% p.a.
Thus, there has been variation in absolute and relative indicators
respectively). Thus, more intensive use of land resources in the major
across Chinese provinces, which implies that certain structural changes
agricultural producing provinces, along with negligible differences in
have been taking place. The LMDI model can thus be implemented to
land productivity, implied robust growth of the labor productivity in the
relate the changes in the underlying factors and labor productivity
major producing provinces.
across the provinces of China.
The dynamics in the input and output levels followed different tra­
jectories across the provinces of China during 1997–2017. Accordingly,
Table 3 summarizes the rates of change in the input and output structure 5.1. LMDI results
in percentage points with the two endpoints of the time period consid­
ered. As one can note, the labor force in Jiangsu province declined to the Agricultural labor productivity increased from 10.7 thousand yuan/
highest extent if compared to the overall quantity of the labor input in worker in 1997 to 34.200 yuan/worker in 2017 (the dynamics are
China. In other words, the share of the province in the national labor depicted in Fig. 1). The results of the LMDI-based decomposition indi­
force shrunk by 2.79 p.p. during 1997–2017. At the other end of the cate that much of the change can be attributed to the land productivity
spectrum, Heilongjiang saw the steepest growth in the share of its labor change. Specifically, the increase in labor productivity of some 17.600
force in the national labor force. Indeed, Heilongjiang showed the yuan/worker (75%) can be explained by an increase in the land pro­
steepest upswing in the share of machinery and agricultural area (along ductivity, i.e., improved farming practices, and better allocation in
with negative changes for the rest of the variables). This indicates an terms of the input/output mix. The increase in land intensity (land area
increasing pace of agricultural development in Heilongjiang compared per worker) contributed to the growth in labor productivity of some
to the rest of China. Regarding the contribution to the national agri­ 6.100 yuan per worker (25.8%). This indicates that, for a given spatial
cultural output, Xinjiang province showed the most serious growth of distribution of the labor force and level of land productivity, the changes
0.96 p.p. These findings indicate that different provinces followed in land endowments relative to the labor force boosted agricultural labor
different paths of agricultural development, and that structural com­ productivity in China. The spatial distribution effect played a negligible
ponents may play a role in shaping labor productivity and, especially, role and was associated with an increase in labor productivity of just 200
regional contributions to its change. yuan/worker. Note that these contributions to the change in labor
We further explored the dynamics in the relative indicators gov­ productivity are defined for the national level. We further broke these
erning labor productivity in China at the provincial level. Table 4 pre­ down along the different dimensions of the data (time and space).
sents the logged rates of change in labor productivity for 1997–2017. A look at the dynamics in labor productivity and the contributions of
Regarding labor and land productivity, all the provinces saw an increase its factors over 1997–2017 (see Fig. 3) reveals that different factors
during 1997–2017. However, the dynamics in land intensity followed appeared as the driving ones across different sub-periods. In particular,
different directions due to increasing urbanization in China [45]. For the sub-period of 1997–2003 saw an annual growth in labor produc­
instance, Beijing, Hainan, Fujian, Guangdong, and Shanxi saw declines tivity of some 500 yuan/worker, with land productivity being the sole
in land area per worker. The highest gains in agricultural labor contributor to its growth. Indeed, this sub-period coincides with agri­
cultural policies that did not encourage agricultural development in
China through such measures as tax reduction or subsidization. The sub-
period of 2003–2015 was marked by more robust growth in labor pro­
ductivity amounting to some 1500 yuan/worker per annum. During the
latter sub-period, both land intensity and land productivity appeared to
be the driving forces behind the growth in agricultural productivity
(with the exception of 1999–2000). Finally, the most recent sub-period
of 2015–2017 exhibited a decline in the contribution of the land in­
tensity component, which indicates that urbanization has been achieved
at a particularly high extent in China. In any event, agricultural pro­
ductivity growth remained stable and even increased during
2016–2017. This suggests that land availability does not preclude labor
productivity growth in China due to the implementation of advanced
farming practices.
Fig. 1. Dynamics in the relative indicators for the whole of China (1997 The drivers of change in agricultural labor productivity across the
= 100). three major regions are summarized in Table 5. Out of the total change

5
T. Baležentis et al. Socio-Economic Planning Sciences xxx (xxxx) xxx

Table 3
The changes in the input and output structure (contributions of the provinces, 2017 compared to 1997, percentage points).
Province Labor Agricultural Machinery Total Areas Sown Pesticides Fertilizers Gross Output

Beijing − 0.02 − 0.89 − 0.28 − 0.34 − 0.35 − 0.26

Tianjin 0.00 − 0.92 − 0.11 − 0.08 − 0.06 − 0.04
Hebei 0.39 − 6.13 − 0.71 − 1.38 − 1.10 0.09
Shanxi 0.67 − 1.99 − 0.33 0.52 − 0.16 − 0.20
Inner Mongolia 0.64 1.05 1.63 1.16 2.19 0.29
Liaoning 0.83 − 0.35 0.15 0.67 − 0.36 0.35
Jilin 0.25 1.49 1.02 1.84 1.24 0.41
Heilongjiang 1.13 2.83 3.01 2.49 1.23 0.73
Shanghai − 0.13 − 0.25 − 0.19 − 0.45 − 0.18 − 0.57
Jiangsu − 2.79 − 0.89 − 0.63 − 3.69 − 2.92 − 0.92
Zhejiang − 1.57 − 2.02 − 1.37 − 2.36 − 1.08 − 1.17
Anhui − 0.64 1.07 − 0.27 − 0.14 − 0.60 − 0.62
Fujian 0.02 − 0.63 − 0.98 − 1.23 − 0.94 − 0.33
Jiangxi − 0.08 0.56 − 0.53 1.48 − 0.72 − 0.25
Shandong − 1.22 − 1.05 − 0.46 − 2.96 − 2.21 − 0.19
Henan 0.92 − 0.14 0.88 0.14 3.14 0.87
Hubei − 0.04 1.36 − 0.24 − 1.80 − 1.16 − 0.29
Hunan − 0.12 2.31 − 0.20 0.09 − 0.22 − 0.42
Guangdong 0.72 − 1.53 − 1.04 − 0.28 − 0.48 − 0.57
Guangxi 0.70 0.89 − 0.44 0.77 0.83 0.34
Hainan 0.30 0.13 − 0.17 1.42 0.32 0.56
Chongqing − 1.14 0.29 − 0.33 − 0.39 − 0.12 − 0.14
Sichuan − 1.64 1.27 − 0.41 − 0.96 − 0.93 − 0.18
Guizhou 0.58 1.13 0.48 0.15 0.00 − 0.01
Yunnan 0.91 0.96 0.69 1.33 1.30 0.73
Tibet 0.11 0.35 0.00 0.01 0.02 0.00
Shaanxi − 0.10 0.17 − 0.48 − 0.08 1.03 0.44
Gansu 0.46 0.11 − 0.18 3.53 − 0.09 0.28
Qinghai − 0.01 − 0.03 − 0.04 − 0.01 − 0.03 − 0.02
Ningxia 0.14 − 0.07 0.05 0.04 0.22 0.13
Xinjiang 0.79 0.94 1.47 0.49 2.17 0.96
Minimum − 2.79 − 6.13 − 1.37 − 3.69 − 2.92 − 1.17
Maximum 1.13 2.83 3.01 3.53 3.14 0.96

in labor productivity over 1997–2017 (23.5 thousand yuan/worker), note, land productivity appeared as the major driver of change in agri­
more than half was attributed to the major agricultural producing cultural productivity for most of the provinces, with the exception of
provinces (13.9 thousand yuan/worker). The minor agricultural pro­ Jiangsu, Xinjiang, Inner Mongolia, Jilin, Chongqing, and Shanghai.
ducing provinces were associated with a much lower contribution to These provinces require a more intensive application of measures for
labor productivity growth (2.2 thousand yuan/worker) if compared to improvement of land productivity and, in turn, labor productivity.
the medium ones (7.5 thousand yuan/worker). Losses of agricultural productivity due to land intensity changes were
Regarding the sources of growth in labor productivity, a positive observed in Guangdong, Fujian, Hainan, Shanxi, and Beijing. Obviously,
spatial distribution effect was observed for the medium and minor these provinces face certain land availability restrictions due to ongoing
agricultural producing provinces, whereas a negative value was urbanization processes. Finally, the negative contributions of the spatial
observed for the major ones. However, the land intensity effect declined distribution effect were observed for Shanghai, Chongqing, Zhejiang,
when moving from the major to minor agricultural provinces. This Anhui, Jiangsu, Sichuan, and Shandong. Most of these provinces are
suggests a certain trade-off between land use per worker and changes in located in coastal areas and experience robust economic growth, which
the spatial distribution of labor. This can be partly explained by is related to labor migration to the urban areas [46].
assuming that higher land intensity ensures proper revenue even
without reallocation to more fertile soils (which may even be impossible
in the case of land supply restrictions). The contribution of land pro­ 5.2. DEA results
ductivity appeared to be a major factor driving growth in labor pro­
ductivity across the three groups of provinces (the contribution The previous sub-section focused on the observed dynamics in labor
amounted to 72–89% of the group contribution to the change in labor productivity and its factors. We can now ascertain whether there is a gap
productivity). in China’s agricultural sector due to technical inefficiency. As described
The minor agricultural producing region contributed to 9.1% of the in Section 4, two technologies were considered, namely restricted and
total change in agricultural labor productivity in China during unrestricted. Restricted technology accounts for the use of variable in­
1997–2017. Indeed, this is comparable to the contribution to the total puts (machinery, fertilizers, and pesticides) aside from land and labor,
output by this region. Specifically, the share of the minor agricultural which are also included in the unrestricted model. DEA was imple­
producing provinces in the national output amounted to some 9.4% over mented in order to quantify the reduction in the use of labor input and
1997–2017. Therefore, these provinces are not likely to act as important possible increase in output. The province-level results were then
contributors to agricultural labor productivity growth. Still, their pro­ aggregated to arrive at the nationwide indicators. Fig. 5 presents the
ductivity should not be left unmonitored, as they could contribute to the results.
governors’ food security responsibility system in China. Under unrestricted technology, labor productivity could increase by
The province-level results are presented in Fig. 4. In this case, the some 45% on average in case full technical efficiency was achieved. As
normalized contributions to the change in national agricultural labor for restricted technology, possible gains of 38% were observed. Thus,
productivity are presented. This allows for the identification of the improving the technical efficiency of China’s agricultural sector would
major drivers of labor productivity at the province level. As one can contribute to the growth in labor productivity to a substantial degree. As
one can note, the gap in labor productivity had been increasing over

6
T. Baležentis et al. Socio-Economic Planning Sciences xxx (xxxx) xxx

Table 4
The changes in relative indicators related to labor productivity in China’s
provinces (2017 compared to 1997, per cent).
Province Labor productivity Land productivity Land intensity
(yuan/worker) (yuan/ha) (ha/worker)

Beijing 64.2 175.5 − 111.4

Tianjin 108.1 108.1 0.0
Hebei 110.4 98.1 12.3
Shanxi 73.3 85.0 − 11.7
Inner 94.5 59.0 35.5
Mongolia
Liaoning 89.3 86.4 2.9
Jilin 120.7 69.0 51.7
Heilongjiang 86.6 62.5 24.1
Shanghai 60.4 41.3 19.2
Jiangsu 167.2 81.7 85.5 Fig. 3. Dynamics in labor productivity and its decomposition for China’s
Zhejiang 146.1 123.9 22.2 agriculture, 1997–2017 (current year is indicated).
Anhui 114.3 74.3 40.0
Fujian 106.3 145.6 − 39.3
Jiangxi 111.1 89.5 21.6
The differences in labor productivity under different technologies (at
Shandong 130.7 88.0 42.7 full efficiency) across the provinces are shown in Table 6. Comparison of
Henan 117.4 83.8 33.6 the labor productivity levels for the restricted and unrestricted tech­
Hubei 112.0 82.9 29.1 nology provides some insight into the role of the variable inputs (ma­
Hunan 110.7 79.2 31.5
chinery, fertilizers, and pesticides). In cases where the ratio of labor
Guangdong 91.1 107.0 − 15.9
Guangxi 111.3 103.5 7.8 productivity under restricted technology to that under unrestricted
Hainan 132.0 172.1 − 40.1 technology falls below unity, a certain province could face difficulties in
Chongqing 156.7 90.0 66.7 obtaining the variable inputs.
Sichuan 135.2 87.3 47.9 The results reported in Table 6 suggest that the lowest ratios of
Guizhou 101.4 67.4 34.0
Yunnan 126.3 91.6 34.7
“restricted” to “unrestricted” labor productivities can be observed for
Tibet 79.8 77.8 2.0 Qinghai, Ningxia, Tibet, Shaanxi, Sichuan, Tianjin, Henan, Guizhou,
Shaanxi 138.3 119.8 18.5 Zhejiang, and Inner Mongolia. Therefore, these regions cannot improve
Gansu 122.9 112.7 10.2 their agricultural labor productivity to the full extent given the con­
Qinghai 112.1 86.6 25.4
straints associated with the variable inputs. Thus, farming practices
Ningxia 122.2 108.9 13.3
Xinjiang 100.8 63.4 37.4 should be improved in these areas by means of advisory services, among
Min 60.4 41.3 − 111.4 other measures. Input support programs could also be directed toward
Max 167.2 175.5 85.5 these areas.
Note: logged rates of growth are provided.
6. Discussion

This research is in line with the work of others, e.g., Ref. [1] who
have reported an increasing labor productivity in Chan’s agriculture
over 1961–2014. It is also consistent with the findings reported by
Ref. [47] who considered the period of 2004–2012. However, the
small-scale farming that prevails in China is associated with a relatively
low labor productivity level when compared to the other regions of the
world. Specifically [48], documented a negative effect of the increase in
land fragmentation on labor productivity in China’s agriculture.
Therefore, further consolidation of the land plots would be beneficial in
China. Also, increasing availability of agricultural services would allow
for the generation of more value to be added on small-scale farms.
Indeed, this paper has shown that the availability of land per labor unit
offered a certain degree of contribution to the increase in agricultural
labor productivity, yet it has become a limiting factor in recent times.
Fig. 2. Coefficient of variation for labor productivity in China’s prov­
The presence of surplus labor in China’s agriculture has been
inces, 1997–2017.
acknowledged in the earlier literature [49,36]. In our paper, we
addressed this issue by applying the DEA model, which adjusted the land
1997–2017, which suggests that catch-up (i.e., movement toward the
and labor inputs in relation to the best-practice frontier. Thus, the es­
production frontier) has become particularly important in China. This
timates of potential labor productivity derived in our paper address this
implies that extension services and similar measures may be beneficial
issue to a certain extent.
in putting the underperforming regions on the production frontier.
Our results indicate that increasing land productivity is a major
An interesting finding is that the restricted and unrestricted tech­
source of labor productivity growth in China’s agriculture. In our cur­
nologies (production possibilities) virtually coincide during 2011–2014.
rent setting, we were unable to take into account capital deepening as
This implies that the availability of variable inputs (machinery, fertil­
was done by Ref. [26]. Therefore, further integration of IDA and
izers, and pesticides) was not a binding factor during the aforemen­
frontier-based approaches is necessary for productivity analysis.
tioned time period. However, it is important to note that restricted
The estimates of productivity gains in China’s agricultural sector are
technology departed from the unrestricted form in 2015, and the recent
based on the non-parametric production frontier. It is important to
result for 2017 shows the largest gap. This indicates that the availability
ascertain whether the results are meaningful and feasible. The study by
of variable inputs plays a key role in further boosting agricultural labor
Ref. [41] can be considered as a reference point. They reported that a
productivity in China.

7
T. Baležentis et al. Socio-Economic Planning Sciences xxx (xxxx) xxx

Table 5
The contributions to the labor productivity change across the three groups of Chinese provinces, 1997–2017.
Group Yuan/worker Per cent

Spatial distribution Land intensity Land productivity Total Spatial distribution Land intensity Land productivity

Major − 1012 4856 10026 13870 − 7 35 72

Medium 729 1072 5707 7508 10 14 76
Minor 109 135 1907 2152 5 6 89
Total − 174 6063 17641 23530 − 1 26 75

Fig. 4. Decomposition of contributions to the change in labor productivity across the provinces of China (%), 1997–2017. Note: Percentages are calculated with
respect to the sum of the absolute contributions to the labor productivity change.

188% increase was possible in labor productivity due to changes in land

use (over 2003–2014). Our study reports lower figures of 1/0.67 =
149% and 1/0.76 = 132% for unrestricted and restricted technologies,
respectively (Table 6). Therefore, the results obtained in our study can
be treated as feasible.

7. Conclusions

As land availability is rather limited in China, the growth in pro­

ductivity remains essential for its agriculture. In this paper, we proposed
a framework for analyzing the dynamics in agricultural labor produc­
tivity. As the process of urbanization proceeds at a varying pace across
the provinces, structural changes occur, i.e., the relative importance of
the provinces changes in terms of the agricultural labor force. We
considered dynamics in agricultural input use (including labor) and
Fig. 5. Observed and potential labor productivity in China’s agricul­
production at both the national and provincial levels, ensuring a
ture, 1997–2017.
residual-free decomposition. Furthermore, we looked at the potential for
improvement in agricultural labor productivity for a given production
technology.

8
T. Baležentis et al. Socio-Economic Planning Sciences xxx (xxxx) xxx

Table 6
The actual and optimal agricultural labor productivity across provinces of China (averages for 1997–2017).
Province Labor productivity, 10000 yuan/worker Ratio

Actual Unrestricted Restricted Actual to unrestricted Actual to restricted Restricted to unrestricted

(1) (2) (3) (1)/(2) (1)/(3) (3)/(2)

Anhui 1.51 2.74 3.07 0.55 0.49 1.12

Beijing 4.75 4.75 4.75 1.00 1.00 1.00
Chongqing 1.26 3.52 3.60 0.36 0.35 1.02
Fujian 2.96 2.96 2.96 1.00 1.00 1.00
Gansu 0.90 3.51 3.83 0.26 0.23 1.09
Guangdong 2.18 2.19 2.18 0.99 1.00 0.99
Guangxi 1.56 2.87 3.62 0.54 0.43 1.26
Guizhou 0.70 3.23 2.64 0.22 0.26 0.82
Hainan 3.17 3.67 3.67 0.86 0.86 1.00
Hebei 2.50 2.87 2.79 0.87 0.90 0.97
Heilongjiang 2.67 3.68 3.84 0.73 0.70 1.04
Henan 1.68 2.48 2.02 0.68 0.83 0.81
Hubei 1.94 2.79 2.81 0.70 0.69 1.01
Hunan 1.72 2.69 2.94 0.64 0.58 1.09
Inner Mongolia 2.82 3.90 3.47 0.72 0.81 0.89
Yunnan 0.98 2.98 3.82 0.33 0.26 1.28
Jiangsu 2.85 2.93 2.89 0.97 0.99 0.99
Jiangxi 1.90 3.38 3.49 0.56 0.55 1.03
Jilin 2.63 3.82 3.92 0.69 0.67 1.03
Liaoning 3.62 3.66 3.62 0.99 1.00 0.99
Ningxia 1.69 4.40 1.83 0.38 0.92 0.42
Qinghai 1.32 4.34 1.50 0.30 0.88 0.35
Shaanxi 1.59 3.35 2.00 0.47 0.80 0.60
Shandong 2.48 2.48 2.48 1.00 1.00 1.00
Shanghai 4.96 4.96 4.96 1.00 1.00 1.00
Shanxi 1.42 3.64 3.68 0.39 0.38 1.01
Sichuan 1.61 2.56 1.61 0.63 1.00 0.63
Tianjin 3.45 4.76 3.74 0.73 0.92 0.79
Tibet 1.02 1.87 1.02 0.54 1.00 0.54
Xinjiang 3.89 4.14 3.89 0.94 1.00 0.94
Zhejiang 2.68 3.15 2.68 0.85 1.00 0.85
Weighted average 1.95 2.96 2.77 0.66 0.71 0.93
Average 2.27 3.36 3.07 0.67 0.76 0.91

The results indicate that structural change (i.e., the distribution of agrochemicals (i.e., organic fertilizers, slow-release fertilizers,
the agricultural labor force across the provinces) played a much less bio-pesticides, and efficient low toxicity, low residue pesticides) will
important role in agricultural labor productivity growth in China during provide more means to increase land and, subsequently, labor produc­
1997–2017 than land intensity (i.e., land endowments per worker) and tivity in China’s agriculture. The results presented here suggested that
land productivity (i.e., output per area unit). Looking at the different Qinghai, Ningxia, Tibet, Shaanxi, Sichuan, Tianjin, Henan, Guizhou,
time sub-periods, the years 2003–2015 were marked by an increased Zhejiang, and Inner Mongolia face constraints related to variable inputs.
effect of the land intensity component. This may be related to the effects Even though the agricultural output generated in these provinces is
of the new agricultural policies in China. However, the sub-period of limited, these regions cannot completely abandon agricultural activities
2015–2017 saw an increasing effect of land productivity, which offset due to food security considerations. Accordingly, the implementation of
the reduction in the land intensity effect. Within increasing urbanization the governors’ food security responsibility system and similar measures
in China, the effect of land intensity is likely to diminish in the future. should consider the actual and potential level of labor productivity in
Further, much of the agricultural labor productivity growth can be different provinces and envisage support for its promotion through
attributed to the major agricultural producing provinces. This suggests modern farming practices in order to avoid excessively high opportunity
that minor agricultural producing provinces still require attention costs.
regarding labor productivity. China has also embarked on the Belt and Road Initiative. Increasing
The promotion of the use of agricultural technologies (both ma­ international cooperation may improve the possibilities for agricultural
chinery and agrochemicals) is still needed to further increase agricul­ trade. Therefore, it is important to determine whether China’s agricul­
tural labor productivity in China. The frontier-based analysis suggested tural labor productivity is comparable to that in the other countries. This
that a gap of 38–45% existed in labor productivity due to the inefficient can be an avenue for further research.
use of inputs. What is more, the input/output mix should be adjusted
within and across the provinces to ensure structural efficiency by CRediT authorship contribution statement
following the comparative advantages in the major and minor agricul­
tural producing areas. Gains in human capital quality could also Tomas Baležentis: Conceptualization, Formal analysis, Methodol­
contribute to labor productivity growth. Increase in the human capital ogy, Writing - original draft. Tianxiang Li: Formal analysis, Writing -
requires the introduction of modern advisory and extension services original draft. Xueli Chen: Conceptualization, Data curation, Formal
with public subsidies covering their costs. analysis, Methodology, Writing - original draft.
In China, a significant amount of attention has been paid to green
and high quality development strategies in agriculture in line with the Acknowledgements
Zero Growth of Chemical Fertilizer and Pesticide Use Actions introduced
since 2015 [50]. Basically, promoting the precise use of fertilizers and Dr. Tianxiang Li acknowledges financial support from the National
pesticides and switching to more effective and environment-friendly Natural Science Foundation of China (No. 71803085, 71773051,

9
T. Baležentis et al. Socio-Economic Planning Sciences xxx (xxxx) xxx

71673142) and Fundamental Research Funds for the Central Univer­ [28] Giannakis E, Bruggeman A. Exploring the labour productivity of agricultural
systems across European regions: a multilevel approach. Land Use Pol 2018;77:
sities (No. SKYZ2020004). Dr. Xueli Chen acknowledges financial sup­
94–106.
port from the Youth Project of National Social Science Foundation of [29] Smit MJ, van Leeuwen, Florax RJ, de Groot HL. Rural development funding and
China (No. 20CXW011). The authors declare that they have no known agricultural labour productivity: a spatial analysis of the European Union at the
competing financial interests or personal relationships that could have NUTS2 level. Ecol Indicat 2015;59:6–18.
[30] Almeida AN, Bravo-Ureta BE. Agricultural productivity, shadow wages and off-
appeared to influence the work reported in this paper. farm labor decisions in Nicaragua. Econ Syst 2019;43(1):99–110.
[31] Cook S. Surplus labour and productivity in Chinese agriculture: evidence from
Appendix A. Supplementary data household survey data. J Dev Stud 1999;35(3):16–44.
[32] Cai F. The Great Exodus: how agricultural surplus laborers have been transferred
and reallocated in China’s reform period? China Agric Econ Rev 2018;10(1):3–15.
Supplementary data to this article can be found online at https://doi. https://doi.org/10.1108/CAER-10-2017-0178.
org/10.1016/j.seps.2020.100967. [33] Wang M. The rise of labor cost and the fall of labor input: has China reached Lewis
turning point? China Econ J 2010;3(2):137–53.
[34] Wang X. An analysis of the characteristics and route of the growth rate of
References agricultural labor productivity in China. The Journal of Quantitative and Technical
Economics 2007;24(4):14–25 [in Chinese].
[1] Fuglie KO. Is agricultural productivity slowing? Global food security 2018;17: [35] Cai F. Has China’s labor mobility exhausted its momentum? China Rural Economy
73–83. 2018;9:2–18 [in Chinese].
[2] Czyżewski B, Matuszczak A, Miśkiewicz R. Public goods versus the farm price-cost [36] Kwan F, Wu Y, Zhuo S. Surplus agricultural labour and China’s Lewis turning point.
squeeze: shaping the sustainability of the EU’s common agricultural policy. China Econ Rev 2018;48:244–57.
Technol Econ Dev Econ 2019;25(1):82–102. [37] Kuang Y. Technology efficiency, technology progress, factor accumulation and
[3] Vasylieva N. Ukrainian agricultural contribution to the world food security: China’s agricultural economic growth. The Journal of Quantitative and Technical
economic problems and prospects. Montenegrin J Econ 2019;14(4):215–24. Economics 2012;1:3–18 [in Chinese].
[4] Hayami Y, Ruttan VW. Agricultural development: an international. Perspective. [38] Li G. Capital deepening, Land-labor ratio and the growth of China’s agricultural
Baltimore, Md: Johns Hopkins University Press; 1985. productivity: an Production function framework. China Rural Economy 2015;1:
[5] Haynes KE, Dinc M. Productivity change in manufacturing regions: a multifactor/ 14–30 [in Chinese].
shift-share approach. Growth Change 1997;28(2):201–21. [39] Lopez RA, He X, Falcis ED. What drives China’s new agricultural subsidies? World
[6] Ang BW. The LMDI approach to decomposition analysis: a practical guide. Energy Dev 2017;93:279–92.
Pol 2005;33(7):867–71. [40] Mao P, Xu J, He X, Zhou Y. Transfer of land and increase of farmers’ labor
[7] Ang BW. LMDI decomposition approach: a guide for implementation. Energy Pol productivity: theoretical and empirical analysis. Econ Res 2015;11:161–76 [in
2015;86:233–8. Chinese].
[8] Chen J, Xu C, Cui L, Huang S, Song M. Driving factors of CO2 emissions and [41] Gai Q, Zhu X, Cheng M, Shi Q. Land misallocation and aggregate labor
inequality characteristics in China: a combined decomposition approach. Energy productivity. Econ Res 2017;5:117–30 [in Chinese].
Econ 2019;78:589–97. [42] Banker RD, Charnes A, Cooper WW. Some models for estimating technical and
[9] Chen J, Cheng S, Song M. Changes in energy-related carbon dioxide emissions of scale inefficiencies in data envelopment analysis. Manag Sci 1984;30(9):1078–92.
the agricultural sector in China from 2005 to 2013. Renew Sustain Energy Rev [43] Färe R, Grosskopf S. Directional distance functions and slacks-based measures of
2018;94:748–61. efficiency. Eur J Oper Res 2010;200(1):320–2.
[10] Li W, Ou Q, Chen Y. Decomposition of China’s CO2 emissions from agriculture [44] National Bureau of Statistics of China. Statistical Database. 2020. http://www.stats
utilizing an improved Kaya identity. Environ Sci Pollut Res 2014;21(22):13000–6. .gov.cn/english/Statisticaldata/AnnualData/ (accessed on: 12 October 2020).
[11] Lv K, ChenXiaosong LX. Study on the duration and influencing factors of China’s [45] Sun J, Wang J, Wang T, Zhang T. Urbanization, economic growth, and
manufacturing export to the countries along the belt and road. Transform Bus Econ environmental pollution. Management of Environmental Quality 2019;30(2):
2019;18(3C):348–66. 483–94.
[12] Cui L, Song M. Economic evaluation of the Belt and Road Initiative from an [46] Wang L, Xue C. Spatio-temporal characteristics and influencing factors of urban
unimpeded trade perspective. Int J Logist Res Appl 2019;22(1):25–46. floating population in China from 2011 to 2015. Chinese Journal of Population
[13] Ma L, Long H, Zhang Y, Tu S, Ge D, Tu X. Agricultural labor changes and Resources and Environment 2019;17(4):359–73.
agricultural economic development in China and their implications for rural [47] Zhang Q, Sun Z, Wu F, Deng X. Understanding rural restructuring in China: the
vitalization. J Geogr Sci 2019;29(2):163–79. impact of changes in labor and capital productivity on domestic agricultural
[14] Astra AE. Labour productivity, economic growth and global competitiveness in production and trade. J Rural Stud 2016;47:552–62.
Post-crisis Period. Procedia - Social and Behavioral Sciences 2014;156:317–21. [48] Lu H, Xie H, Yao G. Impact of land fragmentation on marginal productivity of
[15] Wang X, Shao S, Li L. Agricultural inputs, urbanization, and urban-rural income agricultural labor and non-agricultural labor supply: a case study of Jiangsu, China.
disparity: evidence from China. China Econ Rev 2019;55:67–84. Habitat Int 2019;83:65–72.
[16] Zhang J, Liu X. The evolving pattern of the wage–labor productivity nexus in [49] Bhattacharyya A, Parker E. Labor productivity and migration in Chinese
China: evidence from manufacturing firm-level data. Econ Syst 2013;37(3): agriculture A stochastic frontier approach. China Econ Rev 1999;10(1):59–74.
354–68. [50] Jin S, Zhou F. Zero growth of chemical fertilizer and pesticide use: China’s
[17] Gao F. International experience to improve agricultural productivity and China’s objectives , progress and challenges. Journal of Resources and Ecology 2018;9(1):
policy choice. Fudan Journal (Social Sciences) 2015;57(1):116–24 [in Chinese]. 50–8.
[18] Cao B, Zhu N. Study on paths of agricultural supply side reform from the [51] Plekhanov D. Quality of China’s official statistics: a brief review of academic
perspective of labor productivity. Journal of China Agricultural University 2019;24 perspectives. Cph J Asian Stud 2017;35:76–101.
(1):175–83 [in Chinese]. [52] Rawski TG, Xiao W. Roundtable on Chinese economic statistics introduction. China
[19] Liu S, Wang R, Shi G. Historical transformation of China’s agriculture: productivity Econ Rev 2001;12:298–302.
changes and other key features. China World Econ 2018;26(1):42–65.
[20] Song W, Han Z, Deng X. Changes in productivity, efficiency and technology of Xueli Chen is an Assistant Professor at Chinese Academy of Social Sciences. Her research
China’s crop production under rural restructuring. J Rural Stud 2016;47:563–76. focuses on Media Economics, Efficiency and Productivity Analysis. She has published in
[21] Quan J. China’s agricultural growth path: 1952-2008. Agricultural Economy Issue such outlets as China Economic Review, Sustainable Development, Land Use Policy,
2010;9:10–6 [in Chinese]. Journal of Cleaner Production, Science of The Total Environment and Sustainability
[22] Wu L, Li G, Zhou X. Change of factor endowments and China agricultural growth among others. E-mail: chenxl@cass.org.cn
path selection. China Popul Resour Environ 2015;25(8):144–52 [in Chinese].
[23] Zhao J, Tang J. Understanding agricultural growth in China: an international
perspective. Struct Change Econ Dynam 2018;46:43–51. Tomas Balezentis is a Professorial Research Fellow at Lithuanian Institute of Agrarian
[24] Xin X, Qin F. Decomposition of agricultural labor productivity growth and its Economics and a Professor at Vilnius University. He holds PhD degrees from Vilnius
regional disparity in China. China Agricultural Economic Review 2011;3(1): University and University of Copenhagen. His research focuses on efficiency and pro­
92–100. ductivity analysis, agricultural economics and energy economics. He has published in such
[25] Wang S, Huang J, Wang X, Tuan F. Are China’s regional agricultural productivities outlets as Energy Economics, Energy Policy, China Economic Review and European
converging: how and why? Food Pol 2019;86. Journal of Operational Research among others. E-mail: tomas@laei.lt
[26] Mugera AW, Langemeier MR, Featherstone AM. Labor productivity growth in the
Kansas farm sector: a tripartite decomposition using a non-parametric approach. Tianxiang Li is an Associate Professor at Nanjing Agricultural University. His research
Agric Resour Econ Rev 2012;41(1203–2016-95370):298–312. focuses on Agricultural Economics, Efficiency and Productivity Analysis. He has published
[27] Dorward A. Agricultural labour productivity, food prices and sustainable in such outlets as Applied Energy, Journal of Cleaner Production, Technological and
development impacts and indicators. Food Pol 2013;39:40–50. Economic Development of Economy, China Agricultural Econnomic Review and Sustain­
ability among others. E-mail: garylee0428@gmail.com

10