Total Factor Productivity of Herdsmen Animal Husbandry in Pastoral Areas: Regional Differences and Driving Factors

Abstract

In the context of China’s implementation of the rural revitalization strategy, it is essential to study the total factor productivity of animal husbandry in pastoral areas under the grassland ecological compensation policy, which is essential for promoting the harmonious development of animal husbandry production and grassland ecology in pastoral areas and helping the rural revitalization strategy. Based on the survey data of pastoral areas in Inner Mongolia, this paper measured and comparatively analyzed the differences in the changes in total factor productivity of pastoral households in each region and its convergence and discussed the main factors driving the total factor productivity of animal husbandry. The results of the study show that: (1) Except for Ulanqab City, the annual average total factor productivity of animal husbandry in the region as a whole and in each region is greater than 1, indicating that the animal husbandry production level of herdsmen has been improved to some extent during the policy implementation period. From the phased situation, the overall total factor productivity of animal husbandry in the Inner Mongolia region shows a characteristic of decreasing first and then increasing, while each region shows a different trend of change. (2) In terms of convergence, there is a certain degree of convergence during the policy period for both the region as a whole and each region, indicating that as the grassland compensation policy advances, the spatial differences in herdsmen total factor productivity in animal husbandry show a trend of gradual reduction, and the overall sample represents the sample of all the investigated areas. (3) In terms of driving factors, herdsmen education level, the degree of travel convenience, the degree of by-business, whether they participate in the subsidy policy, and whether they are fined have significant positive effects on their total factor productivity in animal husbandry, while the family dependency ratio and the degree of government regulation have significant negative effects on total factor productivity in animal husbandry. This paper takes the total factor productivity of animal husbandry in pastoral areas as the starting point, providing a new perspective for the research on the effect of the grassland ecological compensation policy. At the same time, it expands the driving factors of total factor productivity in animal husbandry. The conclusion provides a reference for improving the grassland ecological compensation policy and coordinating the harmonious development of production, life, and ecology in pastoral areas.

1. Introduction

In the context of the global pursuit of green development, sustainable economic growth is the goal pursued by all economic entities. As a large agricultural country, China has always been a topic of concern and discussion for scholars. With the development of China’s economy, the contradiction between the limited resource environment and people’s growing material needs is becoming more and more prominent, and the ecological environment has become more and more restrictive to economic development. As an important part of agriculture, animal husbandry in grassland pastoral areas also faces the same problem. The report of the 19th Party Congress proposed a major decision to implement the rural revitalization strategy. China has nearly 400 million hectares of various types of natural grasslands, accounting for 2/5 of the national territory [1]. Pasturing areas have an important strategic position in the overall economic and social development of China. To achieve the goal of a strong socialist modernization country, the revitalization of pasturing areas is imperative [2]. The State Council pointed out in the “Opinions on Promoting the High-Quality Development of Animal Husbandry” issued in 2021, that animal husbandry is an important industry related to the people’s livelihood, and in recent years, the comprehensive production capacity of China’s animal husbandry has been increasing, playing an important role in ensuring national food security, prospering the rural economy, and promoting farmers and herdsmen to increase their income. In 2022, the Central Committee of the Communist Party of China (CPC) and the State Council proposed to strengthen the basic support of modern agriculture in the “Opinions on the Key Work of Comprehensive Promotion of Rural Revitalization in 2022”, while the livestock production of herdsmen, as the basic industry of pasturing areas, still has an irreplaceable role in promoting the sustainable development of pasturing areas and improving the quality of life of herdsmen [3]. As one of the main fronts in the construction of ecological civilization in China, grasslands are not only an important material basis for farmers and herdsmen, who mainly raise livestock to maintain their production and livelihood, but also have an essential role in playing ecosystem service functions and as an ecological security barrier in the northern regions of China. Since most of China’s pasturing areas belong to arid and semi-arid climates with extremely fragile grassland ecosystems, pasturing areas have long been faced with the development of a trade-off between the economic production of livestock and the protection of ecological values [4]. However, grassland ecological services are powerful and benefit a large number of people with strong externalities, which are typically public goods [5]. Therefore, policy intervention is an important means to curb grassland degradation and protect and restore grassland resources [6]. As the most significant compensation policy in China’s pasturing areas, the grassland ecological compensation policy is a policy that has the largest scale of investment, covers the widest area, and involves the largest number of farmers and herdsmen. In 2011, the state began to establish a comprehensive grassland ecological conservation subsidy and award mechanism, and in 2016, the government issued the “Guidance on the Implementation of the New Round of Grassland Ecological Conservation subsidy and award policy”, which further improved the compensation standard of the grassland ecological conservation subsidy and award policy. Research on the grassland ecological conservation subsidy and award policy has been conducted by various scholars from the aspect of policy mechanism development to the evaluation of policy implementation results [7,8].

According to the above research background, the grassland resources in pastoral areas still mainly rely on the operating income from livestock production of herdsmen as the main source of livelihood. The stable and sustainable development of livestock production is the long-term goal pursued by herdsmen. Therefore, confronting the severe status quo of the continuous reduction of high-quality pasture and the deterioration of grassland vegetation structure, the purpose of this study is to study how the changing trends and final state of the differences in the efficiency of the factors of input and output of herdsmen animal husbandry production develop during the implementation of the policy and what driving factors will affect them. It is of great significance to measure and analyze the characteristics of the total factor productivity of herdsmen animal husbandry, objectively understand the level of development of herdsmen animal husbandry in grassland pasturing areas, improve the grassland ecological compensation policy, and realize the harmonious development of herdsmen animal husbandry production and grassland ecology.

Total factor productivity (TFP), as one of the most important indicators reflecting the economic effect of production and operation in a certain industry, is able to integrate multiple indicators of each factor of production. TFP comprehensively reflects the changes in the overall efficiency of its production when studying issues related to production and operation [9]. The methods used by academics in the study of total factor productivity, technical efficiency, etc., are divided into two main categories: parametric and nonparametric estimation methods [10]. Compared with the parametric estimation method, the nonparametric estimation method does not make assumptions about the basic distribution and does not set the production function. It mainly uses the information of the random sampling itself to obtain the optimal estimator from the perspective of input and output. Therefore, some scholars tend to use nonparametric estimation methods to measure production efficiency, among which data envelopment analysis (DEA) is widely used [11]. The data envelopment analysis method was developed by A. Charnes, who was the first to propose it in 1978 [12]. Compared with other measurement methods, DEA method is characterized by its ability to effectively identify multiple inputs and outputs with relative efficiency, so it has been widely used by scholars in many research fields [13,14,15]. Subsequently, Fare et al. combined Malmquist productivity index, proposed by Malmquist in 1953, with DEA theory to form DE–Malmquist index method [16]. Marx (1953) pointed out that one should examine and analyze not only the social productivity of labor, but also its natural productivity [17]. Therefore, the analysis and exploration of productivity in agriculture and animal husbandry are essential. The improvement of agricultural TFP is crucial to the sustainable development of agriculture, as shown by McMillan J (1989), Yin Chao Jing (2022), Li Qiang (2020), and other scholars, who use the macro perspective to evaluate and study China’s agricultural TFP [18,19,20]. At present, when most scholars at home and abroad study total factor production efficiency in agriculture, they usually use inputs such as land, labor, and capital as production input indicators and quantity or income yield as output indicators [21,22,23]. In the driving factors of agricultural total factor productivity, it is found that agricultural total factor productivity will be affected by other subjective or objective variables, including natural factors, institutional factors, economic factors, and social factors, in addition to the input and output indicators of production factors [24,25]. In addition, the improvement of agricultural total factor productivity is also affected by the personal resource endowment of farmers and herdsmen, such as their own age and education level [26,27].

To sum up, there is little research on the combination of grassland ecological compensation policy and total factor productivity of animal husbandry in pastoral areas by scholars at present. The paper takes the relationship between the two as the starting point and discusses the implementation effect of grassland compensation policy from the perspective of total factor productivity of herdsmen’s animal husbandry. Based on this, from the perspective of the micro main body of herdsmen, the paper measures and compares the change and convergence of the total factor productivity of animal husbandry in various pastoral areas in Inner Mongolia, by using field survey data of pastoral areas in Inner Mongolia, and discusses its driving factors, with a view to providing some reference for coordinating the harmonious development of production, life, and ecology in pastoral areas and assisting the revitalization and modernization of pastoral areas. The overall structure of the introduction is shown in Figure 1.

2. Materials and Methods

2.1. Data Sources

The data in this paper were obtained from a field survey of herdsmen, mainly engaged in animal husbandry production before and after the implementation of the first and second rounds of grassland ecological conservation rewards and incentive policy in the Inner Mongolia Autonomous Region, including 2010, 2015, and 2020. In order to maintain the consistency of the sample, all research areas were selected from purely pasturing banners in Inner Mongolia. The research was conducted on the basis of the actual development status and grassland type distribution of each banner and county, using a combination of random sampling and typical sampling for data collection, and the scope of the research included 5 leagues, 8 pure pasturing banners and 23 sumus (towns) in Inner Mongolia, including Hulunbeier City, Xilin Gol League, Chifeng City, Tongliao City and Ulanqab City, with a total of 623 research questionnaires. Excluding the missing samples of important indicators, the remaining valid samples were 536, and this distribution status is shown in

Table 1. Sample Distribution in the Survey Area.

Researching Cities	Research Banner County	Total Sample of League Cities (Households)	Sample Herdsmen Households (Households)	Proportion (%)
Hulunbeier City	Chenbaerhu Banner	139	62	11.57
	Xinbalhoo Left Banner		77	14.36
Xilin Gol League	Abaga Banner	135	78	14.55
	Sunit Right Banner		57	10.63
Chifeng City	Keshiketeng Banner	114	53	9.88
	Balingyu Banner		61	11.38
Ulanqab	Siziwang Banner	69	69	12.88
Tongliao City	Zalute Banner	79	79	14.74
Total	-	536	536	100

.

2.2. Research Methodology

2.2.1. DEA–Malmquist Index Approach

In this paper, DEA–Malmquist index method is used to calculate the total factor productivity of herdsmen in pastoral areas. In the case of constant scale payoffs, it can be decomposed into the integrated technical efficiency change index and the technical progress index, and in the case of variable scale payoffs, the integrated technical efficiency can be further decomposed into the pure technical efficiency index and the scale efficiency index. The specific formula of DEA–Malmquist index for period t + 1 relative to period t is as follows:

$$\begin{array}{ll}{M}_0(x^{{}_{t+1}},y^{{}_{t+1}},x^{{}_t},y^{{}_t})& ={\left[\frac{D_o^t(x^{{}_{t+1}},y^{{}_{t+1}})}{D_o^t(x^{{}_t},y^{{}_t})}\times \frac{D_o^{t+1}(x^{{}_{t+1}},y^{{}_{t+1}})}{D_o^{t+1}(x^{{}_t},y^{{}_t})}\right]}^{1/2}\\ & =\frac{D_o^{t+1}(x^{{}_{t+1}},y^{{}_{t+1}})}{D_o^t(x^{{}_t},y^{{}_t})}\times {\left[\frac{D_o^t(x^{{}_{t+1}},y^{{}_{t+1}})}{D_o^{t+1}(x^{{}_{t+1}},y^{{}_{t+1}})}\times \frac{D_o^t(x^{{}_t},y^{{}_t})}{D_o^{t+1}(x^{{}_t},y^{{}_t})}\right]}^{1/2}\\ & =\frac{D_{vrs}^{t+1}(x^{{}_{t+1}},y^{{}_{t+1}})}{D_{vrs}^t(x^{{}_t},y^{{}_t})}\times \frac{D_o^{t+1}(x^{{}_{t+1}},y^{{}_{t+1}})/D_{vrs}^{t+1}(x^{{}_{t+1}},y^{{}_{t+1}})}{D_o^t(x^{{}_t},y^{{}_t})/D_{vrs}^t(x^{{}_t},y^{{}_t})}\times {\left[\frac{D_o^t(x^{{}_{t+1}},y^{{}_{t+1}})}{D_o^{t+1}(x^{{}_{t+1}},y^{{}_{t+1}})}\times \frac{D_o^t(x^{{}_t},y^{{}_t})}{D_o^{t+1}(x^{{}_t},y^{{}_t})}\right]}^{1/2}\end{array}$$

(1)

where $D_o^t$ denotes the output distance function for period t, $D_o^{t+1}$ denotes the output distance function for period t + 1, x refers to the input, and y refers to the output.

2.2.2. Convergence Study Method

(1) σ convergence. In the paper, using the coefficient of variation as a model of σ convergence, if CV_t > CV_t+1, there is σ convergence in the total factor productivity differences of herdsmen animal husbandry in the study area, indicating that the differences between the total factor productivity of different herdsmen animal husbandry narrows and has a tendency to converge and develop in the presence of objective differences in natural resource conditions [28]; conversely, there is no σ convergence and a tendency to divergence.

$$C{V}_t=\sqrt{\frac{1}{n}{{\displaystyle \sum _{i=1}^n\left[TF{P}_{it}-\overline{TF{P}_{it}}\right]}}^2}/\overline{TF{P}_{it}}$$

(2)

where TFP_it is the total factor productivity of sample herdsmen household i in year t, and n is the sample size of the study area.

(2) β convergence. Absolute β convergence tests whether there is a catch-up effect of herdsmen with lower total factor productivity for those with higher total factor productivity, eventually converging and being in the same steady state [29].

$$\frac{1}{T}\ln (TF{P}_{it+T}/TF{P}_{it})=\alpha +\beta \ln (TF{P}_{it})+{\epsilon }_{it}$$

(3)

where β is the parameter to be estimated and is denoted as the convergence coefficient. If β is negative, it means that there is absolute convergence characteristic of total factor productivity of herdsmen livestock without the existence of herdsmen heterogeneity, and vice versa without convergence [30], and ε is the error term.

$$\ln (TF{P}_{it})-\ln (TF{P}_{it-1})=\alpha +\beta \ln (TF{P}_{it-1})+{\mu }_i$$

(4)

where α is a constant term, and β is the parameter to be estimated. When β < 0, it indicates the existence of conditional β convergence subject to herdsmen household heterogeneity, and i denotes a fixed effect of livestock production in different league cities.

2.3. Description of Input–Output Indicators

This paper combines the actual situation of local herdsmen’s animal husbandry production and operation to construct an index system for assessing total factor productivity in animal husbandry (see

Table 2. Input–output indicators and descriptive statistics.

Type	Variables	Variable Interpretation and Values	Unit	Mean Value	Standard Deviation
Output indicators	Number of livestock slaughtered	Number of livestock slaughtered, that is, the sum of the number of cattle, sheep, horses, and other livestock	Sheep unit	101.67	110.27
	Total operating income	Including income from livestock slaughter, income from pasture rental, income from pasture sale, income from by-products such as wool, sheepskin, and milk	CNY 10,000	13.43	15.93
Input indicators	Actual pasture area operated	Actual pasture contracted area + leased-in pasture area–leased-out pasture area	Hectare	319.51	335.46
	Labor input	Number of herding households engaged in animal husbandry	Person	2.31	0.83
	Number of stocked livestock	Number of livestock in stock at the beginning of the year, that is, the sum of the number of cattle, sheep, horses, etc.	Sheep unit	318.71	274.52
	Animal husbandry cost input	The sum of forage input cost, pasture rental cost, labor cost, medical cost, grazing cost, and fuel cost	CNY 10,000	7.48	10.33

) and in the selection of input and output indicators for measuring the total factor productivity in herdsmen’s animal husbandry, The actual operating pasture area, labor input, number of livestock in stock, and animal husbandry production cost input are used as input indicators of herdsmen’s animal husbandry production, and the number of livestock slaughtered and total operating income are the main output indicators [21,31]. Among them, pasture input, as an important production material for herdsmen’s animal husbandry production, provides the basis for the allocation of other production factors [32]. The input of actual pasture area includes the sum of pasture area owned and rented by herdsmen minus the area of rented pasture. Labor input specifically refers to the number of people engaged in animal husbandry in the household. The number of livestock in stock refers to the number of livestock in stock at the beginning of the year, mainly comprising the number of beef cattle, dairy cows, sheep, goats, and horses and other livestock converted in accordance with the sheep unit. The cost input of animal husbandry mainly includes forage input cost, pasture rental cost, labor cost, medical cost, grazing cost, and fuel cost, etc., among which, forage input cost mainly includes fodder grass and forage salt purchase cost, artificial forage land cost, etc.; labor cost mainly includes machine cost for haying (haying, baling, and hauling, etc.), sheep washing and shearing labor cost, etc.; medical expenses mainly include expenses for medical treatment and epidemic prevention for sheep, cattle, etc. The number of slaughtered livestock mainly includes the number of slaughtered livestock such as beef cattle, dairy cattle, sheep, goats, and horses converted according to sheep units; the total operating income mainly includes income from slaughtered livestock, income from pasture rental, income from pasture sale, and income from by-products such as wool, sheepskin, and milk. The number of livestock is uniformly converted according to sheep units, and the conversion standard is: 2 young animals = 1 adult animal [33], 1 sheep = 1 sheep unit, 1 cow = 5 sheep units, and 1 horse = 6 sheep units (conversion according to the method of commutation of sheep units in the by-laws of the “Regulations on the Protection of Basic Grasslands in Inner Mongolia Autonomous Region”). The interpretation and values of specific input and output indicators of herdsmen are shown in Table 2.

2.4. Description of Variables of Driving Factors

Herdsmen’s animal husbandry production is characterized by individual heterogeneity, and herdsmen’s own resource endowment and external policy and environmental conditions are the main driving factors leading to the efficiency of herdsmen’s animal husbandry production factor input–output allocation [34,35,36]. With reference to the existing research results, four dimensions representing the basic characteristics of herdsmen’s own and external conditions, such as personal and household characteristics, transportation factors, economic factors, and subsidy policy factors, were selected in this paper [34,35,36], so as to study the driving factors of the total factor productivity in herdsmen’s animal husbandry, specifically including gender, age, education level, family dependency ratio, distance from the household to the marketplace, ease of travel, by-business degree, compensation amount, whether to implement the subsidy and award policy, the intensity of government supervision and whether to be fined, and other variables.

Table 3. Variable definition and descriptive statistics.

Type	Indicator	Specific Variable Interpretation and Values	Unit/Assignment	Frequency (%)	Mean Value	Standard Deviation
Personal and household characteristics	Gender	0 = Female,1 = Male	Female = 0	6.95	0.93	0.25
			Male = 1	93.05
	Age	Age of household head	Age	-	47.79	10.66
	Education level	1= Primary school and below, 2= junior high school, 3= senior high school, 4= technical secondary school or junior college, 5= undergraduate degree or above	Primary school and below = 1	26.01	2.14	1.11
			Junior high school = 2	50.89
			Senior high school = 3	15.02
			Technical secondary school or junior college = 4	4.85
			Undergraduate degree or above = 5	3.23
	Family dependency ratio	Number of dependents and supporters/total family size	%	-	0.39	0.24
Transportation factors	Distance to the market	Distance of the household from the nearest market	1000 m	-	27.70	57.60
	Travel convenience	Whether there are travel vehicles, No = 0, Yes = 1	No = 0	32.96	0.67	0.47
			Yes = 1	67.04
Economic factors	By-business	Non-livestock income/total income for herdsmen	By-business degree = 0	71.57	0.09	0.21
			(0, 50%]	22.13
			(50%, 100%]	6.30
	Compensation amount	Grassland ecological bonus standard × grassland contracted area	CNY 10,000	-	2.22	2.37
Subsidy and award policy factors	Whether to implement the subsidy and award policy	Whether the herdsmen carry animal husbandry production in accordance with the load stipulated in the policy, No = 0, Yes = 1	No = 0	70.34	0.30	0.46
			Yes = 1	29.66
	Government regulation	Has the government ever regulated herdsmen’s animal husbandry production; 1 = Not at all, 2 = Rarely, 3 = Some, 4 = Yes, often, 5 = Yes, very strict	Not at all = 1	15.19	2.51	0.93
			Rarely = 2	33.76
			Some = 3	37.00
			Yes, often = 4	13.25
			Yes, very strict = 5	0.81
	Whether to be fined	Whether the herdsman was fined for overloading; No = 0, Yes = 1	No = 0	62.36	0.38	0.48
			Yes = 1	37.64

shows that among the personal and family characteristics, the average age of the household head is 47.79, with a higher proportion of males, and the average household dependency ratio is 0.39. The overall education level of the surveyed herding households is mainly concentrated in primary and junior high school education, accounting for 76.9%, and the overall education level is low. The transportation condition shows that the average household distance from the nearest market is 27.7 km, which is a moderate distance, and 67.04% of herding households have vehicles, so the transportation condition is convenient, but the degree of dispersion is high, and the difference among herding households is large. The results of economic conditions show that 71.57% of herding households took animal husbandry production as their only source of income during the policy implementation period, and only less than 30% of herding households had some degree of by-business income, and animal husbandry production was still the main source of stable income for herding households. In terms of compensation amount, the average compensation amount of households is CNY 22,200, but the dispersion is high and the compensation amount among herding households is widely varied. It can be seen from the conditions of the subsidy and award policy that the percentage of herding households that carry out animal husbandry production according to the prescribed load during the policy period is 29.66%, and 70.34% of herding households do not actually implement the subsidy and award policy and have overloading behavior; the percentage of herding households fined for overloading is 37.64%, and 84.82% of herding households believe that the government has supervised herding households, with a mean value of 2.51, which is between “have, rarely” and “have some”; it can be seen that the policy participation of herding households during the implementation of the policy is not sufficient, and the overall government supervision is moderate to low.

3. Results and Analysis

3.1. Analysis of Regional Differences in Total Factor Productivity of Animal Husbandry

3.1.1. Analysis of Regional Differences in Total Factor Productivity of Livestock Husbandry

Using DEAP 2.1 software, this paper decomposes and measures the total factor productivity (TFP) of animal husbandry of the sample overall and regional herdsmen during the implementation of the grassland ecological conservation subsidy and award policy, with 2010 as the base period using short panel data for three years, 2010, 2015, and 2020, by the DEA–Malmquist index approach (see

Table 4. Total factor productivity of herdsmen’s animal husbandry and its decomposition.

League	Code	Comprehensive Technical Efficiency	Technical Progress	Pure Technical Efficiency	Scale Efficiency	Total Factor Productivity
Overall	ZT	1.689	0.628	1.335	1.265	1.062
Hulunbeier City	HL	1.418	0.809	1.364	1.04	1.148
Xilin Gol League	XL	1.441	0.786	1.34	1.076	1.133
Chifeng City	CF	2.091	0.507	1.408	1.485	1.061
Tongliao City	TL	2.067	0.496	1.367	1.512	1.024
Ulanqab City	WL	1.557	0.613	1.207	1.29	0.954

). On the whole, the total factor productivity (TFP) of herdsmen animal husbandry in Inner Mongolia region during the policy period was 1.062, with an average growth rate of 6.2%, and the overall TFP of the sample of 536 herdsmen was relatively concentrated, and the distribution of herdsmen with TFP > 1 was somewhat more dispersed compared to herdsmen with TFP < 1 (see Figure 2). The comprehensive technical efficiency was 1.689, which was the main driving force of total factor productivity growth of herdsmen’s animal husbandry in Inner Mongolia, while the technical progress was 0.628, with an average growth rate of −34.2%, showing a negative downward trend to the overall growth. Except for Ulanqab City, the average annual animal husbandry TFP of the sample herding households in each region during 2010–2020 is greater than 1, indicating that the level of animal husbandry production in each region has been improved to some extent, with Hulunbeier City and Xilin Gol League showing a greater overall increase and Chifeng City and Tongliao City showing a minor increase in animal husbandry TFP. From the results of the decomposition of Malmquist index, Hulunbeier City, Xilin Gol League, Chifeng City, and Tongliao City are similar to the overall situation of the sample, and the progress of comprehensive technical efficiency has positively contributed to the growth of total factor productivity in their animal husbandry industry, in which pure technical efficiency increased by 36.4%, 34.0%, 40.8%, and 36.7%, respectively, and scale efficiency increased by 4.0%, 7.6%, 48.5%, and 51.2%, respectively. Hulunbeier City and Xilin Gol League mainly rely on the pull of pure technical efficiency, while Chifeng City and Tongliao City mainly rely on the dual drive of pure technical efficiency and scale efficiency. Compared with the comprehensive technical efficiency, the technical progress of herdsmen’s animal husbandry production is less than 1, showing a negative growth, which indicate that herdsmen do not make enough efforts to improve and innovate the resource allocation and production-related technologies, hindering their total factor productivity improvement and progress in the process of animal husbandry production. In contrast, the average annual animal husbandry TFP growth rate during the policy period for the sample herdsmen in Ulanqab City was −4.6%, of which the comprehensive technical efficiency was 1.557, while the technical progress of other factors was 0.613, with an average growth rate of −38.7%, which was the main reason for the negative total factor productivity growth and production inefficiency of herdsmen’s animal husbandry. Figure 2 reports the kernel density of total factor productivity (TFP) in animal husbandry of the sample herdsmen in each league and City. It can be seen that the overall distribution of TFP of herdsmen in Xilin Gol League and Ulanqab City is relatively concentrated, while that of Hulunbeier City, Chifeng City, and Tongliao City are more discrete, and there are large disparities in TFP in animal husbandry among households, and the distribution of herdsmen with TFP > 1 is more dispersed than that of herdsmen with TFP < 1.

3.1.2. Total Factor Productivity in Animal Husbandry by Stage

The results of total factor productivity of animal husbandry in stages (see

Table 5. Total Factor Productivity of Animal Husbandry in Phased Herdsmen and its Decomposition.

League	Time Phase	Comprehensive Technical Efficiency	Technical Progress	Pure Technical Efficiency	Scale Efficiency	Total Factor Productivity
Overall	2010–2015	1.241	0.779	1.009	1.229	0.967
	2015–2020	2.300	0.507	1.767	1.301	1.166
Hulunbeier City	2010–2015	0.839	1.373	0.861	0.975	1.153
	2015–2020	2.397	0.477	2.161	1.109	1.143
Xilin Gol League	2010–2015	0.629	1.699	0.659	0.955	1.069
	2015–2020	3.301	0.363	2.724	1.212	1.200
Chifeng City	2010–2015	2.368	0.485	1.481	1.599	1.149
	2015–2020	1.846	0.531	1.339	1.379	0.980
Tongliao City	2010–2015	1.382	0.533	1.063	1.3	0.736
	2015–2020	3.091	0.461	1.758	1.758	1.426
Ulanqab City	2010–2015	1.702	0.476	1.173	1.451	0.811
	2015–2020	1.424	0.788	1.243	1.146	1.123

) show that the overall total factor productivity of animal husbandry of herdsmen in 2010–2020 shows the characteristics of first decline and then rise, and the overall TFP of animal husbandry of herdsmen in 2010–2015 (before and after the first round of subsidy and award policy) is 0.967, which is 3.3% lower on average than that in 2010. Among them, the comprehensive technical efficiency is 1.241, which can directly promote the growth of total factor productivity of herdsmen’s animal husbandry, while the average growth rate of technological progress is −22.1%, causing the negative growth of total factor productivity. TFP in animal husbandry rose to 1.166 during 2015–2020 (before and after the second round of subsidies and awards policy), up 19.9% compared to the period 2010–2015, where the comprehensive technical efficiency was 2.300, and the index of change in comprehensive technical efficiency of herdsmen also reached its highest point during this period, increasing by 105.9%, which is the main driver of total factor productivity growth in herding households’ livestock industry during the second round of policy implementation in Inner Mongolia, while the average growth rate of technological progress is −27.2%, showing a downward trend of negative growth for the overall sample. In terms of the phased situation in each region, TFP of livestock husbandry in Hulunbeier shows a characteristic of rising and then slightly declining, with an average growth rate of 15.3% from 2010 to 2015, which is mainly influenced by the growth of technological progress. The small decline seen in 2015–2020 was mainly the result of the offset between the significant rise in comprehensive technical efficiency and the decline in technical progress, indicating that the resources for herdsmen’s animal husbandry production in the region need to be further optimally allocated. In contrast, Chifeng showed a rise followed by a significant decline, with the overall animal husbandry TFP for 2010–2015 increasing by an average of 14.9% compared to 2010, mainly relying on the pull of comprehensive technical efficiency. During 2015–2020, animal husbandry TFP declined to 0.980, mainly due to a double decline in pure technical efficiency and scale efficiency. Xilin Gol League, on the other hand, showed a sustained rise, with an average annual growth rate of 6.9% in TFP from 2010–2015 and a sustained rise to 1.200 in the period 2015–2020, which was mainly driven by both pure technical efficiency and scale efficiency in this period. The situation in Tongliao and Ulanqab is similar to the sample overall, both showing the trend of decline first and then rise, with animal husbandry TFP less than 1 in 2010–2015 and the average growth rates of −26.4% and −18.9%, respectively, where the negative growth of technical progress is the main reason for the inefficiency of herdsmen’s animal husbandry production. The period of 2015–2020, on the other hand, showed a significant increase. From the results of technical progress in animal husbandry production during the implementation of the first and second rounds of subsidy and award policy, the overall technical progress in the process of animal husbandry production of herdsmen was less than 1, which hindered the improvement and progress of the total factor productivity in animal husbandry.

3.2. Analysis of Convergence of Total Factor Productivity in Animal Husbandry

3.2.1. σ-Convergence Analysis

In order to understand the trend and final state of total factor productivity of herdsmen’s animal husbandry during the implementation of the subsidy and award policy, convergence analysis is introduced in this paper. The σ convergence results measured according to Equation (2) are shown in

Table 6. Total factor productivity of herdsmen’s animal husbandry σ convergence coefficient.

League/City	2010–2015	2015–2020
Overall	0.6595	0.6388
Hulunbeier City	0.5439	0.6774
Xilin Gol League	0.3375	0.5715
Chifeng City	0.8952	0.5957
Tongliao City	0.5690	0.6968
Ulanqab City	0.7599	0.5264

. From the overall sample, the σ convergence coefficients before and after the implementation of the first and second rounds of the subsidy and award policy are 0.6595 and 0.6388, respectively, showing a small σ convergence as a whole, with a certain magnitude of convergence development trend. From the σ convergence results of each region, Ulanqab City and Chifeng City both showed significant σ convergence, with σ convergence coefficients decreasing from 0.7599 to 0.5264 and 0.8952 to 0.5957, respectively, while the overall total factor productivity σ convergence coefficients of Hulunbeier City, Xilin Gol League, and Tongliao City showed a minor increase. The total factor productivity of herdsmen’s animal husbandry showed a divergence trend during this period. Therefore, during the implementation of the subsidy and award policy, Inner Mongolia showed a slight σ-convergence as a whole, with σ-convergence in Ulanqab and Chifeng, and no σ-convergence in Hulunbeier, Xilin Gol League, and Tongliao.

3.2.2. β-Convergence Analysis

In order to investigate whether the total factor productivity differences in herdsmen’s animal husbandry shrink with the implementation of the subsidy and award policy and eventually converge to a corresponding stable state, the absolute β and conditional β convergence were further analyzed in this paper, and the results are shown in

Table 7. Livestock total factor productivity of herdsmen β convergence test.

Survey Region	Convergence Type	Constant Term a	Β Convergence Coefficient	Standard Error	Adj R2
Overall	Absolute β convergence	−0.0164 ***	−0.2290 ***	0.1422	0.6035
	Condition β convergence	−0.0364 ***	−0.2324 ***	0.1481	0.6043
Hulunbeier City	Absolute β convergence	−0.0222 *	−0.2009 ***	0.17351	0.2828
	Condition β convergence	−0.0957 *	−0.2139 ***	0.1716	0.2984
Xilin Gol League	Absolute β convergence	0.0056	−0.2344 ***	0.1282	0.2321
	Condition β convergence	−0.0332 *	−0.2357 ***	0.1293	0.2183
Chifeng City	Absolute β convergence	−0.0472 ***	−0.2345 ***	0.1108	0.8443
	Condition β convergence	−0.0431 **	−0.2360 ***	0.1124	0.8469
Tongliao City	Absolute β convergence	0.0157	−0.2016 ***	0.1652	0.2605
	Condition β convergence	0.0689 **	−0.2462 ***	0.1925	0.3251
Ulanqab City	Absolute β convergence	−0.0374 **	−0.2506 ***	0.1042	0.8637
	Condition β convergence	0.0861 **	−0.2227 ***	0.0881	0.8749

Note: *, **, and *** indicate significant at the 10%, 5%, and 1% levels, respectively.

. The absolute β convergence coefficient for the overall sample during the implementation of the subsidy and award policy was −0.2290, showing absolute β convergence at the 1% significance level. The absolute β convergence coefficients of all regions, including Hulunbeier City, Xilin Gol League, Chifeng City, Tongliao City, and Ulanqab City, were significantly negative during the implementation of the policy, indicating that the total factor productivity of herdsmen with slow development in animal husbandry production grew significantly faster than that of herdsmen with better development. There is a “catch-up effect” of herdsmen with lower TFP in animal husbandry to those with higher TFP, and their total factor productivity levels will eventually reach the same steady-state level. Further, the conditional β convergence of total factor productivity of herdsmen’s animal husbandry is analyzed, while controlling for herdsmen’s heterogeneity characteristics. The results showed that the conditional β convergence coefficients of the overall Inner Mongolia sample and each region including Hulunbeier City, Xilin Gol League, Chifeng City, Tongliao City, and Ulanqab City exhibited conditional β convergence characteristics at 1% significance level during 2010–2020, further indicating that the spatial differences in animal husbandry TFP of each herdsmen showed a trend of gradual reduction as the policy continued to advance, and the overall convergence to their respective stable levels would gradually accelerate, with a convergent characteristic of convergence in development.

3.3. Analysis of Driving Factors

Since the dependent variable of this study, i.e., total factor productivity in animal husbandry, is a restricted variable, the Tobit regression model was chosen to further investigate the factors influencing its total factor productivity in animal husbandry using the great likelihood method [37]. The correlation coefficients between the variables were used in this paper to test whether there was multicollinearity, and it was concluded that the correlation coefficients between the influencing factors were all less than 0.8, indicating that there was no multicollinearity between the influencing factors and that fixed panel effect analysis could be performed, and the regression results are shown in

Table 8. Analysis on driving factors of livestock total factor productivity of herdsmen.

Type	Variables	Estimated Coefficient	Standard Error	Z Value
Personal and household characteristics	Gender	0.131	0.127	1.04
	Age	0.003	0.005	0.56
	Education level	0.050 **	0.046	2.11
	Family dependency ratio	−0.379 *	0.203	−1.87
Transportation factors	Distance from home to the marketplace	−0.0001	0.000	−0.36
	Convenience of travel	0.004 **	0.105	2.03
Economic factors	Amount of compensation	−0.029	0.025	−1.16
	Degree of by-business	0.023 **	0.291	2.38
Subsidy and award policy factors	Whether or not to implement the subsidy and award policy	0.052 **	0.104	2.05
	Whether to be fined	0.050 **	0.098	2.03
	Degree of government regulation	−0.132 ***	0.050	−3.67
	Constant items	1.400 ***	0.335	4.17
	Prob > chi2	0.000

Note: *, **, and *** indicate significant at the 10%, 5%, and 1% levels, respectively.

. The Wald test passed the significance level of 1%, indicating that the overall effect of the model was positive.

Among the personal and household characteristics variables of herdsmen, the educational level has a significant positive effect on the total factor productivity in herdsmen’s animal husbandry, which is significant at the 5% level. It shows that the improvement of the herdsmen’s cultural level has a strong role in promoting their total factor productivity. In the process of balancing economic and ecological benefits, herdsmen with higher education tend to maintain sustainable development of animal husbandry economy while protecting grassland ecology. Therefore, while implementing the grassland ecological compensation policy, the highly educated herdsmen adjust the input and output factors of livestock production, so as to improve their total factor productivity of livestock industry. The family dependency ratio shows a significant negative effect, i.e., herdsmen with higher family burdens have heavier economic pressure and insufficient labor input. Therefore, it has a negative impact on the total factor productivity of animal husbandry.

Among the transportation factors, the driving effect of the distance between the family and the market is not significant. The convenience of travel has a significant positive effect on the total factor productivity of herdsmen’s animal husbandry, because the more convenient the travel conditions of herdsmen, the better the access to national real-time policy and market change information. Under the restriction of their living environment and resource endowment, herdsmen can make the best choice to adjust the results of resource allocation in animal husbandry production, so as to effectively promote the improvement of their total factor productivity in animal husbandry.

In terms of economic factors, the degree of by-business passed the 5% significance level with a positive coefficient. With other conditions unchanged, the increase in the proportion of non-livestock income has a strong promoting effect on the improvement of herdsmen’s animal husbandry production input–output efficiency. By increasing the proportion of non-livestock income, it can reduce herdsmen’s dependence on traditional animal husbandry production, reduce the pressure of overloading and overgrazing to maintain necessary livelihoods, and make herdsmen consider the sustainability of animal husbandry production while focusing on economic output, thus improving the total factor productivity of animal husbandry production. The amount of grassland ecological compensation has no significant impact on the total factor productivity of livestock husbandry. It can be seen that a single amount of compensation is difficult to cover the economic losses of herdsmen caused by the implementation of the compensation policy, and the grassland ecological environment cannot be effectively protected.

From the perspective of the subsidy and award policy factors, whether or not implementing the subsidy and award policy has a significant positive impact on the total factor productivity in herdsmen’s animal husbandry, i.e., other conditions being unchanged, herdsmen’s animal husbandry production behavior in accordance with the policy load can effectively improve their animal husbandry input–output efficiency, indicating that herdsmen who actually implement the policy during the effective period of subsidy and award policy can improve their animal husbandry production level to a certain extent. Herdsmen who were fined for significantly overloading increased their total factor productivity in animal husbandry, and the greater the degree of positive regulation on herdsmen, the more it was detrimental to their total factor productivity, and the results showed that herdsmen had “loss avoidance” behavior [38], i.e., herdsmen had different sensitivities to losses and gains; relative to unknown gains, herdsmen are more sensitive to known losses. Therefore, reasonable fines can make herdsmen allocate resources to production inputs and outputs more carefully, while strictly strengthening the degree of government regulation cannot effectively promote the total factor productivity of herdsmen’s animal husbandry.

4. Discussion

Herdsmen’s animal husbandry production is a dynamic process of systematic inputs and outputs, so combining inputs and outputs of animal husbandry production factors for comprehensive comparative analysis is of great practical significance for a comprehensive understanding of herdsmen’s animal husbandry production levels in pasturing areas, guiding and encouraging herdsmen to participate in the subsidy and award policy, and protecting the grassland ecological environment.

From the measured results and convergence of animal husbandry production efficiency, the average growth rate of total factor productivity of herdsmen’s animal husbandry in Inner Mongolia during the period of the subsidy and award policy is positive, and there is significant convergence, indicating that the overall animal husbandry resource allocation efficiency of the sample herdsmen has been improved with some stability. Among them, the progress of comprehensive technical efficiency plays a positive role in promoting their total factor productivity in animal husbandry. However, because herdsmen in pasturing areas still rely on the crude animal husbandry production method to maintain their livelihoods and do not make enough efforts to improve and innovate resource allocation and production-related technologies, the overall technological content of animal husbandry remains at a low level, making the overall technological progress present a state of less than 1, which hinders the improvement and progress of its production efficiency in the process of animal husbandry production. Therefore, while steadily improving the comprehensive technical efficiency, it is also necessary to increase the technical innovation of animal husbandry production in pasturing areas, so as to truly improve the level of total factor productivity, vigorously promote the research and development of key core technologies in pasturing around various links including forage production, livestock breeding, and livestock product processing, encourage the development of intensive and ecological-type breeding in animal husbandry, and explore the realization of a sustainable development model of animal husbandry in pasturing areas, so as to realize the steady improvement of the livestock production efficiency of herdsmen.

In addition, the total factor productivity of animal husbandry measured by the DEA–Malmquist index approach will be driven or constrained by herdsmen’s own resource endowment and external environmental conditions, in addition to their factor input–output indicators. From the driving factors of total factor productivity in animal husbandry, it can be seen that the education level of herdsmen in pasturing areas, transportation conditions, degree of by-business, and whether or not to implement the subsidy and award policy are all significantly and positively related to the degree of improvement of their animal husbandry. Therefore, the government should continue to optimize the subsidy and award policy, combine regional differences to appropriately increase the subsidy and award standard, and at the same time, explore diversified compensation methods, further strengthen the publicity work related to the subsidy and award policy, let herdsmen understand more deeply the ecological and economic benefits obtained by reducing livestock in accordance with the regulations, mobilize herdsmen’s enthusiasm to protect the grassland ecological environment, guide and encourage herdsmen’s families to actively implement the subsidy and award policy, and achieve sustainable improvement of livestock production while protecting grassland ecology. At the same time, the government should pay attention to the encouragement of herdsmen’s by-business behavior and improve herdsmen’s economic level by expanding diversified and stable income sources, so as to effectively regulate the allocation efficiency of herdsmen’s animal husbandry production inputs and outputs and promote the transformation and development of herdsmen’s animal husbandry production. In addition, the government should provide supporting measures, including improving the educational and cultural level of herdsmen, introducing highly educated and practical talents in farming and pasturing areas, and providing more convenient transportation conditions, so as to finally and effectively improve the resource allocation efficiency of herdsmen’s animal husbandry and provide a favorable external environment for the development of herdsmen’s animal husbandry production, thus promoting the sustainable development of the pasturing areas. By analyzing the driving factors of total factor productivity of herdsmen’s animal husbandry, this paper provides new ideas for effectively promoting the total factor productivity of herdsmen’s animal husbandry.

Due to various constraints, this paper uses some purely pasture-based banners in eastern and central Inner Mongolia as the survey area, which has certain limitations. This paper also uses short panel data from three periods of field research to measure the production efficiency of two periods, which has limitations in time dimension. Therefore, in a future study, we will further expand the sample size and sample scope on the basis of expanding the time dimension, in order to analyze the differences in spatial and temporal changes and trends in animal husbandry production efficiency of herdsmen in different regions in a more comprehensive and detailed way.

5. Conclusions

The analysis of total factor productivity measures, convergence, and the driving factors in herdsmen’s animal husbandry under regional heterogeneity led to the following conclusions:

(1) From a microscopic perspective, the total factor productivity of herdsmen’s animal husbandry is measured and analyzed, and the average growth rate of total factor productivity of herdsmen’s animal husbandry in Inner Mongolia is 6.2% during the subsidy and award policy period. From the perspective of each region, except for Ulanqab City, the average annual total factor productivity of the sample herdsmen in Hulunbeier City, Xilin Gol League, Chifeng City, and Tongliao City during 2010–2020 was greater than 1, indicating that the level of animal husbandry production in each region has been improved to some extent. The decomposition results show that the progress of comprehensive technical efficiency in the animal husbandry production process of herdsmen in each region has played a positive role in promoting the growth of total factor productivity in their animal husbandry. In terms of the phased situation, the total factor productivity of herdsmen’s animal husbandry in the sample overall, Tongliao City and Ulanqab City showed a feature of declining first and then rising, the total factor productivity of animal husbandry in Hulunbeier City showed a feature of rising and then slightly declining, Chifeng City showed a feature of rising first and then significantly declining, and Xilin Gol League showed a sustained rise during the two phases.

(2) From the convergence results, the overall sample showed a small σ convergence before and after the implementation of the second round of subsidy and award policy compared with the first round. Both Ulanqab City and Chifeng City showed significant σ convergence, while the total factor productivity σ convergence coefficients of animal husbandry in Hulunbeier City, Xilin Gol League, and Tongliao City showed a minor upward trend with dispersion overall. β convergence test results showed that the absolute β convergence coefficients of the overall sample and each region during the subsidy and award policy period showed absolute β convergence at the 1% significance level, indicating that the total factor productivity of herdsmen with slow development in animal husbandry production grows significantly faster than those with better development, and the total factor productivity levels of herdsmen will eventually reach the same steady-state level. In addition, the overall Inner Mongolia sample and each region exhibit conditional β convergence at the 1% significance level, further indicating that the spatial differences in the total factor productivity of animal husbandry of each herdsman showed a trend of gradual reduction as the policy continues to advance, and the overall sample will gradually accelerate convergence to their respective stable levels.

(3) The personal and family characteristics of herdsmen, transportation factors, economic factors, and subsidy and award policy factors are important driving factors of the total factor productivity in herdsmen’s animal husbandry. Among them, education level, the convenience of travel, degree of by-business, whether to implement the subsidy and award policy, and whether to be fined have significant positive effects on the total factor productivity in herdsmen’s animal husbandry, while family dependency ratio and degree of government regulation have significant negative effects on the total factor productivity in herdsmen’s animal husbandry.