Regional Assessment at the Province Level of Agricultural Science and Technology Development in China

Abstract

Science and technology innovation are crucial components underpinning agriculture. We constructed an evaluation framework including 4 pillars and 21 indicators, taking 31 provinces in China as examples to examine the level of agriculture science and technology development from a regional perspective. We found that there is an obvious gap between east and west nationwide, and that only about half of the provinces have reached the high and medium levels. It was worth noting that the innovation conditions in Shanghai and Beijing presented huge advantages, of vital importance to a first-class talent team, a complete innovation system, a stable and prosperous market, and active exchanges and cooperation. In addition, to maximize the transformation of agricultural science and technology achievements into real productivity, local government should also strengthen the construction of agricultural research and innovation platforms, technology transfer, and transformation of results. The findings advance understanding of the strengths and weaknesses of the evaluation subjects’ agricultural science and technology development from a regional perspective and are expected to provide some basis for the government and stakeholders to make relevant decisions.

1. Introduction

Agriculture is the basis of human survival and the economic pillar and foundational industry of many countries. However, its development is still facing urgent challenges such as food security, nutritional structure imbalance, agricultural non-point source pollution, and resource scarcity [1]. There is a widespread belief that the technologies and experiences of the past are no longer well adapted to the opportunities and challenges that agriculture will face in the future [2,3]. Many cutting-edge technologies—biotechnology and genomics, information technology, and microbiology—are transforming the traditional industry, with the scale and the pace of change taking place in global, regional, national, and local agricultural systems being unprecedented. A growing trend in digital and intelligent agricultural practices makes it possible to produce animal and plant products with higher efficiency and lower environmental impact [4,5]. Among the piles of scientific literatures and policy documents, competence in developing and employing science, technology, and innovation (STI), is increasingly acknowledged as an indispensable factor in stimulating agricultural economic growth and reshaping the agriculture system [6,7].

In the past decade or so, much research has been conducted on agricultural economic growth and its determinants [8,9,10]. Compared with other factors, STI has played a significant role in the take-off of the agricultural economy, but this does not reflect the actual degree and level of agricultural science and technology development (ASTD) in a country or region. As early as the 1980s, competitiveness research had attracted much attention from governments, enterprises, and academics due to its ability to identify the capacity and competitive advantage of a country, region, industry, or enterprise to compete in international markets. The World Economic Forum [11] has published the Global Competitiveness Report since 1979 to analyze the key factors of economic growth and explain why some countries are more successful in raising their income levels and employment opportunities. The International Institute of Management [12], in its World Competitiveness Yearbook, focused more on the competitiveness of countries and regions, and in particular on the performance of companies. This stream of research also extended to STI, such as WIPO’s Global Innovation Index [13], which measures an economy or country’s innovation capacity in terms of institutions, human capital and research, infrastructure, market sophistication, business sophistication, knowledge and technology outputs, and creative outputs. The RAND [14] developed a quick-turn and open-source methodology for determining national standing within the science and technology (S&T) fields. The National Science Foundation [15] published The State of U.S. Science and Engineering every two years, providing valuable information concerning the STEM workforce; STEM education at all levels; U.S. and international R&D performance; invention, knowledge transfer, and innovation; and U.S. competitiveness in high-technology industries. Remarkably, in the abovementioned studies on assessing the level of STI capacity, we should give attention to some common indicators: (1) research intensity ratio (agricultural research spending relative to its agricultural gross domestic product); (2) human resources (e.g., number of scientists and engineers per 1000 personal); (3) the gross enrollment rate in higher education (e.g., the proportion of S&E degree recipients); (4) the number of patents and academic publications and infrastructure conditions (e.g., ICT infrastructure); (5) institutions and policy (e.g., social intellectual property protection, stability of the economic environment, etc.).

There have been few evaluations of the level of ASTD. Most articles start from the agricultural innovation system and explore relevant factors that influence the changes in this system [16,17,18,19,20,21,22], such as the limitations of existing regulations and public policies, educated farmers, and stakeholders’ level of cooperation. Some researchers measured the level of ASTD at the macro level, with indicators based on the input–output ratio, showing that fewer inputs can produce more valuable scientific research results [23,24].

Since the 1980s, as a traditional agricultural country, S&T has contributed a lot to the development of Chinese agriculture. Although agricultural growth has slowed in the last decade, the impact of technology should not be underestimated. In 2021, the contribution rate of scientific and technological progress to the growth of the agriculture sector exceeded 60%. However, just as importantly, new challenges persist. ASTD still suffers from insufficient financial investment, low technology conversion rate, inefficient technology diffusion, and obsolete platform service [25]. There is an urgent need to assess the level of ASTD and analyze its strengths and weaknesses.

The significance of science and technology to agriculture development is self-evident; however, ideas on how to measure the level of ASTD have not yet been consistently addressed: (i) with regard to scale, the mesoscale has not received much attention, mainly the national level and farm level; (ii) the existing evaluation indicators appear to be not comprehensive enough, mainly input-type indicators and output-type indicators; (iii) it is difficult to reflect the level of ASTD directly and to make horizontal comparisons among different subjects.

In summary, there is a lack of research on measuring and evaluating the level of agricultural science and technology development, which highlights the need for relevant research in this area. Therefore, this paper is organized as follows: (i) an assessment framework for ASTD is designed and constructed, and a composite index is proposed; (ii) a case study is carried out with the specific situation of 31 provinces in China based on the ASTD framework, and (iii) the advantages and shortcomings of each province according to the evaluation results are identified and revealed.

2. Methodology

2.1. Framework Construction and Indicator Selection

The transformation of traditional agriculture into modern agriculture is a process of harnessing the effectiveness of agricultural factors comprising institutions, natural resources, science and technology, and capital. To maintain sustainable development, ASTD is the vital engine and usually involves multiple elements such as talent teams, capital investment, exchange and cooperation, and policy support [26,27].

The essential attribute of the ASTD is that applicable science and technology are integrated and penetrate traditional agriculture, achieving an increase in agricultural productivity and a leap in production efficiency. In other words, we regard the ASTD as a reflection of the degree of structural optimization and functional performance of the agricultural science and technology innovation system. More specifically, we believe that the assessment framework of ASTD should include four pillars: contribution, technical efficiency, innovation conditions, and knowledge and technology outputs. (see Figure 1).

Pillar 1: Contribution

This pillar indicates the major contributions of agricultural science and technology to social and economic development, with its new goals, including poverty reduction, adequate nutrition, functioning food value chains, and environmental sustainability [28]. The stable operation of the economy and the improvement of people’s living standards are the visual manifestations of the rapid development of agricultural science and technology [29]. The extension of industrial chains and value-added agricultural products is driven by new techniques and business models, making agriculture more competitive. Therefore, four indicators were selected for this pillar: grain yields per capita, agricultural GDP, agro-processing industry, and income inequality.

Pillar 2: Technical efficiency

Modern agriculture has gradually abandoned the reliance on traditional development factors such as resources and labor inputs and pursued more efficient and sustainable development, that is, the higher the output–input ratio of agricultural science and technology, the higher the efficiency of agricultural production activities. Therefore, eight indicators were chosen for this study: labor productivity, land productivity [30], animal-based protein, agricultural mechanization, fertilizer utilization, water utilization, disaster resistance, and green agriculture.

Pillar 3: Innovation conditions

This pillar is designed for this study as the foundation for carrying out STI activities and assessing the basic conditions and the possible activity level in each region [31]. It is necessary to emphasize that the scale and quality of human resources and research funding are significant sources of new technology driving growth in agricultural TFP [32]. In addition, high-tech enterprises, characterized by a high proportion of innovation inputs and high autonomy of core technologies, are the agents of supporting changes in the field of agricultural production with science and technology [33]. National Agricultural Science and Technology Park is a kind of approach that has a leading role in establishing new agricultural research platforms and promoting exchanges and cooperation [34]. Therefore, four indicators were selected under this index: professionalism of the workforce, R&D expenditure, China National Agricultural Science and Technology Park, and high-tech enterprises.

Pillar 4: Knowledge and technology outputs

The pillar is aimed at measuring the extent to which a country or region can absorb and use scientific and technological knowledge, as well as the application level [35,36], which is also a major embodiment of the performance of STI and the active degree of the development of science and technology innovation activities. The pillar is measured by highly cited papers, invention patents, industry standards, crop variety, and the China State Science and Technology Award.

After such preliminary selection, all indicators need to be tested to ensure that they were all available and reliable, including whether the coefficient of variation was sufficiently high. First, we needed to carry out a correlation analysis to check the set of indicators that have some correlation, and then eliminate the indicators, so as to improve the scientificity and rationality of the research results. Generally, according to the research needs, a critical point M (0 < M < 1) is identified to determine the critical value of whether the correlation is close. This paper assumes that the critical point M = 0.8. If the correlation coefficient R_ij > M, it is considered that there is much overlapping information between indicators, which will affect the evaluation results and should be deleted. In addition, it is also necessary to analyze the discrimination of indicators. If the indicators show similar or relatively consistent scores, this indicates that the discrimination of the indicators is weak and should be eliminated; on the contrary, if the indicator has high discrimination, it should be retained. In this paper, 0.15 is taken as the critical value, and the indicators with a coefficient of variation lower than 0.15 are eliminated. After calculation, all indicators can be retained.

In summary, this paper selects a total of 21 indicators from four aspects to comprehensively examine the development level of agricultural science and technology and constructs an evaluation index system (

Table 1. The indicator system of ASTD.

Target Layer	Criterion Layer	Index Layer (Unit)
A. Agriculture S&T Development (ASTD)	B1. Contribution	C1 Grain yields per capita (kg per capita)
		C2 Agricultural GDP (%)
		C3 Agro-processing industry (ratio value)
		C4 Income inequality (ratio value)
	B2. Technical efficiency	C5 Labor productivity (CNY 100 million per million people)
		C6 Land productivity (kg per hectare)
		C7 Animal-based protein (tons per 10,000 pigs)
		C8 Agricultural mechanization (MKw per 1000 hectare)
		C9 Fertilizer utilization (tons per hectare)
		C10 Water utilization (%)
		C11 Disaster resistance (%)
		C12 Green agriculture (%)
	B3. Innovation conditions	C13 Professionalism of the workforce (one per 10,000 people)
		C14 R&D expenditure (CNY 100 million per person)
		C15 China National Agricultural Science and Technology Park (quantity value)
		C16 High-tech enterprises (%)
	B4. Knowledge and technology outputs	C17 Highly cited papers (articles per 10,000 persons)
		C18 Invention patents (number per 1000 persons)
		C19 Industry standards (number per 100 persons)
		C20 Crop variety (number per 100 persons)
		C21 China State Science and Technology Award (number per 100 persons)

).

2.2. Weighing Indicators and Data Source

Composite indices have been extensively used in studying sustainability in agricultural development since they are valuable tools for policymaking and alternatives that integrate and complement different indicators separately. In the present study, a new composite index has been developed to measure ASTD. Each step is shown in the following sections.

2.2.1. Normalization of Indicators

To ensure the comparability of data indicators, it is essential to normalize them before aggregation since they are measured based on various units. They need to be expressed in homogeneous units so that we can perform arithmetical operations on them. Thus, the first step is data standardization, as shown in Formulas (1) and (2). The sets of indicators can be divided into two types: stimulants and de-stimulants. For example, the professionalism of the workforce is regarded as a stimulant or a positive indicator, which means that the better the performance of the indicator, the higher the level of agricultural science and technology is likely to be. On the contrary, if the income inequality between urban and rural residents in a region is large and the farmers’ lives are poor, it may mean that the local agricultural science and technology development level is relatively low and the farmers’ income is low.

$$\mathrm{for}\mathrm{stimulants}:x_{ij}^{\prime }=\frac{x_{ij}-\mathit{min}{x}_{ij}}{\mathit{max}{x}_{ij}-\mathit{min}{x}_{ij}}$$

(1)

$$\mathrm{for}\text{ de-stimulants}:x_{ij}^{\prime }=\frac{\mathit{max}{x}_{ij}-x_{ij}}{\mathit{max}{x}_{ij}-\mathit{min}{x}_{ij}}$$

(2)

where $x_{ij}$ and $x_{ij}^{\prime }$ represent respectively the raw data and standardized data of the j-indicator in the i-pillar.

2.2.2. Weighting Indicators

As one of the most widely used methods for determining index weight, the analytic hierarchy process (AHP) can organize a complex and multi-attribute problem into a hierarchical structure (

Table 2. Weights of the indicator system.

Objective Layer	Criterion Layer	Weight	Index Layer	Weight
A. Agriculture S&T Development (ASTD)	B1. Contribution	0.467	C1 Grain yields per capita	0.527
			C2 Agricultural GDP	0.178
			C3 Agro-processing industry	0.194
			C4 Income inequality	0.101
	B2. Technical efficiency	0.263	C5 Labor productivity	0.237
			C6 Land productivity	0.197
			C7 Animal-based protein	0.119
			C8 Agricultural mechanization	0.157
			C9 Fertilizer utilization	0.072
			C10 Water utilization	0.080
			C11 Disaster resistance	0.071
			C12 Green agriculture	0.073
	B3. Innovation conditions	0.169	C13 Professionalism of the workforce	0.453
			C14 R&D expenditure	0.313
			C15 China National Agricultural Science and Technology Park	0.090
			C16 High-tech enterprises	0.144
	B4. Knowledge and technology outputs	0.102	C17 Highly cited papers	0.146
			C18 Invention patents	0.213
			C19 Industry standards	0.170
			C20 Crop variety	0.373
			C21 China State Science and Technology Award	0.098

). The 23 respondents of the questionnaire are mainly related personnel of agricultural research institutes and agriculture-related institutions of higher learning. The specific steps of AHP are as follows [37,38]:

Step 1: Construct a progressive hierarchical structure analysis model.

When using the AHP method to determine the weight, its indicators are usually built into a progressive hierarchical model according to their relevance and subordinate attributes, which is generally divided into three to four levels. Therefore, this paper constructs a three-level hierarchical model including the target layer, the criteria layer, and the index layer. The target layer is usually an indicator, and its indicator meaning is the ultimate purpose of the article evaluation, representing the overall evaluation level. The criteria layer contains multiple indicator classifications in the system structure, while the indicator layer is the expansion and further decomposition of the criteria layer.

Step 2: Construct a judgment matrix of all layers.

The experts advise taking the indicators of a certain level as the criterion and comparing the relative importance of the data of the next level in pairs, so as to build the corresponding judgment matrix A. Then, according to the scoring data of experts, we take the 1–9 scale method of Professor T.L. Satty to build the judgment matrix A (Saaty’s hierarchy matrix).

Step 3: Conduct the consistency test.

To obtain the maximum eigenvalue of the judgment matrix A, it is necessary to use Formulas (3) and (4) to obtain the CI (the consistency index) and CR (the consistency ratio) of the judgment matrix to test its consistency. The higher the consistency of the judgment matrix, the smaller the CI value. CI is generally considered to be less than 10%. It is noteworthy that each judgment matrix here has passed the consistency test, indicating that the judgment matrix has overall consistency.

$$CI=\frac{\left(\lambda max\right)-n}{n-1}$$

(3)

$$CR=\frac{CI}{RI}$$

(4)

where n corresponds to the dimension of the pairwise comparison matrix A, $\lambda max$ is the largest eigenvector.

Finally, the scores of all the indicators were summed to calculate the ASTD index and development index of the four pillars. The higher the score, the more advanced a province’s ASTD is. The special calculation models are shown in Formulas (5) and (6).

$$A={{\displaystyle \sum }}^{}{B}_i\times {\omega }_i$$

(5)

$$B_i={{\displaystyle \sum }}^{}{x}_{ij}^{\prime }\times {\omega }_{ij}$$

(6)

where $B_i$ and ${\omega }_i$ represent the development index of four pillars and the weight of the four pillars, respectively, ${\omega }_{ij}$ is the weight of these indicators, and A is the ASTD index.

Then, we conduct the cluster analysis based on Ward’s method by R software (version 4.1.0), and we divide the ASTD index and development index of the four pillars of each province into three levels: high level, medium level, and low level. At the beginning of clustering, the distance between n samples needs to be calculated. We chose the Euclidean distance to calculate to measure the distance between objects. The Euclidean distance between the sample i and the sample j is calculated as shown in Formula (7).

$$d_{ij}=\sqrt{{{\displaystyle \sum }}_{k=1}^p{\left(x_{ik}-x_{jk}\right)}^2}$$

(7)

The data sources of this paper are China Statistical Yearbook, China Rural Statistical Yearbook, China Statistical Yearbook On Science And Technology, and the local official website. It is noted that some indicators without direct data are obtained by sorting out and calculating the basic data. The evaluation year was 2020. Appendix A shows calculation progress and data sources of all indicators.

3. Results

Table 2 demonstrated the weight of indicators in each layer and the development indexes of the four pillars, based on which the ASTD index for each province could be derived (

Table 3. Ranking of ASTD index and sub-pillar score in each region.

Region/Province		Contribution	Rank	Technical Efficiency	Rank	Innovation Conditions	Rank	Knowledge and Technology Outputs	Rank	ASTD Index	Rank
North China	Beijing	40.76	9	49.64	6	90.09	1	49.95	5	51.48	2
	Tianjin	44.17	6	47.81	9	26.88	8	42.77	8	42.37	7
	Hebei	37.73	12	48.21	7	20.39	15	34.99	12	37.58	14
	Shanxi	32.96	19	21.60	30	1.59	31	19.80	23	23.93	26
	Inner Mongolia	54.75	3	31.97	24	14.64	23	23.61	21	39.64	11
Regional average score		42.07	2	39.85	3	30.72	2	34.22	4	39.00	4
Northeast China	Liaoning	33.18	18	40.62	11	11.16	28	27.52	14	31.24	19
	Jilin	60.58	2	37.68	18	11.40	27	31.30	13	44.24	4
	Heilongjiang	67.12	1	38.20	14	14.87	22	50.81	4	49.99	3
Regional average score		53.63	1	38.83	4	12.48	7	36.55	3	41.82	1
East China	Shanghai	26.52	25	56.09	4	72.62	2	11.91	29	39.89	10
	Jiangsu	44.02	7	65.52	1	34.17	5	80.94	1	51.76	1
	Zhejiang	38.30	11	63.29	2	42.28	3	16.57	26	43.37	6
	Anhui	43.92	8	36.53	19	18.52	19	62.60	2	39.93	9
	Fujian	34.98	14	57.89	3	24.32	9	27.45	15	38.65	13
	Jiangxi	38.48	10	37.92	15	12.40	26	20.07	22	32.58	17
	Shandong	46.01	5	51.07	5	23.27	10	42.87	7	43.58	5
Regional average score		38.89	4	52.62	1	32.51	1	37.49	2	41.39	2
Central China	Henan	49.23	4	42.33	10	16.40	21	44.95	6	42.01	8
	Hubei	37.70	13	37.91	16	30.33	7	37.01	10	36.57	15
	Hunan	34.74	16	48.08	8	23.13	11	59.89	3	38.92	12
Regional average score		40.55	3	42.78	2	23.28	4	47.28	1	39.17	3
South China	Guangdong	31.54	21	40.10	12	37.32	4	40.78	9	35.57	16
	Guangxi	23.53	28	34.30	22	19.31	16	26.38	16	26.00	23
	Hainan	18.98	31	39.05	13	22.21	13	3.96	31	23.29	29
Regional average score		24.68	7	37.82	5	26.28	3	23.70	6	28.29	5
Southwest China	Chongqing	30.64	22	36.31	20	32.75	6	14.32	27	30.87	20
	Sichuan	34.96	15	34.33	21	18.95	17	24.45	19	31.34	18
	Guizhou	21.00	29	25.71	26	12.83	24	13.85	28	20.30	31
	Yunnan	27.99	24	20.38	31	12.47	25	25.40	18	23.38	28
	Tibet	27.09	30	30.13	23	19.13	18	16.68	30	25.67	30
Regional average score		20.86	6	33.92	6	18.62	5	15.36	7	22.45	7
Northwest China	Shaanxi	29.57	23	25.53	27	8.65	30	24.10	20	24.80	25
	Gansu	26.47	26	21.79	29	18.46	20	36.73	11	25.01	24
	Qinghai	25.51	27	23.39	28	21.22	14	17.49	25	23.52	27
	Ningxia	34.48	17	30.12	25	22.98	12	18.35	24	30.02	21
	Xinjiang	32.24	20	37.90	17	10.34	29	26.04	17	29.80	22
Regional average score		29.65	5	27.75	7	16.33	6	24.54	5	26.63	6
National average score		36.22	—	39.20	—	24.02	—	31.04	—	34.65	—

).

3.1. Analysis of the ASTD Index

The top ten provinces ranked by the ASTD index for 2020 were Jiangsu, Beijing, Heilongjiang, Jilin, Shandong, Zhejiang, Tianjin, Henan, Anhui, and Shanghai, with scores ranging from 39.89 to 51.76, mainly distributed in Northeast and East China. Xinjiang, Guangxi, Gansu, Shaanxi, Shanxi, Qinghai, Yunnan, Hainan, Tibet, and Guizhou ranked in the bottom 10, with scores ranging from 20.3 to 29.8, mainly distributed in Southwest and Northwest China. From the regional scale, the ASTD index showed a decreasing trend from east to west, and the ranking order is Northeast China, East China, Central China, North China, South China, Northwest China, and Southwest China.

Except for Shanxi province, the ASTD indices of provinces in North China, particularly Beijing’s score, were higher than the average level, on both the national and regional scales. Among the three northeast provinces, only Liaoning province performed below average. Three provinces—Jiangsu, Zhejiang, and Shandong—played an outstanding leading role, with Jiangxi province performing below average. In South China, Southwest China, and Northwest China, except Guangdong, the rest of the provinces performed below the national average level. Chongqing and Sichuan were slightly outstanding in Southwest China, as Ningxia and Xinjiang were in Northwest China.

3.2. Analysis of Pillar 1: Contribution

According to Table 2, grain yields per capita was the major factor affecting this pillar, which reflected the level of food security in the region, that is, whether the local population can be provided with sufficient food with as little support as possible from other regions. So, the provinces that scored high on this pillar also tended to have better food production capacity.

It is worth noting that the common characteristics of the economically developed regions such as Shanghai, Beijing, and Tianjin were low agricultural GDP and weak grain production capacity, but the agro-processing industry was relatively flourishing. The income gap between urban and rural residents in Shanghai and Tianjin was not very large, but the agro-processing in Tianjin was relatively well-developed. Therefore, Tianjin performed better than Shanghai.

At the regional scale, the performance on this pillar of each region from best to worst was as follows: Northeast China, North China, Central China, East China, Northwest China, Southwest China, and South China. Northeast and North China are home to many large grain-producing provinces, which have a certain impact on the score.

3.3. Analysis of Pillar 2: Technical Efficiency

In this pillar, labor productivity, land productivity, mechanization, and animal-based protein were the indicators with high weights. Although Shanghai ranked first at the national level in land productivity, it performed relatively poorly in labor productivity, agricultural mechanization, and disaster resistance, so it slightly lagged behind Jiangsu, Zhejiang, and Fujian. Beijing stood out in terms of animal-based protein, water utilization, and green agriculture, but the level of labor productivity and disaster resistance needed to be improved, so its overall performance was slightly inferior to that of Shandong province.

At the regional scale, the development of technical efficiency in South China, Southwest China, and Northwest China was lower than the national average, and Guangzhou, Chongqing, and Xinjiang were the provinces that performed relatively excellently in the above three regions, respectively.

3.4. Analysis of Pillar 3: Innovation Conditions

Two indicators, professionalism of the workforce and R&D expenditure, with a total weight of 76.6%, were the key factors affecting this pillar.

From the raw data, the professionalism of the workforce of five provinces, Beijing, Shanghai, Zhejiang, Tianjin, and Ningxia, was above the national average. Beijing had about 385 agricultural S&T personnel per 1000 people in the primary industry, while Shanghai had 347 people and ranked second, followed by Zhejiang province with 75, ranking third. Although Shanxi province ranked 11th nationwide in terms of the professionalism of the workforce, it performed poorly on the other three indicators in this pillar. Chongqing performed poorly in the high-tech enterprises while scoring relatively high on other indicators, so its innovation conditions ranked sixth in the country.

The scores on this pillar of the seven regions were ranked from best to worst as follows: East China, North China, South China, Central China, Southwest China, Northwest China, and Northeast China. The scores of the three northeastern provinces were even less than half of the national average. From the raw data, Liaoning province needed to invest more in research funding and set up a professional research team. By contrast, Jilin province and Heilongjiang province should focus on investment in research funds and fostering high-tech agricultural enterprises.

3.5. Analysis of Pillar 4: Knowledge and Technology Outputs

Crop variety and invention patents, accounting for more than 50% of the weight, were important influencing factors that affect the scores of this pillar for each assessment object.

From the raw data, Jiangsu and Guangdong performed well in highly cited papers and invention patents, while scoring lower in agricultural science and technology awards. Although the number of highly cited papers and industry standards was not impressive, Henan and Shandong performed better on three indicators: crop variety, invention patents, and State Science and Technology Awards. Shanghai’s performance on this pillar was poor with its rank in the bottom third for both papers and varieties, in addition to its low publication of standards and few awards in the agriculture field.

In terms of regional scores, the order of performance on this pillar was, from best to worst: Central China, East China, Northeast China, North China, Northwest China, South China, and Southwest China. The southwest region’s score was only half of the national average, among which Yunnan and Sichuan provinces performed better, while Tibet was particularly lagging. Hainan Province was at the bottom of the ranking and given its performance on technical efficiency and enabling conditions, attention should be paid to adjusting the input–output ratio to improve the efficiency of innovation.

3.6. Analysis of the Four Pillars

The classification result is shown in Figure 2.

In terms of contribution, North China, Northeast China, Central China, and East China performed better. In the other three regions, six provinces—Sichuan, Ningxia, Xinjiang, Chongqing, Guangdong, and Shaanxi—performed relatively well, while the rest of the provinces performed poorly.

In terms of technical efficiency, except for Shanxi province, North China attained a high level. Anhui and Jiangxi were the only two provinces in East China that performed at a medium level. Northeast and South China both performed at a medium level, while Hunan province in Central China stood out with a high level of performance. The western region (Southwest and Northwest) performed poorly, with 50% of provinces at medium and low levels.

In terms of innovation conditions, 64.5% of the country’s provinces were above the medium level, among which only Beijing and Shanghai were at a high level, while the northeast and western regions performed relatively poorly.

In terms of knowledge and technology outputs, Beijing and Tianjin in North China, Anhui, Shandong, and Jiangsu provinces in East China, Henan in Central China, Heilongjiang in Northeast China, and Guangdong in South China all performed well and reached a high level.

In general, the level of ASTD in China was widely disparate and unevenly developed from east to west. Provinces in the western region (Northwest and Southwest) performed at or below the medium level in all pillars. Only Guangdong province performed better in Southern China. The three provinces in Northeast China performed well in the contribution pillar, among which Heilongjiang province had a relatively excellent overall performance. East China, Central China, and North China performed relatively well, but Jiangxi, Hubei, and Shanxi provinces were mediocre in their respective regions.

4. Discussion

How to shape a competitive and sustainable agricultural system has long been a high priority for governments, and the promotion of agricultural science and STI is one of the top priorities in addressing the many challenges of agricultural development [39].

Food security concerns the survival and well-being of human beings and has been a major preoccupation of the international community. As the two traditional major grain-producing areas, Northeast China and North China, an important reason for their high ranking based on the ASTD index is the ability to keep grain yields stable and productive. The region mainly takes good advantage of crop varieties, advanced technology, and intensive management instead of inputs of labor resources and agricultural items to improve food production supply capacity and market competitiveness.

The development of science and technology is closely related to the professionalism of the workforce and research funding, which are the two supporting forces for the enabling conditions. In this study, The human resources of agricultural science and technology in Shanghai and Beijing present huge advantages over other provinces. It requires us to be aware of the significance of building a strong and professional labor force by training and educating domestic talent and by recruiting and retaining foreign talent. Additionally, the level of agricultural R&D investment in Shanghai, Guangdong, Zhejiang, Jiangsu, and Beijing is much higher than that of other provinces, which is closely related to the thriving economy and market, intensive active exchanges and cooperation activities, and complete agricultural innovation system.

The improvement of technical efficiency is the most direct embodiment of the level of agricultural science and technology, the essence of which is to obtain higher outputs and benefits with fewer inputs. The index of agricultural technical efficiency development in Jiangsu, Zhejiang, Fujian, Shanghai, and Shandong was relatively high but still fell far short of the highest international standards. Some agricultural powerhouses firmly occupy a leading position worldwide by relying on advanced scientific and technological forces and key development fields, such as the United States relying on molecular breeding technology [40], Germany relying on agricultural machinery manufacturing technology, and Israel relying on water-saving irrigation technology [41]. Therefore, local governments can strengthen the regional division of production on the premise of resource and environment carrying capacity and based on comparative advantages, actively develop a sustainable development model suitable for the local area, and realize the transformation and development of green, efficient, multi-functional, high-value, and smart agriculture.

In terms of agricultural science and technology achievements, there is no lack of high-level achievements in China, but there is still a big gap in quality and quantity compared with the world’s most advanced level, and weak competitiveness globally. This is to some extent related to the degree of optimization of the agricultural science and technology innovation system. We can learn from the experience of developed countries. As one of the countries with the highest degree of agricultural modernization, the U.S. agricultural innovation system is mainly composed of scientific research institutions, colleges, and universities, agriculture-related enterprises, farmers, and intermediary organizations. The organizations are closely integrated and have a clear division of labor, which not only ensures that there is no disconnection between scientific research, education, extension, and actual production but also helps popularize the new agricultural technology in agricultural production practice. By contrast, the Netherlands has formed a nationwide agricultural innovation system and network with farmers as the core and farms as the carrier, and has realized the integration of advanced technology and equipment with the goal of industrial development and technology promotion and application [42,43].

In conclusion, a few developed countries have gained an advantage in the fierce international competition in agriculture by relying on ASTI and new models of agriculture [44]. China’s agriculture, with some exceptions, is mainly composed of informal economies, non-standardized production systems, and uneven, though not identical, technological levels, dispersed throughout the territory, which is common in many developing countries. In any case, governments must recognize the need to overcome the constraints of ASTD to produce quality food for the domestic market and for export, and to exert a positive impact on the territories in which they operate.

In terms of relevant research with reference to agriculture, academic studies concerning the level of ASTD have not been carried out. This justifies analyses carried out on the indicated subject, and this paper is an attempt at filling this gap. Some Chinese scholars [45,46,47] have put forward the concept of agricultural scientific and technological innovation capability to reflect the development basis, achievements, and efficiency of ASTI of the evaluation object. In other words, the design process of an indicator system involves three aspects: input, output, and scientific and technological conditions. The differences are mainly reflected in the evaluation object (regional scale or a certain province or a certain National Agricultural Science and Technology Park) and indicator system design. Due to different understandings of agricultural scientific and technological innovation capability, scholars choose different variables and calculation methods to reflect the connotation of indicators, which will affect the evaluation results and analysis. By comparing the results of studies with a regional evaluation scale, the conclusions drawn in this paper are reliable. The level of regional ASTD in China is still uneven, especially in the eastern and western regions. However, the ranking and scores of provinces are slightly different. A point of agreement is that Beijing, Tianjin, Shanghai, Jiangsu, Zhejiang, and Shandong provinces have performed relatively well, and the provinces in the southwest and northwest have performed relatively poorly. Notably, there are some exceptions, such as Heilongjiang Province and Jilin Province ranking very high in this study, which may be because the grain yields per capita is a very important factor in the contribution criterion. This paper believes that one of the important tasks of agricultural science and technology is to ensure food security and increase agricultural added value. However, other papers may not give this indicator a high weight value.

5. Conclusions

This study proposed a framework for evaluating ASTD and conducted an empirical analysis of 31 provinces in China. On the regional scale, the ASTD index showed a remarkable contrast between east and west, with the ranking from best to worst as follows: Northeast China, East China, Central China, North China, South China, Northwest China, and Southwest China. Regarding the four pillars of the 31 provinces, those in the central and eastern regions were concentrated at the high and medium level (Northeast China, East China, Central China, and North China), and the western region had an overall weaker development level. Provinces with a high ASTD index accounted for 9.7% of all provinces, with 41.9% provinces at the medium level, and 48.4% at the low level. These findings are expected to provide some basis for the government and stakeholders to make relevant decisions. The local government should reasonably optimize the interregional allocation of agricultural elements to maximize the transformation of outputs into productivity, especially focusing on the significance of personnel training and research funding.

However, the present study has limitations in the indicator selection of each pillar of ASTD, with the purpose of facilitating the relative ranking and comparison of the examined systems. Second, the AHP method was used to assign weights to the indicators. Future researchers can continue to use a wider selection of variables and other methods allowing a comprehensive assessment of ASTD. Third, in view of the availability of data, the time scale of this study is limited to 2020. In fact, if we can make a comparative study of the static and dynamic conditions of ASTD at different stages, we can better reveal the dynamic change process of ASTD, which is a future research direction.