Food processing industry changes across China regions: The case of flour, rice, oil, and other cereal derivative food

Abstract Faced with the pressure of slowing industrial growth and industrial transformation requirements, it is crucial to analyze the changes and the corresponding driving factors of the food processing industry in China. An analysis using traditional and spatial shift‐share models was conducted to decompose the changes in the food processing industry in each region of China from 2009 to 2019 into five effects: national growth effect (NG), industrial mix effect (IM), competitive effect (CE), neighbor‐nation competitive effect (NNC), and region‐neighbor competitive effect (RNC). Among the five effects from 2009 to 2019, the NG contributed the most to the growth in most regions, indicating that the development of the food processing industry in China was greatly influenced by the industrial base and that China's food processing industry has entered a “growth bottleneck period.” During the period 2009–2014 to period 2014–2019, compared to the IM and CE, the influence of spatial spillover effects was stronger and significantly enhanced. Moreover, the IM, CE, NNC, and RNC in most southern regions were stronger than those in most northern regions. Therefore, China's food processing industry needs and is transforming into high‐quality development. It is necessary to innovate the mode of development of food processing industry and strengthen interregional exchanges and cooperation.

processing industry, as it closely associates with the livelihood of 1.4 billion people (Xie & Li, 2022). On the other hand, China has a large population yet a relatively small proportion of arable land.
Food production and consumption have been in a tight balance for a long time. Based on China's agricultural resource endowment, dietary structure, and dietary habits, China's food security strategy is to ensure "the absolute security of staple foods and the basic selfsufficiency of grains" (Niu et al., 2021;State Council Information Office of the People's Republic of China, 2019). Therefore, grain self-sufficiency, grain cultivation, and grain processing are highly emphasized in China (Cao et al., 2021;Chen et al., 2011;Hang, 2021;Li, Sun, et al., 2021;Niu et al., 2022). Thus, this study focused on China's food processing industry mainly including flour, rice, oil, feed, and other cereal derivative food using data from "Statistics of Grain and Oil Processing Industry".
In recent years, China's food processing industry has gained rapid development. With the continuous improvement of comprehensive grain production capacity, China's grain production has been abundant for years. The annual output has stabilized above 650 million tons (Li & Lv, 2021), laying a good foundation for further development of China's food processing industry.
According to official data from the Food and Strategic Reserves Administration of China, the annual output value of China's food processing industry has remained above 3 trillion yuan since 2018.
In 2019, the total industrial output value of China's food processing industry was 3.15 trillion yuan, which is up 170.39% from 2009. in the past decade, with the increase in residents' income, Chinese consumers' demand for food is changing from "eat enough" to "eat healthy" (Li & Lv, 2021). In order to better meet the needs of consumers and to ensure food security, the Chinese government has guided China's food processing industry to produce food with better quality and to achieve a more sustainable development by improving resource efficiency (Cao et al., 2019;Gao et al., 2021; General Office of the State Council of China, 2017;Lu et al., 2022;Xie & Li, 2022;. It is in line with the transformation trend of the global food processing industry in recent years (Aguilar et al., 2019;FAO, 2019;. Therefore, faced with the pressure of slowing industrial growth and industrial transformation requirements, it is crucial for Chinese policy-makers to analyze the changes and the corresponding driving factors of food processing industry. Shift-share analysis was used to study the changes and the corresponding driving factors of food processing industry. Shift-share analysis is a common structural competitive advantage analysis method, which examines influences of industrial structure, regional competitive differences, and spatial spillover effect on industrial growth (or economic growth) through the decomposition of industrial growth (or economic growth). This method helps to analyze the reasons for the differences in industrial growth in different regions, which broadened the research content of regional difference analysis (Dogru et al., 2020;Polyzos, 2019). Shift-share analysis has been improved and a variety of models have been developed (Costantino et al., 2020;Lahr & Ferreira, 2020;Montanía et al., 2020). Among them, the traditional and the spatial shift-share model are the two most popular models. The traditional shift-share model was first proposed by Jones (1940) and Creamer (1942), then developed by Dunn (1960), and refined by Esteban (2000).
The traditional shift-share model is mainly used to analyze the contribution of industrial structure and regional competition differences to industrial growth (Li, Xing, et al., 2021). Spatial shiftshare model is mainly used to analyze the impact of spatial spillover effects (Espa et al., 2014). Nazara and Hewings (2004) considered the spatial interaction between regions, and first introduced the spatial shift-share analysis. Based on the ideas of Nazara and Hewings (2004), Zaccomer (2006) constructed a complete spatial shift-share model. Many scholars analyzed the changes in different industries in different regions using traditional or spatial shift-share models (Capello & Cerisola, 2022;Dogru & Sirakaya-Turk, 2017;Espa et al., 2014;Li, Xing, et al., 2021;Lv et al., 2021;Mayo & López, 2008). Studies have been carried out in many fields such as regional productivity (O'Leary & Webber, 2015), electricity consumption (Grossi & Mussini, 2018;Lin et al., 2019), tourism (Krabokoukis & Polyzos, 2020), and global rice export (Lakkakula et al., 2015). To the best of our knowledge, this might be the first study applying traditional and spatial shift-share analysis to decompose China's food processing industry changes.
Until now, a number of studies have focused on the performance and development of food processing industry in China.
Productivity and performance in the food processing industry is one of the research hotspots (Hockmann et al., 2018;Kapelko et al., 2015Kapelko et al., , 2016Kapelko et al., , 2017aKapelko et al., , 2017bTriguero et al., 2013). For example, Kapelko (2019) analyzed the productivity changes in food processing industry in EU countries, and Cardamone (2020)    compared the performance of food processing and food manufacturing in China, and the results showed that the performance of food manufacturing enterprises was better than that of food processing enterprises. Sun et al. (2021) studied the spatial pattern of the Chinese green food industry and found that the structure and types of green food enterprises are relatively simple and the regional development is not unified. Furthermore,  explored the green development mode of the Chinese food industry and the total factor productivity of food processing industry based on the Industrial Internet of Things. Cao et al. (2019) investigated the sustainable development of Chinese food processing enterprises.
These studies have provided insights into the understanding of the development and changes in the food processing industry in China. However, China is vast, and the food processing industry includes many subsectors. There is a lack of systematic analysis of the growth differences and the corresponding driving factors of food processing industry in different regions of China.
From the above discussion, we can infer that on the one hand, The rest of this research paper is arranged as follows: Section 2 is about models and data sources; the analysis and discussion of decomposition results are stated in Section 3; and research conclusions and implications are presented in Section 4.

| ME THODOLOGY AND DATA
The methodology for this study is conducting shift-share analysis, and the statistical data of China's food processing industry from 2009 to 2019 is used. The output value growth of the food processing industry in 31 provincial administrative regions in China has been decomposed into five effects. ArcGIS software was applied to analyze and map the spatial and temporal patterns of the actual growth and the five effects.

| Data sources and indicators
The data used for the shift-share analysis are derived from the "Statistics of Grain and Oil Processing Industry" of the National Food and Strategic Reserves Administration of China. Since the latest data are not available, this study takes 2009 as the base period and 2019 as the end period to analyze the output value growth of the food processing industry in different regions. The data span 11 years, including 31 provincial-level administrative regions (excluding Hong Kong, Macao, and Taiwan), which is representative to a certain extent. Some Chinese scholars have used the data from 2008 to 2014 of "Statistics of Grain and Oil Processing Industry" to study the total factor productivity and industrial agglomeration of food processing industry, and reached reliable conclusions (Chen & Zhong, 2017;. It can be seen that the data are applicable for industry analysis. R represents the growth rate of the output value of the food processing industry in China. R i is the national output growth rate of the i industry. R j represents the growth rate of output value of food processing industry in j region. r v ij indicates the growth rate of the i food processing industry in the adjacent region of region j. G t0 j represents the base production value of the food processing industry in region j. G t1 j stands for the final output value of the food processing industry in region j, and G t0 ij represents the base production value of industry i in region j. The first part on the right of the equation is the national growth effect (NG), which represents the amount increased by the food processing industry in region j according to the growth rate of the national food processing industry. The second part is the industrial mix effect (IM), which represents the sum of the amount increased by the subdivisions of the food processing industry in region j according to the difference between the growth rate of the output value of subdivisions and that of the whole industry in China. Its value is positive, indicating that most food processing subdivisions in this region have industrial growth advantages, and that this region has an industrial structure advantage. The third part is the competitive effect (CE), which represents the sum of the amount increased by the subdivisions of food processing industry in region j according to the difference between the growth rate of actual output value of the subdivisions in region j and that of the subdivisions in China. Its value is positive, indicating that the growth rates of most food processing subdivisions in region j are higher than that of other regions in the country and that this region has a competitive advantage.

| Spatial shift-share model
Juxtapose with the traditional shift-share model, the spatial shiftshare model considers the interaction between regions. Studies have shown that spatial spillover effects have an important impact on industrial development.
In this study, a 30 × 30 weight matrix W is constructed to represent the interaction among 30 provincial administrative regions. Since Hainan does not border other regions, a 30 × 30 weight matrix is constructed. w jk is the spatial weight, representing the intensity of the interaction between regions j and k. Different definition standards of spatial weight will influence the results of empirical analysis to some extent. Therefore, the most classical (0-1) space-weight matrix construction method is selected. Geographically adjacent counts as 1, and not adjacent counts as 0. And the matrix W is normalized by row.
The calculation of the spatial shift-share model is shown in Equation (2). The calculation of r v ij is shown in Equation (3), and other symbols are consistent with the symbolic meaning of Equation (1). In Equation (3), v is the number of adjacent regions of region j, and w jk is the row-normalized spatial weight defined earlier. G t0 ik represents the base output value of the industry i in the adjacent regions of region j.

G t1
ik stands for the final output value of the industry i in the adjacent regions of region j.
In Equation (2), the first part on the right is the national growth effect (NG), which is consistent with the national growth effect in Equation (1). The second part on the right side of the formula is the neighbor-nation competitive effect (NNC), which represents the sum of the amount increased by the food processing subdivisions in region j according to the difference between the growth rate of the output value of the subdivisions in adjacent regions of region j and that of the whole industry in China. Its value is greater than 0, indicating that most of the food processing subdivision industries in adjacent regions of region j have industrial growth advantages, and that the region j has a spatial structure advantage. The third part is the region-neighbor competitive effect (RNC), which indicates the sum of the amount increased by the food processing subdivisions in region j according to the difference between the growth rate of actual output value of the subdivisions in region j and that of the subdivisions in the neighboring regions of region j. Its value is greater than 0, indicating that the growth rates of most food processing subdivisions in region j are higher than that of the industry in neighboring regions, and that the region j has a spatial competitiveness advantage.   advantage of industrial structure. The larger the value, the more reasonable is the regional structure layout.

| RE SULTS AND D ISCUSS
According to Figure

| Spatiotemporal change in region-neighbor competitive effect
A positive region-neighbor competitive effect indicates that the growth of most food industries in this region is better than that in its adjacent regions, with spatial competitiveness advantages and spatial spillover effects.
According to Figure  in terms of raw grain cultivation. The raw grain output in North China was higher than that in the South, and the grain logistics obviously showed a trend of "grains being transported from north to south."

| Comparison between different effects
Firstly, among the five effects of output value growth decomposition from 2009 to 2019, the national growth effect contributed the most, indicating that China's food processing industry has entered a "growth bottleneck period." Table 2 shows the sum of the absolute value of regions with the five effects. According to Table 2 Table 2, the sum of the regional absolute value of industrial mix effect and competitive effect did not change much in these two periods. The regional absolute summation of neighbor-nation competitive effect increased from 391.301 billion yuan to 1875.561 billion yuan, and the regional absolute summation of region-neighbor competitive effect increased from 547.905 billion yuan to 2176.62 billion yuan, thus, indicating a significant increase in spatial effects. Figure 5 shows the percentage of the effects of Spatial model for each region in both periods. According to Figure 5, for most regions, the influence of neighbor-nation competitive effect and region-neighbor competitive effect enhanced, no matter whether it was positive or negative. According to the data from these two periods, the spatial spillover effects increased significantly. Therefore, it can be inferred that promoting interregional resource flow and cooperation will be more conducive to the development of China's food processing industry.

| CON CLUS IONS
In recent years, the growth rate of China's food processing industry has slowed down, and the Chinese government is guiding  The implications of this study are as follows. Firstly, with the increase in people's income and the improvement in living standards, the diet structure of Chinese people has become diversified.
Therefore, the intake of meat, seafood, soy products, vegetables, fruits, and nuts has increased, while the intake of staple foods such as rice and flour products has decreased. The demand for green and high-quality food products increased, while the demand for ordinary food products decreased. For China's food processing industry, it is both a challenge and an opportunity. Secondly, China's food processing industry has slowed down in growth and

CO N FLI C T O F I NTE R E S T
The authors declare no conflict of interest.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the National Food and Strategic Reserves Administration of China.
Restrictions apply to the availability of these data, which were used under license for this study. Data are available from the authors with the permission of the National Food and Strategic Reserves Administration of China.