The Evaluation of Shared Prosperity: A Case from China

Abstract

This research investigates the disparities, trends and spillovers of shared prosperity for all in China during the period of 2012–2021. Taking a representative region consisting of 18 urban and rural areas as a case study, using 10 indicators such as economic development, population density and education level, along with the spatial lag model, we explore the impact of social and economic factors on common prosperity as well as the associated spillovers. Results revealed that there existed huge regional disparities in common prosperity in the short term, namely the unbalanced level of prosperity across China’s mainland, while in the long term, the common prosperity level appears to be gradually enhanced with the convergence of income ratio lines. Meanwhile, common prosperity is spatially correlated with each other, with the spatial distribution features of high–high and low–low agglomerations. Based on the model analysis, there are mixed spillovers in the evolution of common prosperity: factors like education level and population density have positive spillovers while the rest of the factors have negative spillovers. To recap, population density and education level can significantly abridge the disparities in urban and rural areas.

1. Introduction

Shared prosperity implies that the fruits of socioeconomic development should benefit people in all segments of society in an inclusive and sustainable way. Most importantly, economic achievement is the source of shared prosperity. At present, the global economy remains sluggish. China and the US economy are growing relatively rapidly, followed by India and other global south countries, while the European economy is kind of still. Indeed, growing the economy is pivotal for all countries, because without it, the other 16 sustainable development goals (SDGs) will never be achieved, let alone shared prosperity. In particular, the first goal, namely poverty eradication (SDG1), the second—zero hunger (SDG2), the fourth—quality education (SDG4), the eighth—decent work and economic growth (SDG8), the ninth—industry, innovation and infrastructure (SDG9) and the tenth—reduced inequality (SDG10), are closely related to this research [1]. Notably, goals such as SDG4, SDG8, SDG9 and SDG10 are fundamental to gaining and keeping prosperity for both developed countries and developing ones. Thus, shared prosperity is closely related to the SDGs in a global sense. In other words, the latter four goals mentioned above can be viewed as primary factors that impact the evolution of shared prosperity.

Traditionally speaking, there are two types of countries, i.e., developed and developing countries. Currently the former are also called the global north and the latter the global south. Developing countries outnumber developed ones already. Roughly speaking, developing countries or global south countries are relatively poor or less developed, while the developed countries are relatively rich in terms of GDP per capita and technology power, among others. One can observe that shared prosperity depends largely on the quality-driven common development of developing countries, especially their economic vitality and resilience and sustainability as they account for over 70 percent of the world population and land area in total. Their economic growth rate has exceeded their developed counterparts in recent years and become key engines in boosting global economic growth.

Developing countries like Liberia, Burundi and other global south countries face even tougher challenges, inclusive of SDG1 and SDG2 as well. There are roughly 24 poorest countries in the world [2]. As of 2021, approximately 128 million people had faced or experienced the problem of hunger due to the impact of the COVID-19 pandemic [3]. Currently, China holds the rank of developing country because of its comparatively low GDP per capita and unbalanced regional development, among others. Moreover, China is a rags-to-riches tale of a developing country in the world. For example, China’s per capita GDP was less than one half the Asian average in 1978 [4]. However, China is currently the world’s second largest economy, contributing one-third to the world economy. More importantly, China has historically eliminated absolute poverty (SDG1) and has contributed significantly to global poverty alleviation. As the report of the 20th national congress of CPC pointed out, China adheres to a targeted poverty alleviation strategy and won the campaign of the largest scale of poverty reduction. As of the year 2022, 832 poverty-stricken counties across China have been entirely lifted out of poverty, approximately 100 million rural impoverished people escaped poverty, and over 9.6 million poverty-stricken population realized relocation.

In terms of food and energy security, the total output of grain ranks first in the world, which effectively guarantees the food security of 1.4 billion Chinese people [5]. When it comes to nutrition, data show that the Engel coefficient, a general indicator to measure people’s living standards proposed by [3], was 29.8% and 32.4% for China’s national average and rural Engel coefficient in 2023, respectively. Note that if the value of the Engel coefficient falls in the interval of 40–50%, it is in the rank of moderately prosperous, while the interval of 30–40% and 20–30% denotes the good living standard and the wealthy level, respectively. In addition, according to data from National Bureau of Statistics, China has a population of 1411.78 million, with 90.199 million, or roughly 64%, living in rural areas. Based on the above [3,5] as well as the most recent rural and urban population data, the overall living standard of Chinese people has reached moderately prosperous but is on the journey to good living standard and further to the wealthy level. Regarding the rural areas, overall, poverty and hunger as well as nutrition are no longer a problem [6].

Obviously, there is a close nexus between shared prosperity, SDGs and economic developments for both developed and developing countries in the modern world. In particular, shared prosperity can embody the essence of the SDGs and economic and social development. As such, it is meaningful to evaluate the progress of shared prosperity in depth, especially from the perspective of a developing country like China, using finer data spanning 2012–2021.

The main contributions of this paper are threefold: first, it compares three related concepts, namely shared prosperity, common prosperity and universal affluence; to the best of our knowledge, this might be the first article to distinguish between those synonyms, thus enriching the literature of common prosperity. Second, it analyzes nine factors that may affect the evolution of common prosperity in rural–urban areas in China and beyond, which are used to explore in depth the impact of these components on the evolution of shared prosperity, therefore contributing to the research through the lens of a developing country. Finally, the research employed the recent finer data along with a spatial lag model to capture the mixed spillover effect of common prosperity. This also makes it possible to examine the links between the factors and common prosperity from a spatial perspective. According to the extant literature, there are few, if any, studies applying this model to conduct a shared prosperity-themed analysis.

This paper is organized as follows. The background literature section follows the introduction, and Section 3 introduces the materials and methods. Section 4 presents the empirical results. Section 5 is a discussion of the results. The final section concludes the paper.

2. Background Literature

As previously mentioned, three related terms are applied to decode prosperity, namely, universal affluence/opulence, common prosperity and shared/global prosperity in literature. The concept of universal affluence was put forward by [7] in the late eighteenth century, who highlighted that due to the division of labor, universal opulence extends itself to the lowest ranks of the people. In addition, universal affluence was described thus: ‘It is the great multiplication of the productions of all the different arts, in consequence of the division of labor, which occasions, in a well-organized society, that universal affluence which extends itself to the lowest ranks of the people’ [7]. Prior to Smith, in 91 B.C., [8] documented that in managing a nation, it is of huge significance to enrich the people. It embodied the idea of common prosperity in ancient China. Like ‘universal affluence’, ‘common prosperity’ per se is no new concept either. Basically, it is similar to both ‘universal affluence’ and ‘shared prosperity’ in the literature. It is only that the former is mostly used in China, while the latter two are used more widely and outside China. At present, China has already entered the stage of substantially promoting common prosperity, but there is no one-size-fits-all pattern in pushing this work. Due to the imbalanced development between urban and rural areas, the rural areas have lagged far behind during the process of national shared prosperity. Therefore, it is vital to focus on the concepts of inclusive growth and people-oriented development so as to achieve the goal of common prosperity [9,10], which is deeply rooted in traditional Chinese culture [11].

In a recent review of literature, ‘common prosperity’ is described as ‘People must work together to create better lives for themselves and ultimately achieve well-rounded human development and common prosperity for everyone’ [9]. The notion particularly underlies the people-centered philosophy in all aspects; however, it is not that people become rich at the same pace.

Regarding the most recently emerged term ‘shared prosperity’, it is one of the two goals for [12], namely to end extreme poverty for the 1.2 billion people who continue to live with hunger and destitution and to promote shared prosperity. In essence, ‘shared prosperity’ is the same as ‘inclusive economic growth’. Based on the above connotations, it is obvious that ‘common prosperity’ is closest to ‘shared prosperity’, but the latter is more prevalent, whereas universal affluence is mainly focused on production output and exchange in the context of labor division. Moreover, both shared prosperity and common prosperity focus more on inclusive development for all, especially those vulnerable and marginalized groups who receive little attention. In this sense, it is valuable to delve into shared prosperity or common prosperity. For simplicity, common prosperity and shared prosperity can be used interchangeably in this study. Regarding inclusive growth pattern, it is the synergy of economic growth, environmental protection and educational development, which should be encouraged as well [10].

Briefly, the concept of prosperity has evolved with time, from the original common prosperity or gòng tóng fù yù in Chinese, and the subsequent universal affluence/opulence, to the most recent shared or global prosperity. In a sense, each of the terms holds a significant position in the global history of economy. One can observe that there are similarities and differences between these three terms mentioned above, which are listed below in

Table 1. Comparisons between common prosperity, universal affluence and shared prosperity.

Concept	Source	Similarities	Differences
common prosperity or gòng tóng fù yù in Chinese	It was documented in 91 B.C. in Historical Record (Shiji) by Qian Sima, a renowned ancient Chinese thinker.	Pursuit of better lives and well-rounded human philosophy. Its basic feature is the broadly equitable distribution of national wealth among the ordinary people.	It evolved from the rural agriculture-based economy and is mostly used in China. It is currently embedded in the process of modernization with Chinese characteristics.
universal affluence or universal opulence	It was documented around 1776 in An Inquiry into the Nature and Causes of the Wealth of Nations by Adam Smith, a renowned British economist and philosopher.	Pursuit of sufficient wealth and a well-organized society. It underscores fairer and more accessible job opportunities for the general public, especially people in the lowest ranks of society.	It emerged prior to the industrial revolution and evolved from the commercial and trade-based economy. It highlights labor divisions and the associated social order in capitalist society.
shared prosperity or global prosperity	It was first employed by the World Bank Group in 2013.	Pursuit of inclusive development for all, especially the more vulnerable and marginalized groups. It is a global prosperity, taking the holistic development of all countries into consideration.	It has been widely used since 2013 on and is more closely connected with the 17 SDGs put forward by the United Nations. It focuses on common development and prosperity on a global scale.

.

Relatively speaking, although the SDGs are widely discussed and comprise most of the contents of common prosperity, there is scant empirical research on achieving pinpoint target through the lens of a country. Note that China had 770 million people who were in extreme poverty in 1978, of which 98 percent were in rural areas under the poverty line. However, empowered by China’s opening up policy and reform since 1978, along with growing its national economy, China has hugely reduced the global poverty rate from 44 percent to 9 percent [13]. Moreover, job creation, including but not limited to network marketing platforms or/and e-commerce for those who lack access to regular jobs, are the most fundamental means to poverty reduction and prosperity [14]. Essentially, only when every country has achieved the goal of common prosperity can one say that the universal prosperity is realized. In this regard, this paper attempts to fill in this gap in the literature and aims to empirically analyze the determinants of common prosperity from the viewpoint of a developing country like China.

So far, there is no consensus on the definition of developing and developed countries. The standards to distinguish the two kinds of countries are mostly based upon indicators such as the GDP per capita and differences in social, technological, educational, environmental and sustainable aspects. In practice, the classification is made according to their GDP, human development index, industry competitive index, etc. [15]. In our case, by the standard of GDP per capita, developing countries comprise most countries like China and India in Asia; Liberia and Burundi in Africa; and Bolivia in Latin America, to name only a few. As was mentioned previously, developing countries are relatively poor, whereas their counterparts are relatively rich based on the above criteria.

As developing countries outnumber developed ones, it is critical to focus on the progress of shared prosperity in the former. Only when developing countries achieve shared prosperity will the future of the world be prosperous and promising. Notably, being the largest developing country in the world, China has made tremendous progress since its reform and opening up in 1978 in terms of economic growth and poverty reduction, as well as other social and technological developments. It is a tale of rags to riches. In this sense, China can provide an exemplary case for most developing countries.

In addition, according to the seventh Chinese population consensus, the proportion of rural population accounted for 36.11% or about 51 million farmers in 2020 [16], and rural areas across China surpass 94 percent of the total national area [17]. In our case, the term ‘rural areas’ mostly refers to the scattered towns and counties where farmers live and work, while a large and densely populated area is referred to as urban area, including cities and their administrative districts and counties.

At present, China’s common prosperity evolution has gained its initial momentum; to keep the momentum going, targeted measures must be taken to promote the high-quality development of economy and society. In fact, to fulfill the targeted common prosperity, both economic efficiency and social equity should be enhanced in the path toward that goal. The features of high-quality development comprise the principal of ‘innovative, green, coordinated, opening up and sharing’. Pathways to obtain common prosperity consists of nurturing high-quality development, ameliorating income allocation, intensifying social safety net, maintaining effective digital governance and promoting high level opening up and reform [11]. Moreover, the rural demonstration projects can help accelerate the progress of rural common prosperity through a state–society interaction approach [18].

Given the present stage of common prosperity in China, what are the factors affecting this prosperity? Generally speaking, there are at least 10 factors that may affect the process of common prosperity, namely the urban–rural income gap, economic development, industrial structure, government scale, population density, education level, opening up, consumption power, finance level and foreign investment. Consequently, 10 variables were applied to conduct the subsequent study. Specifically, the urban–rural income gap factor denoted by Ratio, which is the urban resident disposable income per capita over that of their rural counterpart rather than [19], is used to capture the level of common prosperity; technically, the smaller the value of Ratio, the higher the level of common prosperity.

Regarding economic development, it is closely related with household income level. The logic is like this: if a poor area can take advantage of favorable policy and develop better, then its economy may grow significantly and help to boost common prosperity [20]. On the other hand, if the economy grows fast while less attention is paid to efficiency and fairness, then the urban–rural income gap expands [21].

Meanwhile, industrial structure also affects common prosperity, since unreasonable structure could widen the imbalanced development between urban and rural areas; for instance, areas with heavy industry might increase urban income levels while simultaneously enlarging the urban–rural gap [22]. Nevertheless, the upgrading of industrial structure could lead to changes to employment structure, boosting employment opportunities and channels for farmers, which is favorable for narrowing the urban–rural gap.

Similarly, consumption power is another driver of economic growth; the optimization of residents’ consumption structure can sustain economic growth and reduce the income gap between urban and rural areas accordingly. With respect to the national opening up policy, it will fuel economic growth, but this effect is akin to the Matthew effect, which triggers the Ratio in a vicious cycle [23].

Concerning government scale, it matters a lot in economic development. For instance, [24] found that governance level and economic growth are positively correlated. Taking fiscal tools as another example, fiscal spending is a powerful instrument for income redistribution, and therefore factors like taxation structure and the relations between government and market could exert dissimilar effects in different regions or residents, thus affecting the income level between urban and rural residents. In fact, how the government guarantee mechanism works in the fields of education, Medicare, and people’s other livelihood issues can make a difference [25].

Regarding the population factor, it will affect the labor market and further resident income. At present, it is undeniable that the problem of both an aging society and the one-way labor migration from the countryside to cities are much more severe in China’s rural areas. The population decline coupled with the aging society will have a negative impact on different regions [26].

The role of education development is uncertain, because it is likely to raise human capital via education and training in rural areas, while the expansion of education could intensify its imbalance and make a difference to human capital levels between urban and rural areas and, accordingly, the Ratio [27,28].

Finally, as to foreign investment, the larger the amount of investment, the more advanced industry and technology will be, and so it is with the local labor demand and productivity. Moreover, financial development can also boost local enterprises and raise GDP per capita.

There are three research questions in this study. The first research question is What are the main factors affecting common prosperity? Followed by How can the evolution trends of common prosperity be identified and better understood? The third one is How can the spillovers of common prosperity be captured using a spatial lag model?

3. Materials and Methodology

3.1. Background of the Case

In Figure 1, the study region, located in eastern China’s Shandong Province, is referred to as the Yimeng Region. It encompasses 18 districts or/and counties, each of which is labeled with one digit along with its proper names, through 1–18.

It has to be stressed that given the comprehensive strength among China’s 31 provinces, particularly economic power, Shandong Province ranks in the top tier. Despite the overall strong competitiveness of Shandong Province across China, the Yimeng Region shows predominantly low rural performance in terms of overall economic development level. As of 2021, the proportion of total population and regional GDP in the Yimeng Region was 15.1% and 8.7% of that of the whole province, respectively.

Due to historical reasons, for instance, one reason behind is that prior to 1949, before the foundation of China, the Yimeng Region, a mountainous area, used to be a well-known Chinese revolutionary base because of its geographical position, which is suitable for warfare at that time, but nowadays it has become an obstacle in terms of the inconvenient transport facilities, thus impeding common growth. Another reason is that after the reform and opening up policies implemented in the year 1978, due to urban bias and economic efficiency priority policy guidance, this clustered region developed at a relatively slower pace as a whole. Therefore, the development of Yimeng Region has lagged behind compared with the average level of the whole province. However, through the lens of the rural–urban integration system, it is roughly equivalent to the national average level of affluence in current China. Based on the above reasons, we choose the Yimeng Region as an example to explore the evolution of common prosperity in China.

3.2. Variables and Data

Our dataset is a dynamic panel data, which covers the 10 variables mentioned below, and the corresponding variable definitions are showcased in

Table 2. Variable definitions.

Variable Name	Symbol	Descriptions
common prosperity	Ratio	Disposable income per capita of urban residents over that of their rural counterparts.
economic development	Ln(GDP)	Logarithm of regional or county GDP.
industrial structure	str	Added value of secondary and tertiary industries over regional GDP.
government scale	gov	Fiscal expenditure over GDP.
population density	pop	Year-end population over area of administrative district.
education level	edu	Number of secondary school students over total population.
opening up	open	Total import and export volume over GDP.
consumption	con	Regional or county total retail sales of consumer goods over total population
finance level	fin	Year-end loan balance of regional or county financial institutions over regional GDP
foreign investment	FI	Region or county actual use amount of foreign capital over regional GDP

.

The 10 variables in Table 2 comprise 1 dependent variable—Ratio to denote common prosperity, or rather, the urban–rural income gap or simply gap; and 1 core independent variable, namely economic development. The remaining 8 variables, including industrial structure, government scale, population density, education level, opening up, consumption, finance level and foreign investment, are controlling variables.

Additionally, the controlling variables were utilized because economic development is mostly affected by various factors in the real world. To put it differently, controlling variables are independent variables as well; hence, there are 9 explaining variables in total, and by controlling the latter 8 variables in subsequent modelling, we can better understand how economic development affects common prosperity.

Regarding the raw data, they were obtained from various provincial statistical yearbooks published from 2013 to 2022, and the missing value was processed by linear interpolation. In line with the variable specification,

Table 3. Summary statistics.

Variable(s)	Observations	Mean	Std. Dev.	Min	Max
Ratio	180	2.307	0.314	1.641	3.015
Ln(GDP)	180	5.668	0.463	4.781	7.196
str	180	0.066	0.029	0.030	0.207
gov	180	0.886	0.077	0.586	1.146
pop	180	0.050	0.012	0.019	0.099
edu	180	0.130	0.039	0.047	0.224
open	180	0.123	0.181	0.003	1.411
con	180	1.776	0.996	0.736	6.786
fin	180	0.882	0.464	0.332	2.715
FI	180	0.013	0.019	0.000	0.140

provides summary statistics for the data.

3.3. Method

Firstly, given the nature of our data, it is proper to conduct both time and spatial analyses so as to analyze our research questions holistically and precisely. Since the study region contains 18 areas that are spatially interconnected, it is necessary to construct either adjacent matrix or geographic distance matrix or other types of matrices to delineate the nexus among neighboring geographical units. This is also the basis for spatial econometric analysis. According to our data, adjacent and geographic distance matrices are sufficient to meet our analytical demand.

Secondly, one needs to build Moran index (hereafter MI) to perform the spatial correlation test, which can be classified as global MI (GMI) or local MI (LMI). Technically, GMI refers to a rational number, whose values can be normalized to an interval set of [−1, 1] by global variance normalization. Specifically, if GMI is greater than 0, it means that the variables have positive spatial correlations; that is, similar regions have a significant space conglomeration effect. Conversely, if GMI is less than 0, it means that the variables have negative spatial correlations, namely that the dissimilar regions have a significant space conglomeration effect. The equations of GMI are listed below.

$$I={\displaystyle \frac{n\sum _{i=1}^n\sum _{j=1}^n{w}_{ij}\left(x_i-\overline{x}\right)\left(x_j-\overline{x}\right)}{\sum _{i=1}^n\sum _{j=1}^n{w}_{ij}\sum _{i=1}^n{\left(x_i-\overline{x}\right)}^2}}={\displaystyle \frac{\sum _{i=1}^n\sum _{j=1}^n{w}_{ij}\left(x_i-\overline{x}\right)\left(x_j-\overline{x}\right)}{s^2\sum _{i=1}^n\sum _{j=1}^n{w}_{ij}}}$$

(1)

$$s^2={\displaystyle \frac{1}{n}}\sum _{i=1}^n{\left(x_i-\overline{x}\right)}^2$$

(2)

In Equations (1) and (2), I represents GMI; X_i(X_j) the observations in sample region; $\overline{x}$ the mean of 18 sample regions; W_ij the element values of spatial weight matrix; and $s^2$ sample variance. With respect to LMI, its spatial correlation can use a scatterplot to analyze. To be specific, a scatterplot can be divided into four quadrants; each quadrant stands for one sort of spatial correlation pattern. Of these, the first quadrant (high–high, H-H) represents the club of high-level agglomeration; i.e., the high-level agglomeration area is surrounded by high-level ones; the second quadrant (low–high, L-H) indicates a low-level agglomeration area surrounded by high-level ones; the third quadrant (low–low, L-L),where a low-level agglomeration area is surrounded by low-level ones; and the fourth quadrant (high–low, H-L), where a high-level agglomeration area is surrounded by low-level ones. Mathematically, the local MI can be written as follows:

$$I={\displaystyle \frac{\left(x_i-\overline{x}\right)}{s^2}}{\sum }_{j=1}^n{w}_{ij}\left(x_j-\overline{x}\right)$$

(3)

where I signals LMI, and the other variables remain the same as those in Equations (1) and (2).

In terms of constructing the spatial weight matrix W, we employ two types of matrices, namely, an adjacent matrix (0–1 matrix) and a geographic distance matrix; the respective definitions are as follows. The adjacent matrix is a popular spatial matrix whose principle is simple: if areas i and j have a common border, then $W_{ij}$= 1; otherwise, $W_{ij}$= 0. Regarding the geographic distance matrix, it is based on the square of the reciprocal geographical distance. This paper takes the adjacent matrix and geographic distance matrix to explore the spatial relations of income level change between urban and rural residents. Particularly, we employ MI to judge whether the variables are spatially dependent on each other. Only when the variables are spatially interdependent can they be analyzed econometrically.

Thirdly, it is essential to build a spatial econometric model to further the empirical analysis. With respect to the commonly used spatial econometric models, there are spatial error model (SEM), spatial lag model (SLM), and spatial Durbin model (SDM), which is the combination of the former two. It is worth mentioning that in order to determine the right form of spatial econometric model, we first use MI to conduct spatial correlation test to ensure the feasibility of building a spatial panel data model. The following step is to adopt Lagrange multiplier (LM) test to analyze whether SEM and SLM are significant or not [29]. In our case, it revealed that the lag effect is significant at 1 percent level, and thus we chose the SLM, then applied the Hausman test to ascertain whether the random effect model should be chosen; coupled with likelihood ratio (LR) test, which demonstrated that time fixed effect and individual fixed effect are both statistically significant, thus proving that SLM with double fixed effects is the right model to construct. Accordingly, the empirical SLM applied in this study is specified in Equation (4):

$$\begin{array}{ll}{Ratio}_{it}=& {\beta }_0+\rho W{Ratio}_{it}+{\beta }_1\ln \left({GDP}_{it}\right)+{\beta }_2{str}_{it}+{\beta }_3{pop}_{it}+{\beta }_4{edu}_{it}+\\ & {\beta }_5{gov}_{it}+{\beta }_6{open}_{it}+{\beta }_7{con}_{it}+{\beta }_8{fin}_{it}+{\beta }_9{FI}_{it}+{\epsilon }_{it}\end{array}$$

(4)

where ${\beta }_0$is a vector of intercept terms,$W$ is the spatial weight matrix, and${Ratio}_{it}$ denotes the income ratio of urban and rural residents for area i in year t; $\ln \left({GDP}_{it}\right)$ is economic development; ${str}_{it}$ denotes industrial structure; ${pop}_{it}$ denotes population density; ${edu}_{it}$ signals education level; ${gov}_{it}$ signals government scale; ${open}_{it}$ is opening up level; ${con}_{it}$ is consumer consumption level; ${fin}_{it}$ is financial development level; $\rho$ is the spillover effect of Ratio, and ${\epsilon }_{it}$ is the error term.

In addition, amidst the above steps, robust checks, including employing geographical distance weight matrix as well as heterogeneity test, were conducted to make sure that the finding results associated with modelling are reliable. In summary, the above blended methods and steps are delineated in Figure 2.

In Figure 2, primary steps like spatial correlation test, and model regression are presented on the far left, while the methods used are showcased on the far right, and the middle section is the core of this study.

4. Results

4.1. Time Trends of Shared Prosperity

Figure 3 depicts the evolution of common prosperity in the period of 2012–2021. In Figure 3, the legends marked by digits 1–18 possess the same respective specifications as those in Figure 1, where the horizontal axis represents the time denoted by year, the vertical axis represents the urban–rural income gap denoted by Ratio as was mentioned previously, and the colored lines represent different districts or counties of the sample region. Given the specific value of each Ratio, one can see they all varied in the sample period, specifically before 2015, which fluctuated more heavily compared with the change after 2015. Interestingly, 3 pairs, namely 1 (Lanshan District) and 16 (Tai’an County), 2 (Luozhuang District) and 17 (Ju County) and 3 (Hedong District) and 18 (Ju County), demonstrated almost identical trends. In addition, 2 (Luozhuang District) and 17 (Ju County), 15 through 18, i.e., Sishui County (15), Tai’an County (16), Wulian County (17) and Ju County (18), all displayed a relatively small magnitude of fluctuation but an encouraging downward trend as a whole. Note that the lower the Ratio, the better the evolution of common prosperity, namely that the regional urban–rural income disparities are dwindling continuously.

Figure 4 describes the evolution of common prosperity in sample area using a kernel density approach. To explore the dynamic features of common prosperity, we employed Matlab to generate the three-dimensional (3D) kernel density graph to demonstrate the changes in urban–rural income disparity more visually.

In Figure 4, the kernel density curve exhibited a flat but wide shape, with a smaller peak value and larger horizontal breadth, as well as the discrete data distribution, revealing a vast disparity in regional common prosperity in the period 2012–2014. Nevertheless, from the year 2016 on, this kernel density curve shifted right in the shape of a waveform, implying the constant expansion of the Ratio, or rather, the temporary retrogress of common prosperity. Meanwhile, the breadth of the waveform diminished, whereas the vertical height of the wave crest increased gradually, indicating that the data distribution in 2016–2021 was growing denser and the Ratio was dynamically converging. In other words, the degree of common prosperity was enhanced gradually, which is consistent with the previous time trend analysis using a two-dimensional graph approach.

In summary, Figure 3 and Figure 4 showcase the dynamic tendency of common prosperity using two different approaches. Overall, Figure 3 illustrates the changes in income gap between urban and rural residents; hence, the common prosperity in the sample region tends to flatten out in the long term, albeit being steeper in the short term of 2014–2016. It is worth noting that most units emerged their peak value in 2015. As expected, the overall tendency of common prosperity moves downward in the long term, indicating the promising signs of common prosperity in the Yimeng Region year by year. Similarly, Figure 4 illustrates that the income gap of urban and rural residents in sample areas appears to close in the long term. In a sense, that trend is encouraging and may be attributed to sustainable development in either the form of the digital economy or the green transition in rural areas.

4.2. Spatial Correlation Tests of Shared Prosperity

Table 4. GMIs in 2012–2021.

Year	0–1 Matrix GMI	p-Value	Geographic Distance Matrix GMI	p-Value
2012	0.209	0.070	−0.059	0.034
2013	0.201	0.085	−0.059	0.021
2014	0.026	0.589	−0.059	0.143
2015	0.360	0.007	−0.059	0.006
2016	0.390	0.004	−0.059	0.004
2017	0.375	0.006	−0.059	0.005
2018	0.375	0.006	−0.059	0.005
2019	0.374	0.006	−0.059	0.005
2020	0.376	0.005	−0.059	0.004
2021	0.373	0.006	−0.059	0.005

provides the respective GMI for the period 2012–2021 based on the adjacent matrix and the geographic distance matrix. Results show that all the MIs in 0–1 matrix are volatile slightly in sample period. Specifically, except for 2014, the Ratio is significant at 10 percent level in terms of spatial correlation, while in geographic distance matrix, it is significant at 5 percent level. On the whole, based on Ratio, common prosperity is spatially correlated with each other during the period of 2012–2021.

In Table 4, even though the GMI in 2014 is insignificant, it still might illustrate that the spatial agglomeration of common prosperity exists, because the rest of the GMIs in their corresponding years are remarkably significant.

In the meantime, the LMI of 2013, 2017 and 2021 are presented in Figure 5 to demonstrate the spatial state of common prosperity in the sample region. As is shown in Figure 5, according to the adjacent matrix, despite the different spatial distribution features of Ratio of various districts and counties, they showcase a stable but positive correlation in nature, generally with the characteristics of the high–high and low–low agglomeration effects. According to geographic distance matrix, compared to 0–1 matrix, the Ratios of various districts and counties present quite similar spatial distribution features of high–high and low–low agglomerations.

In Figure 5, the Ratios of the 18 regions display a declining trend starting in 2016, reflecting the promising progress of shared prosperity year-on-year. In light of Figure 4, local Moran scatterplots using either adjacent matrix or geographic matrix consistently confirm the general positive correlations, which is appropriate to build a SLM in the subsequent model regression procedure.

4.3. Results of Spatial Autoregression Analysis of Shared Prosperity

Table 5. Results of spatial lagged model.

Variables	Adjacent Matrix	Geographic Distance Matrix
Ln(GDP)	0.528 * (0.003)	0.543 * (0.003)
str	0.957 * (0.007)	0.940 * (0.010)
open	0.449 * (0.000)	0.459 * (0.000)
gov	3.457 * (0.004)	3.673 * (0.002)
edu	−10.639 * (0.002)	−10.798 * (0.002)
pop	−3.640 * (0.002)	−3.694 * (0.002)
con	0.003 (0.942)	0.002 (0.958)
fin	−0.022 (0.423)	−0.028 (0.325)
FI	0.025 (0.958)	0.035 (0.943)

Notes: p value is in parenthesis. * denotes 10% significance level.

provides the results of the SLM. In Table 5, regarding the adjacent matrix, among the various determinants of common prosperity, factors including economic development level, industry structure, degree of opening up and government scale are positively correlated with urban–rural income Ratio, while education level and population density are negatively correlated with Ratio. To ensure the reliability of empirical results, we used geographic distance matrix to conduct the corresponding robustness test; the coefficients and signs in the latter are consistent with those of the former, thus proving the results are reliable.

4.4. Heterogeneity Test

Table 6. Results of heterogeneity test.

Variables	Group 1 (Marked 1–12)	Group 2 (Marked 13–18)
Ln(GDP)	0.789 ** (0.038)	0.146 *** (0.082)
str	0.336 (0.318)	0.540 * (0.011)
open	0.006 (0.988)	0.168 * (0.001)
gov	5.112 * (0.014)	−0.350 (0.582)
gov	−16.641 * (0.002)	−3.297 *** (0.073)
pop	−6.690 * (0.004)	10.219 ** (0.032)
con	0.050 (0.372)	0.025 (0.102)
fin	−0.083 (0.152)	−0.038 ** (0.046)
FI	0.561 (0.133)	1.121 * (0.002)

Notes: p value is in parenthesis. *** 1%, ** 5% and * 10% significance level.

showcases the results of the heterogeneity test. In Table 6, due to local differences, we split the total 18 regions into two sub-samples based on their actual development situations; specifically, areas 1–12, belonging to Linyi City, are classified as Group 1, while the other 6 areas (13–18), located outside Linyi City, are classified as Group 2. As is shown in Table 6, the heterogeneity test results suggest that in both groups, the economic development level retarded the decrease pace of the income gap, while education level reduced the Ratio. Meanwhile, the upgrading of industry structure and the deepening of opening up to the world significantly hindered the shrinking of the income gap for both groups. Concerning population density and government scale, they played differing roles in terms of accelerating common prosperity in our case.

4.5. Spillover Effect of Shared Prosperity

Table 7. Spillover effect of shared prosperity.

Variables	Direct Effect	Spillover Effect	Total Effect
Ln(GDP)	0.549 ** (0.000)	0.236 ** (0.040)	0.785 *** (0.000)
str	−3.826 ** (0.011)	−1.657 * (0.096)	−5.484 ** (0.016)
open	1.018 *** (0.000)	0.442 * (0.065)	1.460 *** (0.002)
gov	−11.083 *** (0.000)	−4.828 ** (0.036)	−15.911 *** (0.000)
edu	3.606 *** (0.000)	1.545 ** (0.044)	5.515 *** (0.001)
pop	0.468 *** (0.000)	0.204 ** (0.0304)	0.672 *** (0.000)
con	0.004 (0.897)	0.002 (0.898)	0.006 (0.895)
fin	−0.024 (0.596)	−0.010 (0.655)	−0.034 (0.608)
FI	0.074 (0.902)	0.027 (0.925)	0.101 (0.908)

Notes: p value is in parenthesis. *** 1%, ** 5% and * 10% significance level.

showcases the results of the decomposed effects of the above SLM. As is shown in Table 7, there exist spillovers in the process of common prosperity. Specifically, variables like economic development, industrial structure, opening up, government scale, education level and population density all impact common prosperity at different significant levels. Among them, economic development level, opening up, education and population had positive signs, while industrial structure, government scale and finance had negative signs. The signs of the corresponding coefficients are consistent with those of the above analysis, thus confirming the reliability of the associated conclusions from a novel perspective.

5. Discussion

Shared prosperity is a major goal for China to fulfill around 2049. The Yimeng Region can serve as a case of the general level of common prosperity across Chinese mainland. Due to historical reasons, the Yimeng Region has lagged behind in economic development. Determinants as aforementioned cover both the economical and societal aspects. In this study, we chose the period of 2012–2021 to capture the spillover effects of common prosperity for all in China. The year 2012 is an important milestone in common prosperity, when it was placed on the agenda by China’s central government. The year 2021 is also important because it revealed the impact of the COVID-19 pandemic on economic growth and thereby common prosperity in China. Table 5, Table 6 and Table 7 report the key modelling results. When compared with other similar studies, there are more dissimilarities than similarities to our case.

Table 5 revealed that the enhancement of economic development level significantly expanded the gap between residents in rural and urban areas, which is consistent with the conclusion of [25], who conducted a Granger causality test on the relationships between China’s economic growth and its income gap between urban and rural areas and found that the former may restrain the shrinking of the latter. However, this contradicts [20], who argued that economic growth can boost wealth and enhance efficiency based on the test of the nexus between economic growth and income gap. Regarding industrial structure, it is positively associated with the Ratio: every unit increase of non-agricultural level results in an average 0.949 unit increment of the Ratio; the bigger the proportion of secondary and tertiary industries, the larger the percentage of the Ratio in the Yimeng Region. Thus, the development of non-agricultural industries is unable to narrow the disparities in urban and rural citizen income levels. Refs. [26,27] also observed, using provincial panel data, that the upgrading of industrial structure would enlarge the income gap.

As to the level of opening up, in both matrices, the corresponding coefficients are highly significantly positive, implying the greater the level of opening up, the larger the urban–rural income gap. Indeed, only high earners can benefit from opening up, while rural residents cannot take full advantage of overseas markets [28]. The possible reasons behind were that the opening up allows urban citizens to have more dividends while expanded the gap between the urban and rural regions simultaneously. From the perspective of [30], the optimization of trade structure may narrow the urban–rural income gap, so there might be unreasonable trade structure in the Yimeng Region.

Also in Table 5, regarding government scale, we found as Ratio grows, the income gap between urban and rural areas expanded with the increase of government scale, which is consistent with [31], who proved that initial distortions coupled with urban fiscal deviation may enlarge disparities in urban and rural income; fiscal expenditure had a low efficiency regarding regional synergy effect and thus did not narrow the gap between urban and rural areas. This is likely to trigger the expansion of the income gap. In addition, even if local government augments fiscal expenditure, it does not narrow the income disparities, due to the fact that there are too many low-income groups in rural areas and much greater expenditures are needed, but local governments cannot satisfy this demand [32].

The parameter of education level was negatively correlated with Ratio, which means that the development of education can help reduce urban–rural income gap in our case. However, our conclusion contradicts [24,33,34,35], who concluded that improper allocation of education resources was a crucial reason that resulted in the current expansion of the income gap and common prosperity. Furthermore, the gap is exacerbated via inter-generational transmission.

Concerning population density in Table 5, it is also negatively associated with Ratio, implying that population increase can reduce the urban–rural income gap. With respect to factors like consumption level and foreign investment level, they are positively correlated with Ratio, albeit at an insignificant level; the level of finance development is insignificantly negatively associated with Ratio. Nevertheless, financial development or financial inclusion is still crucial in terms of poverty reduction and further common prosperity, especially in rural areas, in a global sense [36]. Although the above three factors have little effect on Ratio, due attention should still be paid in future studies.

In Table 6, in the benchmark result in Group 1 column, the effects of financial development and foreign investment were less significant. By contrast, regarding the effects revealed in the Group 2 column, financial development level can considerably reduce Ratio. It is presumable that the latter had poor finance infrastructure or there were high entrance barriers to the financial market, thereby producing a less desirable common prosperity spillover effect. In addition to that, due to urban bias, fewer resources are allocated to rural areas, thus impeding the pace of prosperity in the latter.

In Table 7, one can find that there are mixed spillover effects associated with shared prosperity. In particular, three variables, including industrial structure, government scale and finance development, have negative coefficient signs, indicating negative spillovers. This finding is inconsistent with [37] in terms of the spillover direction in the urban–rural tourism industry. By contrast, the rest of the corresponding coefficients present the expected positive signs of the relations between the variables and common prosperity. In other words, these factors empower the advance of shared prosperity through positive spillovers.

Last but not least, there are scant empirical research results from the international academic community regarding this prosperity theme. In particular, the Belt and Road Initiative (BRI) was commissioned as a study for a broader perspective. Some researchers have acknowledged the logic of BRI rather than the concrete social system behind it. Based on our empirical results, one can derive the conclusion that the goal of shared prosperity cannot be achieved overnight. Essentially, various factors can influence economic prosperity and further shared prosperity. In a much broader sense, the level of innovation along with the path to shared prosperity in developing countries will contribute to the realization of SDGs [38]. Despite these achievements of shared prosperity, challenges for future economic growth of developing countries remain, as evidenced by the less developed rural inclusive finance, low levels of public consumption, unequal allocation of education resources between rural and urban areas, unsustainable industrial structure, etc. These challenges may result in relatively low economic growth and hence less sustainability in shared prosperity in the global south. Consequently, supportive strategies as well as flexible policies including but not limited to BRI and the global development initiative (GDI) are needed to put in place to tackle these prominent problems so as to achieve global prosperity.

6. Conclusions and Limitations

6.1. Conclusions

This research was conducted in five steps. First, time and spatial analysis of the data were investigated so as to better understand the general trends of shared prosperity in sample country. The results revealed that the evolution of common prosperity are promising, as expected, because both the Ratios and kernel density exhibit diminishing trends.

In the second step, spatial correlation tests of shared prosperity were performed using GMI and LMI along with two spatial weight matrices. The results confirmed that, according to GMI, on the whole, common prosperity among the 18 regions in the Yimeng Region is spatially correlated with each other in the given period of 2012–2021; given LMI of the above mentioned two kinds of matrices, both feature similar spatial distributions of high–high and low–low agglomerations.

In the third step, the regression of the SLM was employed to study the effect of various variables on common prosperity. The results indicated that, first and foremost, the low and imbalanced level of rural education, with the parameter being −10.72, is a key factor that may significantly abridge the income gap. Consequently, to reasonably allocate education resources coupled with improving rural education level can help mitigate the present imbalanced situations and promote common prosperity.

Regarding population density, with the mean being −3.67, it is conducive to narrowing the urban–rural income gap, but the current rural labor migration to cities has enlarged low-income groups, which can be addressed by increasing population density in rural areas in the future. This conclusion is also applicable to other regions due to the pace of societal aging in China. Moreover, elements like economic development level, industrial structure, the degree of opening up and government scale all expanded the income gap in our case, implying these factors may negatively impact common prosperity through poor allocation effect.

In the fourth step, the robust check using heterogeneity test was performed to validate the benchmark regression results of the third step. In light of the result of heterogeneity test, it is noticeable that population density and government scale had different roles in two groups, which might be attributable to imbalance regional development as well as policy heterogeneity.

Lastly, in the fifth step, one can observe there is spillover effect resulting from common prosperity, revealing that the level of common prosperity associates directly with local economic development and industrial structure, meanwhile, indirectly affects common prosperity of adjacent areas in one way or another.

6.2. Limitations

This study has some limitations. Firstly, it primarily touches upon six UN SDGs that are pertinent to shared prosperity with Chinese characteristics and further extracts the corresponding factors from the goals for empirical analysis. For broader generality, subsequent research should take more common cases to further enhance the sample representation and validation. Secondly, the focus was on developing countries like China, distinct from some developing countries that have no aging population issue and employ different state interventions. Given the small sample size, the conclusions of this study are not applicable universally. Studies in varied social systems and culture contexts may provide more comparative and comprehensive insights. Finally, our evaluation for common prosperity utilized specific variables linked to shared prosperity for all.

In future research, we will adopt diverse variables such as Gini coefficient and Theil index for a refined design. Apart from that, we will develop our study in a much broader sense, taking the example of the BRI, implemented a decade ago, which was participated in by over 150 countries globally. Through cooperating in global trade, manufacturing and infrastructure, the BRI has created considerable job opportunities and revenue, thus producing positive spillovers and pressing common development and shared prosperity in a global sense, which deserves to be further delved into in subsequent studies.