Income inequality represents a common challenge in the economic development process of major global economies. However, existing research has insufficiently examined the logical relationship embedded within environmental constraints among industrial upgrading, high-quality development, and common prosperity. This paper aims to address this gap by investigating how environmental issues arising from non-agricultural industrial development influence income distribution, thereby exploring the inherent connections between industrial transformation and shared prosperity under environmental limitations.
Based on World Bank data, which corroborates the environmental problems associated with non-agricultural industrial development, this study employs theoretical analysis and empirical modeling to elucidate the mechanisms through which environmental factors contribute to the widening income gap. The research specifically examines the lagged impact of environmental pollution on human capital and incorporates this temporal dimension into both theoretical frameworks and econometric tests.
The study reveals that environmental pollution exerts a delayed effect on human capital, creating a “concealment effect” that obscures its full socio-economic impact. This phenomenon provides a theoretical and empirical explanation for the prevalent “pollute first, treat later” approach in industrial development. Furthermore, the process alters the demand structure of labor factors, enlarges the skill premium, and consequently worsens the distribution pattern of factor income, thereby exacerbating income inequality.
This paper innovatively integrates environmental constraints into the analysis of the “industrial upgrading high-quality development-common prosperity” nexus, highlighting the concealed and lagged influence of pollution on human capital and labor markets. It provides a new perspective on how environmental degradation indirectly aggravates income disparity through changes in factor demand and skill premium expansion. The findings underscore the necessity of considering the synergistic interaction among technological upgrading, industrial coordination, and green production in fostering sustainable and inclusive growth on the new journey toward comprehensively building a modern socialist country.
1. Introduction and a literature review
The 20th report of the Communist Party of China points out that China-style modernization is the modernization of common prosperity for all people, but the income gap is a common problem faced by the economic development of major economies. As shown in Figure 1, the Gini coefficient in the USA rose from 0.423 in 1978 to 0.494 in 2021, and the United Kingdom once surpassed a high of 0.4 between 2000 and 2003. China is no exception. According to the data released by the National Bureau of Statistics, the Gini coefficient of Chinese residents was 0.317 in 1978, but it has continued to rise since then. The high period was from 2003 to 2012, between 0.47 and 0.49. After reaching the highest 0.491 in 2008, the overall trend gradually fell to 0.466 in 2021, but it was still hovering at a high level.
The horizontal axis of the line graph lists years from 1978 to 2022 in increments of 2 years. The left vertical axis ranges from 0 to 100 with increments of 10. The right vertical axis ranges from 0 to 0.6 with increments of 0.1. Two lines are plotted. The legend at the bottom identifies the dashed line as “Non-agricultural added value of industries accounted for (percent)”. The other line represents “gini”. The left axis is for “Non-agricultural added value of industries accounted for (percent)” and the right axis is for “gini coefficient”. The line for “Non-agricultural added value of industries accounted for” starts at 72 percent in 1978, increases with smoothly to 80 in 1995, and continues upward and ends at 92 in 2022. The line for “gini coefficient” starts at 0.31 in 1978, increases with fluctuations to 0.48 in 1996, and then stabilizes after 2003 with value of 0.47, and ends at 0.47 in 2022. Note: All numerical values are approximated.The ratio (%) of non-agricultural industry added value in 1978–2022
The horizontal axis of the line graph lists years from 1978 to 2022 in increments of 2 years. The left vertical axis ranges from 0 to 100 with increments of 10. The right vertical axis ranges from 0 to 0.6 with increments of 0.1. Two lines are plotted. The legend at the bottom identifies the dashed line as “Non-agricultural added value of industries accounted for (percent)”. The other line represents “gini”. The left axis is for “Non-agricultural added value of industries accounted for (percent)” and the right axis is for “gini coefficient”. The line for “Non-agricultural added value of industries accounted for” starts at 72 percent in 1978, increases with smoothly to 80 in 1995, and continues upward and ends at 92 in 2022. The line for “gini coefficient” starts at 0.31 in 1978, increases with fluctuations to 0.48 in 1996, and then stabilizes after 2003 with value of 0.47, and ends at 0.47 in 2022. Note: All numerical values are approximated.The ratio (%) of non-agricultural industry added value in 1978–2022
In fact, income comes from production activities, and the relationship between the income gap and industrial change is an empirical law to follow. Since the Industrial Revolution, technological progress has promoted the rapid transformation of the industry, and thus realized the increase of the total income (Yuezhou and Yifu, 2017; Jiaxu and Yao, 2020). In this process, the world’s major economies have generally experienced the process of increasing the proportion of the secondary and tertiary industries and the widening income gap (Acemoglu and Guerrieri, 2008; Acemoglu et al., 2016). Recently, the literature has started to systematically analyze the relationship between the two from the theoretical or empirical aspects (Alvarez-Cuadrado et al., 2018), which provides an important theoretical basis for further narrowing the income gap of Chinese residents. And the empirical tests based on these studies also confirm the causal relationship between the non-farm proportion and the widening income gap (Wanzong et al., 2018; Grossman and Oberfield, 2022).
Therefore, to promote common prosperity in high-quality development, the overall productivity level and productivity structure shaped by the industrial structure and the relationship between various industries are the key, which fundamentally determines the quality and height of the common prosperity of a society. The formation of China's income distribution pattern itself is carried out against the background of industrial adjustment. The proportion of the added value of the three industries in GDP changed from 28.2%, 47.9% and 23.9% in 1978 to 10.1%, 45.3% and 44.6% in 2012. In 2013, the added value of the tertiary industry exceeded that of the secondary industry for the first time. By 2022, the added value of the three industries was 7.3%, 39.9% and 52.8%, respectively. The transformation and upgrading of industrial structure has become an important economic feature under the new development stage. In the process of industrial adjustment, the competition between capital and labor never stops. In the early days of capital liberalization, capital profit takes the upper hand. In the later stage, imitation and innovation advance together, technology upgrades, and industrial structure changes continuously, presenting a situation of capital and labor competition driven by the two wheels of capital and technology. Compared with the fluctuation characteristics of China’s Gini coefficient, it is not difficult to find that the high hovering period of China’s Gini coefficient is also the critical period of industrial structure adjustment.
According to the perspective of new structural economics, the transformation and upgrading of industrial structure and the sustained high-quality economic growth mainly depend on the resource endowment structure of a country and a region and its dynamic changes (Lin, 2018). However, empirical research shows that there are many contradictions in “industrial development based on ecological occupation” (Hong and Zhenxing, 2017). The development of industry has always been accompanied by environmental pollution. A large number of studies have shown that the reform of industrial structure leads to environmental deterioration (Grossman and Krueger, 1995). The London smog event and the Los Angeles haze event are all typical cases that scholars pay attention to in the research process. The climate targets achieved in the Paris Agreement make deep decarbonization of carbon-intensive industries crucial (Xianhua and Qingquan, 2016; Jianlei et al., 2018; Bachner et al., 2020a, b). Furthermore, the academic research on the relationship between environmental pollution and income distribution not only discusses the issue of welfare distribution from a macro perspective (Nam et al., 2010; Ravallion et al., 2000), but also explains the impact of environmental quality on people from the micro perspective of healthy human capital (Janke, 2014).
China also has such problems. How to narrow the income gap and achieve common prosperity in the new journey of building a modern socialist country with comprehensive prosperity? In addition to clarifying the law of industrial reform, we should also explore the characteristics of environmental constraints, and then further analyze the realistic path of forming the income gap (Chinese Style Modernization Research Group, 2022). The task of jointly promoting “carbon reduction, pollution, green and growth” is arduous, and the positive interaction between industrial reform and the improvement of distribution structure has become a difficult problem.
An analysis of China’s industrial reform and environmental protection process can be found that the industrial development process resonates with the evolution of the government’s environmental protection strategy and the change of the pollution situation. In 1974, the Leading Group for Environmental Protection of the State Council was formally established. In 1979, the Environmental Protection Law (Trial) was enacted. Later, the Marine Environmental Protection Law and other special laws were passed and amended several times. While improving the legal system, China also took an active part in international activities. In 1992, the United Nations Conference on Environment and Development adopted Agenda 21, and China attended the conference as a developing country. After the 18th National Congress of the Communist Party of China, ecological civilization construction has been paid unprecedented attention, and the “two-carbon” goal was formally proposed at the 75th Session of the United Nations General Assembly in 2020. During this period, China’s environmental pollution situation has also undergone phased changes. In 1978, China’s energy consumption and emissions were 1.42 billion tons, and reached 8.12 billion tons in 2011. Among them, from 2000 to 2011, the average annual growth rate reached 8.38%, which was a stage of rapid growth. In the following decade, from 2012 to 2021, carbon dioxide emissions increased from 9.00 billion tons to 11.47 billion tons, with an annual growth rate of less than 1.5%, a low growth stage. According to the characteristics of the World Bank data, China is also in line with the synchronous fluctuation of industry proportion and carbon dioxide emissions (Figure 2) [1].
The horizontal and vertical axes of the scatter plot are labeled in Chinese text. The horizontal axis ranges from 0 to 80 with increments of 20 units. The vertical axis ranges from 0 to 20 with increments of 5 units. Numerous circular data points are plotted across the graph, forming a dense cluster primarily between horizontal values of 5 to 50 and vertical values of 2.5 to 17. A straight fitted line is drawn across the scatter points. The fitted line in the scatter plot passes through the following data points: (5, 7), (20, 9), (40, 11), (60, 13), and (80, 15). Note: All numerical values are approximated.The relationship between the proportion of the secondary industry and the logarithm of carbon dioxide emissions in various countries since 1978. Source: The World Bank
The horizontal and vertical axes of the scatter plot are labeled in Chinese text. The horizontal axis ranges from 0 to 80 with increments of 20 units. The vertical axis ranges from 0 to 20 with increments of 5 units. Numerous circular data points are plotted across the graph, forming a dense cluster primarily between horizontal values of 5 to 50 and vertical values of 2.5 to 17. A straight fitted line is drawn across the scatter points. The fitted line in the scatter plot passes through the following data points: (5, 7), (20, 9), (40, 11), (60, 13), and (80, 15). Note: All numerical values are approximated.The relationship between the proportion of the secondary industry and the logarithm of carbon dioxide emissions in various countries since 1978. Source: The World Bank
The frequency of industrial structure adjustment in China is accelerating, especially with technological progress and the proposal of dual-carbon targets, and the industrial structure is undergoing rapid transformation (Shuai et al., 2022). How to deal with the relationship between technological progress and dual-carbon strategy and employment is a major issue that must be paid great attention to for a long time in the future. In formulating national economic plans and making major adjustments to industrial structure and industrial layout, common prosperity should be considered as a high-level goal of social and economic development, and a long-term mechanism of benign interaction between industrial structure adjustment, environmental protection and income distribution should be established and improved (Zhibiao and Yonghui, 2020).
In short, industrial upgrading is an important influencing factor for achieving high-quality development and common prosperity, while environmental constraints determine the difference in resource endowment, and ultimately have an impact on the logical relationship of “industrial upgrading-high-quality development-common prosperity”. However, the existing literature rarely considers the theoretical basis and empirical evidence from the perspective of industrial adjustment, environmental protection and income gap, so this paper tries to give a beneficial supplement. In addition, in many developed countries, there has been serious environmental pollution caused by industrial development, which is of great concern to the international community. However, why do many late-developing countries still take the path of “treatment first, pollution later”? There is no reasonable explanation. This is also the answer that this article hopes to find. This part is the introduction and literature review, mainly combing the existing research contributions; the second part is the theoretical model, aiming to find the relationship between industrial structure change, environmental change and income distribution; the third part is the empirical test to verify the influence path of industrial change and ecological change on income distribution with the World Bank data; and the fourth part is the conclusion and suggestions.
2. Theoretical analysis and research hypotheses
In this paper, the theoretical research model of Pengfei (2017) and other scholars is improved. Based on the current situation of technological differences between industries, the role principle of environmental pollution, health human capital investment and income gap between high-tech departments and low-tech departments is analyzed through a two-stage model.
Production model by department
Here, a production model is based on departmental technical differences. Similarly, to simplify the analysis, it is assumed that only one highly skilled worker and one low-skilled worker participate in economic activities, and the same consumption is equal during the interperiod consumption selection process, and the consumption of the second phase depends on the income of the first phase. In the first phase, it is assumed that highly skilled workers and low-skilled workers have the same healthy human capital. There is an inherent income level gap between industries, with highly skilled workers producing in high-end industries, low pollution and energy consumption, and high income; low-skilled workers, high pollution and energy consumption, and low income. The production function is expressed as follows:
In the first issue,
Among them, and represent the income obtained by the high-tech production workers and the low-tech production workers in the first phase respectively. The healthy human capital owned by the workers in the first phase is equal, and the marginal remuneration of the healthy human capital is unchanged, and there is no law of decline. And represent the material capital possessed by the workers in the high-tech production sector and the low-tech production sector, respectively, so that the material capital meets the law of diminishing marginal return. And for the technical level of the workers owned by the two departments, among which, the workers who master the technology will increase the income.
A two-sector production model considering healthy human capital investment
In the second phase, according to the pollution production function (Stokey, 1998, 2021), environmental pollution will reduce the human capital of workers, and workers can increase their human capital from the income of the first phase.
Among them, the environmental pollution level in the second phase is determined by the total output level and production cleanliness of the first phase. Since the total output level of the first phase is determined, the environmental pollution level of the second phase is determined by the production cleanliness. And is the income level of workers in the high-tech production and low-tech production sectors in the second phase, respectively; and indicates the healthy human capital level of workers in the two sectors; and the healthy human capital investment for workers in the two sectors.
The decision-making model of the workers in different departments
According to the consumption theory, workers can choose the optimal investment scale of healthy human capital investment in the second phase to achieve the maximum utility. The consumption decision model of workers in the high-tech production sector is as follows.
Equation (5) is the selection process of workers in the high-tech production sector, which represents the consumption of workers in the first phase and the second phase respectively, and is the discount rate. It indicates that the consumption of the workers in the high-tech sector is equal in the two periods, which meets the assumption of smooth consumption in the life cycle consumption theory. In order to simplify the analysis, the investment level of healthy human capital can be deduced as follows.
Similarly, the consumption decision model of workers in low-tech production sectors is as follows.
Available, the level of healthy human capital investment in the low technology sector is as follows.
Through Equations (7) and (8), it can be seen that when there is no environmental pollution, immediately, the healthy human capital investment of workers in the two departments is 0, and when there is environmental pollution, immediately, both workers will invest in healthy human capital, and the investment level will increase with the increase in environmental pollution level. The difference between the healthy human capital investment of the workers in the high-technology sector and the low-technology sector is as follows.
Because, it is not difficult to find that workers in the high-tech sector invest more in healthy human capital than those in the low-tech sector. This gap is not only caused by the income gap between different sectors in the first phase. With the increase in environmental pollution, the investment gap in health human capital investment of workers in the two sectors will widen. In fact, this is a model based on utility optimization, but in reality, it is difficult to say that there is an optimal health human capital input value. On the one hand, under the condition of the shortage of medical resources, it is difficult to make proactive health investment; on the other hand, human behavior is often short-sighted, health investment is a project of reducing income in the previous period, and the process of reducing health human capital caused by environmental pollution is often not foreseen in the short term.
The measure of the income gap
Further, in order to measure the income gap between the two sectors in the second phase relative to the first phase, it can be defined as follows.
However, the gap in healthy human capital investment leads to a greater difference in the healthy human capital of workers in the two sectors in the second phase, and expands the income gap between workers in different sectors in the second phase.
From the above equation (9),
The income gap in the second phase has widened compared with the first phase, which is mainly caused by the difference in the health human capital investment level of workers in different sectors caused by environmental pollution. If it is assumed that the level of health human capital is optimal in the first phase, then the investment difference, which means that the pollution caused by the first phase of production activities will have a negative impact on the health human capital of workers. As it indicates that the difference in healthy human capital investment caused by environmental pollution will widen the income gap.
At the same time, it can also be judged that the second phase of healthy human capital level investment is subject to environmental pollution, which makes different workers in different sectors have heterogeneous investment. Further, we can find that about the partial derivative,
Obviously, if the technical level of the low-tech sector remains unchanged, the income gap between the two phases will further widen as the technical level of the high-tech sector improves.
Similarly, if the partial derivative can be found,
If the technical level of the high-tech sector remains unchanged, the income gap between the two phases will be further narrowed with the improvement of the two phases.
Formula (12) and Formula (13) show that technology convergence between high-tech and low-tech sectors helps to narrow the income gap. It can be seen that the coordinated development of technology among the three industries is more conducive to narrowing the income gap.
Furthermore, assuming that the same material input is given to different sectors, in other words, taking a neutral attitude towards the investment in the high-tech and low-technology sectors, according to Equation (11), it can be found that the bias guidance,
This is an additional function, which means that if the same material input is indiscriminate to the high technology and low technology sectors, the more the investment, the greater the income gap between workers.
Research hypothesis
Based on the above theoretical analysis and factual characteristics, this paper proposes the research hypothesis. According to general experience, the current historical development of China and even the world is characterized by the non-agricultural sector being more technology-intensive compared with the agricultural sector. Therefore, it is advisable to simplify the analysis and initially test the influence direction, characteristics and results of the non-agricultural sector structure, that is, put forward hypothesis 1: the increase of the proportion of the non-agricultural sector has a significant positive effect on the widening of the income gap. On this basis, given the model (9), historical data were used to verify the environmental problems caused by the increased proportion of the non-agricultural sector and the impact on the income gap. Considering the short-sighted nature of people and the non-short-term foresight of the reduction of healthy human capital, Hypothesis 2: Environmental pollution will bring a concealment effect.
3. Measurement model and empirical test
Measurement model and data selection
The empirical study in this paper tries to explore the degree of environmental problems caused by industrial structure adjustment, and then evaluate the impact of this process on income distribution. Using the global data from 1978 to 2021, the following measurement model is set as follows:
The model dependent variable was the income gap, which was mainly measured by the Gini coefficient. The core explanatory variables of the model are industrial structure, and represent the proportion of industrial added value and added value of service industry in the total added value respectively. To save space, only the identified equations are listed here, and the service industry is similar. The subscript and reference shall indicate the country (region) and year, respectively. In equation (a), it is the total effect of the proportion of the added value of the secondary industry on the dependent variable. In equation (b), it is the effect of the independent variable on the log of the carbon dioxide emissions per unit GDP of the masked variable. In equation (c), the effect of the independent variable on the dependent variable after controlling the independent variable, and the direct effect of the independent variable on the dependent variable after controlling the dependent variable. And are the national fixed effects of the three equations; and are the corresponding time fixed effects; and are the regression residuals. Where the coefficient is negative and has a significant increase, the cover effect exists; if significant, a partial cover effect, if not significant, a complete cover effect.
It is a group of control variables. This paper refers to the research of existing literature, and selects the following control variables from the perspectives of overall economic operation, labor market operation and labor force characteristics, urbanization level, foreign trade, financial development situation, and investment level. (1) Actual GDP per capita, Use it to measure the level of economic development; (2) Inflation, Measured by the consumer price Index (CPI); (3) Urbanization level, Measured by the ratio of the urban population to the total population; (4) The dependency ratio, Measured by the ratio of the working-age population to the working population; (5) Educational level, Measured by the ratio of the population who has completed high school education to those aged 15 and older; (6) Annual average population growth rate; (7) Foreign trade dependence, By the ratio of total imports and exports to GDP, Measure the degree of trade openness; (8) Investment rate, The investment rate of the country is measured as the ratio of social fixed asset investment to GDP. Referring to the traditional practices, in order to ensure the robustness of the data, two methods are adopted to supplement the missing data. The first method is to use the proximity method to fill the missing data, and then use the proximity method to make up for the data that has not been interpolated. The second method is to fill in the missing value. It aims to match the changing trend of data in various countries.
The statistics for the description of primary variables are shown in Table 1 below. Based on the descriptive statistics of the sample, the mean Gini coefficient in the sample countries was 38.2%, with the highest 65.8% (Malawi in 1997) and the lowest 20.7% (Jack in 1992), with a standard deviation of 9.1. And the 25% quantile is 32.9%, while the 75% quantile is 45.0%. According to the international general standard, the Gini coefficient exceeds 0.4, indicating that the Gini coefficient is higher, and about 43.2% of countries are higher than this standard. The average proportion of industrial output was 27.5%, with a standard deviation of 11.9. The average carbon dioxide emissions were 330 million tons, with a standard deviation of 1.5.
Statistics for description of primary variables
| Variable | Mean | Standard deviation | Least value | Crest value |
|---|---|---|---|---|
| Gink coefficient | 0.3959 | 0.8751 | 0.2070 | 0.6580 |
| Industrial proportion | 26.51% | 12.01% | 2.36% | 90.51% |
| Service sector proportion | 49.52% | 19.13% | 0.09% | 96.08% |
| Log of the CO2 emissions per unit of GDP | 9.2722 | 2.8212 | 2.3026 | 16.9066 |
| Annual GDP per capita (USD) | 8760.54 | 17075.57 | 12.78 | 234315.00 |
| Urbanization level | 50.72% | 25.34% | 2.07% | 97.23% |
| The dependency ratio | 70.78% | 20.49% | 16.17% | 91.64% |
| The ratio of total imports and exports to GDP | 50.77% | 46.29% | 2.72% | 87.78% |
| Annual population growth rate | 1.78% | 1.73% | −27.72% | 20.47% |
| The ratio of the population who have completed their high school education | 55.65% | 34.38% | 0.00% | 96.13% |
| The proportion of fixed assets investment in the whole society | 0.7558 | 0.7886 | 0.0054 | 5.4356 |
Basic regression model
The model regression results are shown in Table 2 below. Columns (1) and (2) are tests of main effects, where the first column is the model without a control variable and the second column is the model with a control variable. Both models showed that the increase in the industrial output ratio and the increase in services output ratio had a positive impact on the income gap, but the latter was not significant. This means that non-farm industries, representing higher technology, have a significant positive impact on widening the income gap, assuming 1. This is consistent with the study of Kuznets (2019). According to the current empirical data, the relationship between economic development level and income gap is still in the left half of the Kuznets curve. That is, the development of non-agricultural industries, especially the industrial development, is to widen the income gap (Stern, 2004).
Basal regression models
| Explained variable | (1) | (2) | (3) | (4) |
|---|---|---|---|---|
| Income divide (Gini) | Income divide (Gini) | Environmental pollution (log of CO2 emissions) | Income divide (Gini) | |
| The logarithm of the CO2 emissions | −0.5390*** | |||
| (0.0695) | ||||
| The proportion of industrial output | 0.0160* | 0.1706*** | 0.0352*** | 0.1885*** |
| (0.0095) | (0.0130) | (0.0031) | (0.0131) | |
| Proportion of service sector output | 0.0105 | 0.0026 | 0.0029* | 0.0042 |
| (0.0066) | (0.0083) | (0.0016) | (0.0063) | |
| The log of GDP per capita | 0.0002*** | 0.00001*** | 0.0002*** | |
| 0.0000 | 0.0000 | 0.0000 | ||
| Urbanization rate | 0.0027 | 0.0031* | 0.0084 | |
| (0.0070) | (0.0018) | (0.0070) | ||
| Inflation (CPI) | 0.0031*** | 0.0003* | 0.0033*** | |
| (0.0006) | (0.0001) | (0.0006) | ||
| Dependency ratio | 0.2576*** | −0.0412*** | 0.2367*** | |
| (0.0110) | (0.0025) | (0.0112) | ||
| Foreign trade dependence degree | −0.0346*** | −0.0187*** | −0.0452*** | |
| (0.0033) | (0.0007) | (0.0036) | ||
| Annual population growth rate | 0.4506*** | (0.0230) | 0.4682*** | |
| (0.1012) | (0.0200) | (0.1005) | ||
| Education level | 0.0307*** | (0.0015) | 0.0335*** | |
| (0.0073) | (0.0018) | (0.0072) | ||
| Investment in the fixed assets | −1.6794*** | −0.1662*** | −1.6261*** | |
| (0.1251) | (0.0317) | (0.1243) | ||
| State fixed effect | Control | Control | Control | Control |
| Time fixed effect | Control | Control | Control | Control |
| Observations | 10,416 | 3,844 | 4,650 | 3,844 |
| R2 | 0.0046 | 0.4603 | 0.4128 | 0.4686 |
Note(s): And are significant at 1%, 5% and 10%, respectively, with robust standard error in parentheses
At the same time, the model’s multicollinearity was tested simultaneously during the basic regression model fitting process. Among them, the correlation coefficient between each variable is not high, the variance inflation factor is less than 5, the highest level of education is 4.53, and the smallest variable is CPI, 1.07(As shown in Table S(1).
On this basis, continue to test the hiding effect of carbon dioxide emissions. That is to verify that a country's industrial development brings environmental problems, and environmental problems will bring a cover-up effect. After controlling carbon dioxide emissions under column (4), the impact coefficient of industrial output proportion on the income gap increases from 0.0352 shown in column (2) to 0.1885 shown in column (4). The product of column (3) the proportion of industrial output (0.0352) and column (4) the coefficient of the logarithm of carbon dioxide emissions (0.5390) is negative, which means. Since the coefficient of the proportion of industrial output in column (2) is significantly positive, environmental pollution has a concealment effect to a certain extent. At this point, hypothesis 2 is proved. This means that the industrial development leads to the deterioration of the environment, and when the environmental deterioration intensifies, the direct effect of the widening income gap caused by the industrial development will be underestimated. In addition, the healthy human capital investment itself has subjectivity. In reality, the laborer really establishes health consciousness, the utility optimization of health human capital allocation has time lag, the cover effect occurs, workers may not yet reached the optimal human capital investment. Combined with the constraints of social health care resources, the actual health human capital investment is insufficient. Therefore, the phenomenon of “pollution first, treatment later” is easy to explain.
At the same time, in the model regression results, all the estimated coefficients of the control variables in the regression are very significant, which are basically consistent with the existing literature. Specifically, the increase in per capita GDP, dependency ratio, population growth rate, education level, fixed asset investment rate and other indicators will widen the income gap, and all are significant at 1%. Increasing dependence on foreign trade will narrow the income gap, indicating that opening up will help a country's low-income groups get more opportunities to increase their income. From the R2 of the regression model, after the addition of control variables, the interpretation of the residents’ income gap was significantly improved, and the corresponding p-values of Sobel, Aroian and Goodman statistics were all less than 0.001.
Robustness and endogenicity test
First, consider the robustness of the model. In this paper, the robustness of the model from three aspects is verified. Firstly, adjusting the logarithm of CO2 emissions to the logarithm of PM2.5 emissions. PM2.5 was chosen as the explained variable of the robustness test model for the following three reasons. First, both are important indicators to measure the level of environmental pollution. Second, from the perspective of environmental theory, there is no necessary connection between the two in their formation, and the two variables are relatively independent. Third, although there is carbon dioxide greenhouse effect, in most people's subconscious, high concentrations of PM2.5 harm health. Inhaling a lot of solid particles can easily lead to diseases such as tracheitis, and serious cases can destroy human body function. This is also an important reference index for the weather forecast and people when deciding whether wear masks or travel. In other words, from the perspective of healthy human capital investment, people refer more to the PM2.5 index, but pay less attention to the carbon dioxide emission index, the latter suggests that PM2.5 may be more associated with the investment choice of healthy human capital. Second, the model form was adjusted and tested with the help of quantile regression. Although in the process of data processing, 1% tail is reduced, there is a big gap between countries and regions in industrial development and environmental protection. In order to avoid the influence of extreme value data on the model, this paper tests the robustness of the model by using non-parametric methods.
Secondly, the endogeneity of the model is considered. First, in order to verify whether the model has the endogenous problem caused by the explanatory variables and the explained variables being mutually causal, the industrial structure data of the lag period is used as the tool variable (IV) to estimate the model again. Second, to verify whether the model has endogeneity problems caused by the lack of important control variables, control variables are added here. On the one hand, among the many control variables selected above, there are few variables related to financial development. The extent of the development of a country's financial system is here measured by the ratio of domestic credit provided by the banking sector to GDP. On the other hand, due to the different systems of different countries, the controllers in the energy sector are significantly different, so the proportion of private investment in the energy sector is used as the control variable to fit the model again. The above tests yielded similar results and space constraints, and the results adjusted for the explained variables and IV estimates are listed in Table 3.
Results of the robustness test and the endogeneity test
| Robustness test | Endogenicity test: IV estimation | |||||
|---|---|---|---|---|---|---|
| Explained variable | (1) | (2) | (3) | (4) | (5) | (6) |
| Income divide (Gini) | Environmental pollution (log of PM2.5 displacement) | Income divide (Gini) | Income divide (Gini) | Environmental pollution (log of CO2 displacement) | Income divide (Gini) | |
| Log of the number of contaminants | −0.0603*** | −0.5390*** | ||||
| (0.0110) | (0.0695) | |||||
| The proportion of industrial output | 0.1655*** | 0.0486** | 0.1653*** | 0.1706*** | 0.0352*** | 0.1885*** |
| (0.0130) | (0.0194) | (0.0129) | (0.0130) | (0.0031) | (0.0131) | |
| Proportion of service sector output | −0.0026 | −0.0128 | −0.0023 | (0.0032) | (0.0034) | (0.0052) |
| (0.0083) | (0.0128) | (0.0082) | (0.0086) | (0.0021) | (0.0086) | |
| The log of GDP per capita | 0.0002*** | 0.0001*** | 0.0002*** | 0.0002*** | 0.0000*** | 0.0002*** |
| 0.0000 | 0.0000 | 0.0000 | 0.0000 | 0.0000 | 0.0000 | |
| Urbanization rate | 0.0027 | −0.0495*** | 0.0006 | (0.0002) | 0.0038** | 0.0057 |
| (0.0070) | (0.0111) | (0.0070) | (0.0075) | (0.0019) | (0.0074) | |
| Inflation (CPI) | 0.0032*** | 0.0003 | 0.0032*** | 0.0029*** | 0.0003** | 0.0031*** |
| (0.0006) | (0.0009) | (0.0006) | (0.0006) | (0.0002) | (0.0006) | |
| Dependency ratio | 0.2572*** | −0.1088*** | 0.2535*** | 0.2639*** | −0.0426*** | 0.2432*** |
| (0.0110) | (0.0155) | (0.0110) | (0.0114) | (0.0026) | (0.0117) | |
| Foreign trade dependence degree | −0.0346*** | −0.0077* | −0.0353*** | −0.0357*** | −0.0189*** | −0.0460*** |
| (0.0033) | (0.0042) | (0.0033) | (0.0035) | (0.0007) | (0.0038) | |
| Annual population growth rate | 0.4508*** | 1.0255*** | 0.4800*** | 0.4531*** | −0.0211 | 0.4727*** |
| (0.1013) | (0.1251) | (0.1011) | (0.1060) | (0.0208) | (0.1052) | |
| Education level | 0.0303*** | −0.0328*** | 0.0289*** | 0.0373*** | −0.0022 | 0.0397*** |
| (0.0073) | (0.0109) | (0.0073) | (0.0076) | (0.0019) | (0.0076) | |
| Investment in the fixed assets | −1.6791*** | −0.1195 | −1.6860*** | −1.6578*** | −0.1680*** | −1.6046*** |
| (0.1251) | (0.2007) | (0.1246) | (0.1288) | (0.0328) | (0.1280) | |
| National effect | Control | Control | Control | Control | Control | Control |
| Time effect | Control | Control | Control | Control | Control | Control |
| Observations | 3,844 | 4,588 | 3,844 | 3,562 | 4,344 | 3,562 |
| R2 | 0.4602 | 0.1822 | 0.4643 | 0.4612 | 0.4167 | 0.4692 |
Note(s): And are significant at 1%, 5% and 10%, respectively, with robust standard error in parentheses
Re-inspection of the formation mechanism of the income gap: based on the technical perspective
Above models are trying to find the cause of income increase, and gives the “pollution, after governance” behavior, but did not consider the green technology, namely its implied assumption is based on the use of new energy of green technology is the same between departments, obviously this is a relatively rough estimate, so consider the theoretical model of production cleanliness, namely the further exploration of green technology application is valuable. Some scholars believe that the development of some industries has the innate “clean production characteristics” (Zhiguang, 2021), which is the additional effect of technology. It is also an important aspect of the development force of green technology. For example, the development of digital technology has achieved energy intensive in some areas, paperless office and network transactions have reduced energy consumption, and its leading role comes from the market. Second, subjective technology research and development for environmental protection goals. As people pay attention to environmental issues, the governments’ measures to participate in governance are increasing. China is a typical example. After the 18th National Congress of the Communist Party of China, China has significantly reduced its carbon dioxide emissions thanks to the promotion of energy conservation and emission reduction technologies. Therefore, environmental issues are not only related to the choice preference of workers’ individual human capital investment, but also an activity based on market and non-market behavior related to technological development. In view of this, this paper takes green technology as the entry point to construct a model to further analyze the mechanism of the formation of income gap.
In fact, in the process of specific analysis, it is difficult to distinguish the additional effects of technology and the impact of technology research and development on environmental conditions and income distribution, but it is feasible to conduct a more macro study on the effects of the two aspects. In view of this, the ratio of renewable energy consumption to the total energy consumption is selected as the proxy variable and embedded with interaction terms, aiming to effectively measure the impact of green change based on different industries. The model can be constructed as follows:
Among them, the multiplication term is used to measure the interaction of new technology and industrial structure, and the ratio of renewable energy consumption to total energy consumption and its cross term are embedded in the service industry (.
Clearly, the sectors capable of cleaner production have higher technology, so all models show a significant positive correlation between the proportion of renewable energy consumption and the income gap, similar to the previous study, suggesting that the greater the technology gap between the high-tech and low-tech sectors, the wider the income gap. However, in the context of cleaner production factors, all the models in Table 4 below show that the interaction term coefficient between the industrial proportion and the proportion of new energy consumption is negative for both industry and service industry, indicating that the use of renewable energy within the industry partially weakens the widening income gap to some extent. In addition, all models show that the coefficient of the “proportion of industrial output to renewable energy consumption” is smaller than the proportion of renewable energy consumption in the interactive item, because the use of clean energy technology in services is more efficient than that in industry, and the effect of weakening the widening income gap is stronger.
Considers the mechanistic effect test of cleaner production
| Explained variable | Consider the interaction of cleaner production and its industrial output | Consider the interaction of cleaner production and service sector output | Consider the interaction of cleaner production and the two non-farm industries | ||||||
|---|---|---|---|---|---|---|---|---|---|
| (1) | (2) | (3) | (4) | (5) | (6) | (7) | (8) | (9) | |
| Income divide (Gini) | Environmental pollution (log of CO2 displacement) | Income divide (Gini) | Income divide (Gini) | Environmental pollution (log of CO2 displacement) | Income divide (Gini) | Income divide (Gini) | Environmental pollution (log of CO2 displacement) | Income divide (Gini) | |
| CO2 Loggarithm of emissions | −0.9232*** | −0.6754*** | −0.9495*** | ||||||
| (0.0709) | (0.0696) | (0.0710) | |||||||
| The proportion of industrial output | 0.0854*** | 0.0323*** | 0.0832*** | 0.1573*** | 0.0428*** | 0.1877*** | 0.0746*** | 0.0493*** | 0.0845*** |
| (0.0141) | (0.0031) | (0.0138) | (0.0138) | (0.0032) | (0.0139) | (0.0160) | (0.0031) | (0.0156) | |
| Proportion of service sector output | 0.0027 | −0.0048** | 0.0010 | 0.0111 | 0.0030 | 0.0129 | 0.0085 | 0.0035 | 0.0101 |
| (0.0081) | (0.0020) | (0.0079) | (0.0097) | (0.0024) | (0.0096) | (0.0096) | (0.0024) | (0.0093) | |
| The proportion of renewable energy consumption | 0.0414*** | 0.0035*** | 0.0435*** | 0.0413*** | 0.0022* | 0.0428*** | 0.0404*** | 0.0026** | 0.0422*** |
| (0.0046) | (0.0012) | (0.0045) | (0.0047) | (0.0012) | (0.0046) | (0.0046) | (0.0012) | (0.0045) | |
| Industrial output as a proportion of renewable energy | −1.73e−07*** | 2.26e−08*** | −2.44e−07*** | −1.67e−07*** | 2.47e−08*** | −2.34e−07*** | |||
| (1.59e−08) | (1.58e−09) | (1.65e−08) | (1.70e−08) | (1.56e−09) | (1.74e−08) | ||||
| Consumption The portion of renewable energy consumption | −0.0008*** | −0.0002*** | −0.0009*** | −0.0006** | −0.0002*** | −0.0006*** | |||
| (0.0002) | (0.0001) | (0.0002) | (0.0002) | (0.0001) | (0.0002) | ||||
| Controlled variable | Control | Control | Control | Control | Control | Control | Control | Control | Control |
| Observations | 3,844 | 4,526 | 3,844 | 3,844 | 4,526 | 3,844 | 3,844 | 4,526 | 3,844 |
| R2 | 0.4862 | 0.4605 | 0.5090 | 0.4740 | 0.4323 | 0.4867 | 0.4870 | 0.4622 | 0.5099 |
Note(s): And are significant at 1%, 5% and 10%, respectively, with robust standard error in parentheses
4. Conclusions and recommendations
This paper studies the relationship between industrial change, environmental protection and income gap through theoretical modeling and empirical analysis. In terms of theoretical analysis, this paper embedded the pollution production function Stokey (1998, 2021) into a two-stage production model creatively, and found that the investment level of healthy human capital increases with the increase of the difference and the increase of the income gap in the low technology sector, the income gap results in the increase of the income gap. In terms of empirical analysis, this paper proves the following conclusions with the help of empirical data, mediation effect model and IV model. First, the development of the non-agricultural production sector, which represents more advanced technology, has led to a widening income gap, and the environmental pollution problem caused by it has to some extent masked the trend of this income expansion. In other words, the widening income gap caused by environmental pollution may be confused with the widening income gap caused by industrial development, which leads to insufficient attention to environmental problems. At the present stage, this phenomenon is particularly pronounced in the industrial sector. Environmental constraints are an important variable to be considered in studying the relationship between industrial development and income gap. Second, both in the industry and in the service sector, the use of clean technology has weakened environmental pollution to a certain extent, thus playing a role in curbing the expansion of income, but this inhibitory effect is far less than the gap in technological progress between sectors. Third, compared with industry, the service industry is more efficient in using clean technology, more effective in reducing pollution, and thus has a higher effect on curbing the income gap.
Based on the above conclusions, this paper puts forward the following policy suggestions. First, the government plays a key role in narrowing the income gap and achieving common prosperity. The development of technology is an inevitable trend, and the emergence of new technology is often faster than the supply of human capital. In addition, the difference in investment in healthy human capital by the workers with different incomes under the environmental constraints, will inevitably lead to the widening of the income gap, including the application of green technology. In fact, after experiencing rapid economic growth, many countries have reduced the income gap by using inheritance tax and expanding the scope of social security. Of course, we need to pay attention to the coordination of efficiency and fairness in the process of designing policies. Second, we should strengthen policy guidance and continue to promote the use of clean energy. On the one hand, the inherent characteristics of the industry determine that some production projects are non-clean, so to achieve the goal of environmental governance, the government needs to use its visible hand, give full play to the effects of environmental protection, tax and other policies, and promote the use of clean energy. On the other hand, due to the differences in technological efficiency of different departments and the use efficiency of clean energy, government guidance contributes to the balanced development of green technology among various departments, so as not to cause the further widening of the income gap caused by technological imbalance.
Note
See China’s Agenda 21 —— White Paper on China’s Population, Environment and Development in the 21st Century.
The supplementary material for this article can be found online.
