Risk and self-optimization analysis of the carbon tax policy based on scenario simulation

The double dividend of the carbon tax policy has been a controversial topic. To comprehensively evaluate the benefits and risks brought by the carbon tax policy and contribute to China’s emission reduction goals, this paper establishes a carbon tax policy cycle simulation model based on China’s economic and energy data from 2010 to 2020 to explore the winner-curse phenomenons of the policy. To alleviate the winner’s curse of the carbon tax policy, this paper introduces a consumer behavior model to explore the optimization degree of loss aversion effect on the carbon tax policy. The research results show that the carbon tax policy has three kinds of winner’s curse phenomenons, namely, the improvement of environmental quality and the reduction of market capital, the decline of national carbon intensity and the increase of carbon intensity of three major industries, and the reuse of the tax revenue and the increase of economic loss. The loss aversion of consumers can alleviate the negative effect of the carbon tax policy and strengthen the positive effect. In addition, during the implementation of the carbon tax policy, the loss aversion effect can also reduce the polluted population by about 2%. Finally, based on the research results, the paper puts forward some feasible policy suggestions.


Introduction
By playing the role of price signal, the carbon pricing mechanism guides economic entities to reduce greenhouse gas (GHG) emissions, reduce environmental pollution behaviors, and promote the transformation of economic and social development to green development.It mainly includes the carbon tax and carbon trading.Compared with the carbon trading system, the carbon tax policy is easier to implement [1], and the long-term emission reduction effect is more significant [2].In addition, the carbon tax policy, as a fiscal tool, has a very high legally binding force.In recent years, more than 3,500 US economists, including 27 Nobel laureates, and 1,600 European economists have advocated a carbon tax [3].According to the 'State and Trends of Carbon Pricing 2022' report released by the World Bank, as of April 2022, a total of 68 carbon pricing mechanisms have been implemented in the world, covering about 23% of the total global GHG emissions, of which 37 are carbon tax systems.
A carbon tax is an environmental tax levied on the producers and users of fossil fuels that contain carbon [4].The core of the carbon tax policy is price control, through the price mechanism intervening and distorting economic entities to optimize production and business behavior, to achieve the purpose of carbon emission reduction (the carbon emissions mentioned in this article are all carbon dioxide emissions).And the revenue from a carbon tax could be used for public utilities or to offset the distorting effects of other tax systems on the market economy [5].Therefore, the carbon tax policy has the theoretical double dividend property: reducing environmental pollution and improving social welfare.
In terms of environmental benefits, the carbon tax policy has significant advantages in energy conservation and emission reduction.In Finland, for example, the carbon tax policy has a significant negative effect on carbon emissions per capita [6,7].In addition, while imposing a carbon tax, other supporting policy tools can further promote carbon reduction and improve environmental quality.Xiao et al [8] found that the interaction effect of the carbon tax policy and the new energy policy is greater than the comprehensive effect of a single strategy.Based on considering the synergies of other environmental policy tools, Hajek et al [4] found that the carbon tax policy in the energy sector is effective for environmental protection, and an increase of one euro per ton of the tax rate can reduce annual per capita emissions by 11.58 kg.In addition, when the carbon price is low, Li et al [9] found that the combined use of carbon cap-and-trade and carbon tax policies is more effective in reducing emissions.
However, in terms of economic benefits, some scholars believe that the economic dividend of the carbon tax policy does not exist.For example, Van Heerden et al [10] believe that the reinvestment of the carbon tax will greatly reduce the negative impact on gross domestic product (GDP), while subsidies to all industries in the form of production subsidies will minimize the negative impact on GDP, but there is a problem that the recycling of the revenue will reduce the degree of emission reduction.Gemechu et al [11] and Liang et al [12] agree that environmental and economic goals cannot be achieved at the same time.The carbon tax policy reduces the consumption of carbon-containing energy by increasing the cost, but the tax collected will be detached from the economic market, thus inhibiting the normal development of the economy.Although the recycling of the tax can make up for the negative economic impact of tax collection to a certain extent, Zhang and Lu [13] believe that due to the delay between the collection and recycling of the carbon tax, the government will always retain a part of the tax revenue.Therefore, the recycling of the carbon tax cannot completely offset the restraining effect of the carbon tax on the market economy.Wesseh Jr et al [14] found that the reuse of the carbon tax did not create welfare for low-income countries and regions, but made people in these regions worse off.
To mitigate the negative economic effects of the carbon tax policy, the most direct and effective way is to use additional public revenue to compensate those economic entities affected [11].However, this approach will further exacerbate the fiscal deficit of the government sector, which is not a good choice for some countries or regions with high fiscal deficits.Based on the internal driving mechanism of the carbon tax policy, the economic benefits of a carbon tax can be realized by adjusting the forms of the tax collection and reuse.In terms of carbon tax collection, Chen and Nie [15] found that the carbon tax in the production of products can improve social welfare, while the opposite is true in the consumption and redistribution links.In terms of the reuse of the carbon tax, Zhou et al [16] believe that the recycling of the tax revenue to households and enterprises will help realize the 'weak double dividend' effect of the carbon tax.In addition, from the perspective of policy complementarity, adopting a production tax reduction policy to offset the carbon tax can effectively reduce the economic cost of carbon tax collection [17].
Based on the above analysis, three conclusions can be drawn.First, the carbon tax policy has the effect of reducing GHG emissions.The essence of the carbon tax policy is to increase the cost of obtaining carboncontaining energy and reduce the actual consumption of such energy.Based on the price-demand relationship of a commodity, it can be seen that when the price of a commodity rises, the demand for that commodity will decline.Therefore, the carbon reduction effect of the carbon tax policy is real.Second, the carbon tax policy has the phenomenon of a winner's curse.The main reason for this paradox is that the carbon tax policy is to achieve carbon emission reduction through price signals, which leads to the imposition of a carbon tax to reduce the activity of the market economy to a certain extent, thus inhibiting the economic growth of a country or region.Due to the complexity of the social system, even if the government can efficiently collect and reinvest, there will always be a certain delay between collection and reinvestment, which will lead to a more serious financial crisis for low-income groups during this delay time.Third, the winner's curse of the carbon tax policy can be weakened.Through the detailed analyses of the social development process of the region where a carbon tax is imposed and every detail of the recycling from the collection of a carbon tax, it is possible to clarify the object of a carbon tax and optimize the process of policy implementation to minimize the negative impact.In addition, while imposing the carbon tax, introducing other complementary policies can also make the carbon tax infinitely close to the 'Pareto optimum'.
To realize the double dividend effect of the carbon tax policy, three questions need to be answered.First, how does the carbon tax policy affect the variables in the social system, and to what extent?Based on this problem, this paper explores the changes of variables in the social system under three scenarios: no carbon tax, carbon tax collection, and carbon tax reuse, to deeply analyze the internal mechanism of the winner's curse in the carbon tax policy.Second, from a micro perspective, can carbon tax policies reduce the carbon intensity of different industries and help achieve emission reduction targets?The imposition of a carbon tax will promote industrial restructuring, but it may also lead to the closure and layoffs of some enterprises and factories [18], especially for energy-intensive enterprises [19].When a carbon tax is imposed on energy-intensive enterprises with low added value, the economic damage of the carbon tax to this type of enterprise is greater than the reduction effect of energy consumption.Therefore, the paper divides the economic input into the primary, secondary, and tertiary industries, and explores the impact of the carbon tax policy on the carbon intensity of the three industries respectively.Third, based on not increasing the implementation cost of the carbon tax policy, is there any measure to enhance the double dividend effect of the carbon tax policy?The refinement of the implementation process of the carbon tax policy and the introduction of other auxiliary policies will increase the implementation cost of the carbon tax policy to a certain extent.From the perspective of behavioral economics, this paper introduces a consumer behavior model to explore how loss aversion enhances the double dividend effect of the carbon tax policy.
Based on the above three problems, this paper first establishes a carbon tax policy cycle model based on the loss aversion effect to analyze the internal cycle mechanism of the carbon tax policy.Secondly, through the collection and analysis of economic, energy, population, sulfur dioxide (SO 2 ), and other data, we determine the functional relationship between variables, and then establish a carbon tax policy cycle simulation flow diagram.Thirdly, multiple simulation scenarios are constructed by modifying the relevant parameters of variables in the cycle model.Finally, based on different policy scenarios, the paper analyzes the risks and benefits of the carbon tax policy in detail and gives the corresponding optimization path.
The remainder of this paper is organized as follows.Section 2 describes the application of the method and the data source.Section 3 introduces the construction process of the simulation flow diagram and the setting of the scenarios in detail, and uses the mental model to test the reliability of the simulation flow diagram on the basic scenario.Section 4 is the results and discussion.This part confirms that there are three kinds of winner's curse phenomenon in the carbon tax policy, and analyzes the reasons for each phenomenon.In addition, this part also explores the optimization degree of the loss aversion effect on the carbon tax policy.Section 5 is the conclusions and policy implications.

Application of loss aversion effect in the carbon tax policy
The loss aversion effect means that when people face the same amount of gain and loss, the negative emotion caused by loss is higher than the happiness caused by the same amount of gain [20].The realization of the double dividend of a carbon tax is to tax carbon dioxide (CO 2 ) emissions and then return the tax to taxpayers.This means that under the loss aversion effect, consumers will strengthen the collection of the carbon tax and weaken the subsidy of the tax revenue in their senses, that is, the collection of a carbon tax weakens consumers' consumption desire, while the same tax subsidy cannot completely make up for the decrease in consumption.The loss aversion effect provides a theoretical basis for analyzing consumer behavioral motivation under the carbon tax policy.
After the implementation of the carbon tax policy, the tax revenue will cycle back and forth between the government and the economic market.Figure 1 shows the cycle of the carbon tax based on the loss aversion effect.A more detailed introduction is as follows.
Step 1: Determine the form of the carbon tax policy.
From the perspective of tax distribution, the carbon tax policy can be divided into two types.The first is that the tax revenue brought by the carbon tax policy is used to make up the fiscal deficit.The second is the use of carbon tax revenues to compensate those affected by the carbon tax policy.Compared with the first form, the second can reduce the impact of the carbon tax policy on regional economic development.In addition, the loss aversion effect is caused by consumers' psychological perception based on loss and gain.Therefore, this paper adopts the second form as the basis of model construction.
Step 2: Consumers' perception of economic loss caused by the carbon tax.
After the carbon tax is imposed, some output will be taken out of the economic market.The amount of the part depends on the amount of CO 2 emitted from fossil fuel consumption and the carbon tax rate.Because the carbon content of different energies is different, the carbon emissions of different types of energy need to be calculated separately.
Where EL j represents the economic loss of industry j. E ij represents the total consumption of energy i in industry j, and C i represents the carbon emission coefficient of energy i. i represents coal, crude oil, and natural gas, respectively.j is for primary, secondary, and tertiary industries.TR represents the tax rate of the carbon tax policy.Step 3: Reuse of the carbon tax revenue under the delayed effect.The realization of the double dividend effect of the carbon tax policy is based on the collection and reinvestment of the carbon tax.Based on the research of Fan and Wang [21], the paper sets the delay time for the collection and reuse of the carbon tax to 1 year.In addition, to maximize the economic dividend of the carbon tax policy, the paper assumes that the amount of the carbon tax reuse is based on the total tax revenue retained by the government in the current year, rather than the carbon tax revenue of the previous year.
Where TE represents the amount of the carbon tax reused.And TEL represents the amount of the tax revenue retained by the government, which is determined by economic losses and the amount of the carbon tax reuse.DELAY1I is the delay function, and the initial value of the carbon tax reuse is 0.
Step 4: The impact of the loss aversion effect on energy consumption after imposing the carbon tax.The higher the degree of loss aversion, the lower the desire for energy consumption, that is, the greater the loss aversion value (L), the less energy consumption (E).Energy consumption after implementing the carbon tax is the product of the sum of GDP after implementing the carbon tax (GDP EL ) and the amount of the recycling of the carbon tax under the loss aversion, and unit energy consumption (UC).
Step 5: Assessing the spillover effect of the carbon tax policy.
In addition to carbon dioxide, other concomitant pollutants will be produced in the process of energy consumption.The greater the energy consumption, the more concomitant pollutants are produced.To prevent repeated assessment of multiple concomitant pollutants on the same object damage, SO 2 was selected as the concomitant pollutant to assess the health effects.
( ) Where W represents the actual pollution level of SO 2 , C i ¢ represents the SO 2 emission coefficient of energy i, and s represents the space size of SO 2 diffusion.N represents the total population affected by SO 2 , P represents the total population of the region, R represents the actual mortality rate, l represents the coefficient of the model (0.00075) [23], and W 0 is the threshold level that does not affect health.
Figure 1 constructs a carbon tax policy cycle model based on the loss aversion effect.This model analyzes the internal circulation mechanism of the carbon tax and is more in line with reality.Under the loss-aversion effect, the public will significantly underestimate the amount of revenue reuse [3].At the same time, this model also answers why carbon tax policy can achieve carbon emission reduction when the total amount of the market economy remains unchanged.

Data
Based on the three problems explored in this paper, the basic data in the simulation flow chart includes GDP, population, total energy, the proportion of consumption of each energy, SO 2 emissions, and so on.The data comes from the China Statistical Yearbook from 2011 to 2021.The data on GDP and energy consumption are broken down by industry.The types of energy are divided into coal, crude oil, natural gas, and renewable energy.In addition, the respective unit energy consumption of the three major industries is calculated.The amounts of CO 2 and SO 2 produced by different types of energy are calculated from the Industrial Pollutants Production and Discharge Coefficient Manual [22].Since both the carbon tax policy and carbon trading policy belong to the carbon pricing mechanism, this paper sets the tax rate based on the current carbon trading price in China.Since 2022, the daily closing price of the national carbon market has been between 56 and 62 CNY/ton.Therefore, the basic tax rate of the carbon tax is set at 60 CNY/ton.In addition, the formula for people affected by SO 2 and the proportion of energy consumption is based on other studies [13,23].See appendix A for a more detailed description.

Construction and scenario setting of simulation flow diagram
Based on the loss aversion and causal relationship between variables, this paper constructs a simulation flow diagram from the four aspects of the economy, energy, environment, and health, as shown in figure 2. The simulation diagram of figure 2 can be divided into four sub-modules, namely primary industry, secondary industry, tertiary industry, and others.The construction of the first three sub-modules is based on the model of section 2.1, and the construction of the last one is based on the relevant literature and calculation rules.In each sub-module, the equations and descriptions of the variables are detailed in appendix A. Figure 2 contains four alterable variables, namely loss aversion index (L), loss aversion adjustment factor (S2), tax rate (TR), and tax adjustment coefficient (S1).By adjusting the loss aversion index of consumers and the carbon tax rate and their adjustment coefficients, this paper sets scenarios 1-6 to explore the winner's curse phenomenon and reveal the optimization path of the carbon tax policy.The Settings of scenarios are shown in table 1.
As a basic scenario, scenario 1 is used to compare with scenarios 2 and 3 to reveal the impact mechanism and degree of the carbon tax policy on the social system.In addition, based on these three scenarios, the paper also

Scenario
Variable value Description In this scenario, no carbon tax is imposed.This scenario is used to be compared with other scenarios to show the effect of the carbon tax policy on the social system. 2 L = 0, TR = 60, S1 = 0, S2 = 0 The scenario represents the imposition of a carbon tax, but the tax revenue is not reused.This scenario is mainly for comparison with scenario 3 to highlight the shock effect of scenario 3 on social development.
The scenario represents the imposition of a carbon tax, and the tax revenue is reused.
This scenario indicates the extent to which the loss aversion effect strengthens the double dividend effect of the carbon tax policy and weakens the winner's curse when the consumer's loss aversion degree is low.
This scenario indicates the extent to which the loss aversion effect strengthens the double dividend effect of the carbon tax policy and weakens the winner's curse when the consumer's loss aversion degree is medium.6 L = 0.9, TR = 60, S1 = 1, S2 = 1 This scenario indicates the extent to which the loss aversion effect strengthens the double dividend effect of the carbon tax policy and weakens the winner's curse when the consumer's loss aversion degree is high.
explores the root of the winner's curse of the carbon tax policy.Scenarios 4-6 mainly analyze the optimization degree of the loss aversion effect on the carbon tax policy.

Model testing
To ensure the accuracy and robustness of the model, this paper uses a mental model test to verify the four core state variables in the simulation flow diagram.The mental model test analyzes whether the behavior simulated by the system can replicate the data changes in the real-world time series.The test contains two types.The first is to test the coincidence of absolute data, that is, the error value between the simulated data and the real data is small.The second is to test whether the simulation trend of the variable is consistent with the real one.To this end, the mean relative error (MRE) and the synthetic degree of incidence (SDI) are used to evaluate the simulation data of the total population and the GDP of the primary, secondary, and tertiary industries from 2010 to 2020 in the simulation model.MRE is the ratio of the absolute error caused by the measurement to the true value, usually expressed as a percentage.SDI reflects the rate of change of two sequence data relative to the starting point.
The evaluation results of MRE and SDI are shown in Table 2. table 2 shows that the MRE of the four variables is less than 10%, and SDI is greater than 0.9.According to the evaluation criteria of MRE [24] and SDI [25], the error evaluation results of the four variables are good, and their trend evaluation grade is level 1. Combining these test results, it is concluded that the simulation flow diagram has good numerical accuracy and can reflect the changing trend of variables.Therefore, a simulated flow diagram can be used.

Results and discussion
4.1.The Winner's curse 4.1.1.Impact of the carbon tax policy on social system To explore the impact of the carbon tax policy on the economy, environment, and health, GDP, CO 2 emissions, and the population affected by SO 2 are respectively simulated in scenarios 1-3, and the simulation results are compared, as shown in figure 3. Compared to figures 3(b), and (a) shows that when a carbon tax is imposed but not reused, a carbon tax policy can reduce more CO 2 emissions, but also lead to a greater loss of GDP.This shows that although the imposition of a carbon tax can slow down GHG emissions, it will also have corresponding negative effects on the economic market.figure 3 (b) reveals a cyclical phenomenon, that is, the collection and reuse of the carbon tax will lead to periodic fluctuations in the social system, and this cycle is 7 years.The  environmental and health benefits of figure 3 (b) are about ten times smaller than those of figure 3 (a), so figure 3 confirms that the recycling of the tax revenue reduces the effectiveness of emissions reductions [10].It is worth noting that regardless of whether the tax revenue is reused or not, figure 3 shows that the positive benefit of the carbon tax policy on the environment is higher than the negative benefit on the economy, except for the first five years of the simulation period, the health benefit brought by the carbon tax policy is much higher than the negative impact on the economy.In addition, as time goes on, the negative benefit of the carbon tax policy on the economy will gradually weaken, and the difference between the positive and negative benefits of the carbon tax policy will gradually increase.The increasing difference in the positive and negative benefits is mainly due to the progress of science and technology and the optimization of energy structure (figure B1). Figure 3 reveals the first winner-curse phenomenon of the carbon tax policy, namely that the carbon tax policy has environmental and health benefits, but increases the risk to market economy.This result is consistent with the conclusions of previous studies [11][12][13].For China, the imposition of a carbon tax would worsen the economic situation, although, in the long run, the negative economic impact would diminish relative to environmental improvement.Although the reuse of the tax revenue can mitigate the economic risks brought by a carbon tax to the greatest extent, it will also weaken the emission reduction effectiveness of the policy.Therefore, it is almost impossible to reduce environmental pollution without damaging the market economy, unless additional public revenue is used to make up for the economic losses caused by the carbon tax [11].But using other revenue to make up for the economic loss made by imposing a carbon tax would undoubtedly exacerbate the government's fiscal deficit.Take China's economic situation in 2022 as an example, China's fiscal deficit has reached 3.37 trillion yuan, if other revenues are used to make up the economic loss, the fiscal deficit will increase by more than one-third.In addition, even if the government introduces other revenues into the carbon tax system, the emission reduction effect of the carbon tax policy will face the risk of halving again.This winner-curses phenomenon is one reason why there is currently no carbon tax in China.
From the perspective of the stability of social development, the collection and reuse of the carbon tax will aggravate social instability.Economic fluctuation will lead to the instability of residents' income, which will produce unpredictable risks.Reducing energy consumption per unit of output and improving the energy consumption structure can not only reduce the negative effects of the carbon tax policy (figure B1), but also mitigate the economic fluctuations caused by the carbon tax policy (figure 3 (b)).

Simpson's paradox
To explore the impact of the carbon tax policy on different industries in more detail, this paper analyzes the impact of the carbon tax policy on the carbon intensity of different industries and compares it with the national carbon intensity.The results are shown in table 3. Table 3 shows that with the collection of the carbon tax, the national carbon intensity will start to decrease in the second year of the implementation of the carbon tax policy, but the reuse of the tax revenue will reduce the actual utility of the carbon tax policy in the fourth year.Overall, the carbon tax policy help to reduce total carbon intensity.However, when assessing the impact of the carbon tax policy on the carbon intensity of the three industries separately, an interesting phenomenon is that the imposition of a carbon tax would lead to an increase in the carbon intensity of the three industries starting from the third year, and the reuse of the tax revenue would further increase the carbon intensity of the three industries in the sixth and seventh years respectively.Therefore, Simpson's paradox exists in the assessment of the carbon intensity of the three major industries and the total, that is, the carbon tax policy leads to the increase of the carbon intensity of the three major industries, while the total carbon intensity decreases.One of the reasons for Simpson's paradox is that there are significant differences between the proportion values of the items in the statistical object.Figure B2 shows that the carbon intensity of the secondary industry is 4-10 times greater than that of the primary and tertiary industries.Therefore, the reason for Simpson's paradox caused by the carbon tax policy is that the carbon intensity of the secondary industry is significantly higher than that of the primary and tertiary industries.
The second winner's curse of a carbon tax is that while it reduces the total carbon intensity, it increases the carbon intensity of the three major industries.At the end of 2021, China's Ministry of Industry and Information Technology proposed to reduce CO 2 emissions per unit of industrial-added value by 18% by 2025.Industry belongs to the secondary industry.table 3 shows that in 2025, compared with the carbon intensity of the secondary industry in scenario 1, scenario 2 increases by 1 ton/billion CNY, while scenario 3 increases by 1.1 tons/billion CNY.Therefore, the implementation of the carbon tax policy will erode the carbon intensity targets to a certain extent.According to previous reports [26], China's industrial sector currently accounts for 65% of the country's CO 2 emissions.Therefore, the carbon tax policy will be more corrosive to the carbon intensity targets of the industry.In addition, Zhang and Lu [13] also found that a carbon tax would erode China's carbon peak goals.Simpson's paradox shows that although the carbon tax policy has the effect of energy saving and emission reduction, it is not always beneficial to some strategic goals of the country.B1 (a)), there are significant differences in the impact of the carbon tax policy on the three industries.figure 4 shows the simulation results of the carbon tax policy on industrial upgrading and economic losses of various industries.Figure 4 (a) shows that the collection of the carbon tax can significantly promote the upgrading of industrial structures, but the reuse of the tax revenue will slow down the process of industrial structure upgrading.The main reason why the collection of the carbon tax promotes the upgrading of industrial structure is that the carbon intensity of the secondary industry (table 3) and the proportion of economic output (figure B3) are too high, resulting in a more significant impact of the carbon tax policy on the secondary industry.In addition, the reuse of carbon tax revenue will lead to the increase of capital in the secondary industry market, which will inhibit the adjustment of the industrial structure by the carbon tax policy to a certain extent.Figures 4(b)-4(c) show that the economic loss caused by the carbon tax policy is the largest in the secondary industry, followed by the tertiary industry, and the least affected is the primary industry.The reason for the obvious difference in economic losses among the three industries is due to the unit energy consumption (figure B1 (a)) and the proportion of economic output (figure B3).Since the secondary industry has the largest unit energy consumption and the proportion of economic output has advantages, the economic loss of the secondary industry is the largest.Since the secondary industry has the largest unit energy consumption and the proportion of economic output is larger, the economic loss of the secondary industry is the largest.Compared with the economic output of the secondary industry, although the proportion of the tertiary industry in 2035 will reach 64.1%, due to its small unit energy consumption, the impact of the carbon tax policy on the economic loss of the tertiary industry is not as great as that of the secondary industry.Since the unit energy consumption of the primary industry and the proportion of economic output are small, the carbon tax policy has the least impact on the economic loss of the primary industry.An interesting phenomenon is that the reuse of the carbon tax revenue will increase the economic losses of the three industries.In essence, the reuse of the carbon tax revenue is to reduce the negative effect of carbon tax policy on the economy but increases economic losses.This is mainly because although the collection and reuse of the carbon tax to a certain extent ensure that the total market capital will not be reduced too much, the reuse of the carbon tax increases the monetary turnover rate, and further increases the total social wealth, and then lead to the increase of economic losses in numerical terms.The third winner's curse of the carbon tax policy is that although the reuse of carbon tax revenue can reduce the outflow of market capital, it will also slow down the upgrading of industrial structure and increase the economic losses of the three major industries.From the perspective of economic losses, the carbon tax policy can improve public welfare, which is a theoretical 'scam'.The process of collecting and reusing the carbon tax does not, in essence, produce substantial value production.In this process, a series of management behaviors such as GHG emission measurement, data management, and personnel organization also has some costs, which will also consume part of the carbon tax revenue.Under the dual effects of rising prices caused by the carbon tax collection and disguised increase of total social wealth caused by the carbon tax recycling, the recycling of carbon tax will aggravate economic losses in the market.The most important manifestation of industrial upgrading is the transfer of economic output from the secondary industry to the tertiary industry.If the reuse of the tax revenue in the secondary industry is tilted to the tertiary industry, although it can accelerate the upgrading of industrial structure, it will make the carbon tax policy far away from the Pareto optimum, and have a huge negative impact on the secondary industry.Therefore, returning the tax revenue to the secondary industry is a necessary option.

Optimization of the carbon tax policy
To reduce the negative effects of the winner's curse and realize the Pareto optimality to the maximum extent, this paper, starting from the internal mechanism of the carbon tax, explores the degree of consumer loss aversion to the optimization of carbon tax policy from the perspective of double dividends.Here, scenarios 1 and 3 are used as base scenarios to evaluate the actual effects of the carbon tax policy on the economy, environment, and health under different loss-aversion scenarios.Since the collection and reuse of the carbon tax will lead to periodic fluctuations of variables in the social system, the trough points of each cycle (2013, 2019, 2025, and 2031) are selected as observation years to more clearly analyze the impact of the carbon tax policy.

Economy
Table 3 shows that there are significant differences in the impact of the carbon tax policy on the three major industries and the whole.To this end, the economic losses of each industry and the whole are simulated under scenarios 4-6, and compared with scenario 3. The results are shown in table 4. Horizontal and vertical comparisons of each variable in table 4 reveal two phenomena.First, in the vertical comparison, with the increase of the loss aversion index, the economic losses of the three industries and the whole will tend to decrease.Taking the peak point of the first cycle (2013) as an example, for every 0.3 increase in loss aversion, the three industries and the total economic loss will decrease by 0.11%, 0.81%, 0.17%, and 0.68%, respectively.This is mainly because the higher the loss aversion of consumers, the less energy consumption, thus reducing the economic loss.Second, in the horizontal comparison, with time, under the same loss aversion degree, the reduction proportion of economic loss will gradually decrease.Taking the total economic loss as an example, when the loss aversion level is 0.3 (Scenario 4), the economic loss can be reduced by 0.68% in 2013, but only 0.19% in 2031.In addition, in the same variable, although loss aversion increases regularly, the economic loss reduction of the trough point shows a decreasing change.Taking the economic loss of the secondary industry as an example, with the increase of loss aversion, the economic loss reduction rates at the four-cycle points are 0.81%, 0.54%, 0.37%, and 0.23%, respectively.The reason for this phenomenon is mainly due to the rationalization of the resource structure and the reduction of unit energy consumption (figure B1).To reduce the impact of the carbon tax policy on the development of the market economy, the reuse of tax revenue is an effective method.In addition, increasing consumers' loss aversion can also mitigate economic losses.It should be noted that the use of consumers' loss aversion can not fundamentally solve the negative economic effects caused by the carbon tax policy, but can only alleviate them to a certain extent.To maintain the reduction of economic losses under the condition that the carbon tax rate remains unchanged, increasing the loss aversion of consumers is a feasible method.

Environment
Table 4 shows that the loss aversion effect can mitigate the economic losses caused by the carbon tax policy.Does the loss aversion effect reduce the environmental benefits of the carbon tax policy due to the first winner-curse phenomenon in section 4.1.1?When the economic loss is reduced, the carbon intensity value will increase if the CO 2 emissions increase at a large rate.Therefore, selecting carbon intensity as an indicator to assess the environment further reveals the optimization of the loss aversion effect on the carbon tax policy, and the results are shown in table 5. table 5 shows that with the increase of loss aversion, the carbon intensity shows a decreasing phenomenon.Combined with the results in table 4, it can be found that the loss aversion effect can not only reduce the economic disutility of the carbon tax policy, but also enhance the positive environmental effect of the carbon tax policy.In addition, as compared with table 3, Simpson's paradox does not exist in the collection and reuse of the carbon tax when loss aversion is taken into account.When the loss aversion degree is 0.3, compared with scenario 1, the carbon intensity of the three industries decreases by 3.42, 190.9, and 7.92 tons/million CNY respectively at the peak point of the first cycle.And with the reduction of carbon intensity in the three major industries, the total carbon intensity has also been reduced by 201.2 tons/billion yuan.
The loss aversion effect can alleviate the negative effects and strengthen the positive effects of the carbon tax policy.Therefore, the loss aversion effect can effectively solve or alleviate the three kinds of winner's curse phenomenon of the carbon tax policy.As for the first and third winner's curse phenomena of the carbon tax policy, the loss aversion effect reduces energy consumption through different perceptions of consumers' psychological loss and income, thus promoting the reduction of economic losses and reducing the drain of market capital.However, it should be noted that this may inhibit the activity of economic markets to some extent.In addition, for the second phenomenon of the winner's curse of the carbon tax policy, the loss aversion effect can fundamentally eliminate Simpson's paradox.Therefore, in reality, using only real-world data, the carbon tax policy does not lead to Simpson's paradox.The analysis results in section 4.1.2only exist in the theory of this paper.Although Simpson's paradox of the carbon tax policy is a less important phenomenon, this paradox cannot be ignored by policymakers.If policymakers focus on increasing the perception of the reuse of the carbon tax revenue, Simpson's paradox will become a reality.

Health
Tables 4 and 5 show that the loss aversion effect can promote the realization of the double dividend of the carbon tax policy to a certain extent.To reveal the health benefits attached to the carbon tax policy under the loss aversion effect, this paper simulated SO 2 -affected populations in scenarios 4-6 respectively, and compared them with scenario 1.The results are shown in table 6.Table 6 shows that with the increase in loss aversion, the number of people affected by SO 2 will gradually decrease.This is because the implementation of the carbon tax policy will reduce the consumption of fossil fuels, and therefore the emissions of pollutants associated with CO 2 will also be reduced.China's SO 2 emissions showed a precipitous decline in 2016, so table 6 shows that except for 2013, the ratio in the other three years shows an upward trend, but the increase is limited.It can also be found from table 6 that under the effect of loss aversion, the implementation of a carbon tax policy can roughly reduce the polluted population by about 2%.
Based on the analysis results in table 6, it can be found that the loss aversion effect can further strengthen the spillover effect of the carbon tax policy on public health.As the level of consumer loss aversion increases, the public health benefits of the carbon tax policy will gradually increase.Therefore, when analyzing the spillover effect of the carbon tax policy, analyzing the actual utility of the carbon tax policy from multiple dimensions can better amplify the advantages of the carbon tax policy and properly avoid its disadvantages.

Conclusions and policy implications
To comprehensively evaluate the benefits and risks brought by the implementation of the carbon tax policy, this paper constructs a cycle simulation flow diagram of the carbon tax policy based on real-world data and consumer behavior models.Based on the simulation flow diagram, different simulation scenarios are set up by adjusting the coefficients of the variables.To ensure the accuracy and robustness of the simulation, the numerical errors and trend errors of the four core variables in the simulation flow diagram are verified.Finally, through the simulation results of the main variables under different scenarios, the winner's curse phenomenons of the carbon tax policy and the optimization degree of loss aversion effect on the carbon tax policy are explored.The main results of this paper are as follows.
First, the carbon tax policy has an inhibitory effect on the development of the market economy, but the selfoptimization attribute of the social system can reduce the inhibitory effect.The carbon tax policy takes the price as a signal to reduce the consumption of fossil energy and reduce environmental pollution, which leads to the loss of the market economy.To reduce the negative effects of the carbon tax policy on economic markets, tax revenue should be used to compensate those groups affected.Therefore the carbon tax policy theoretically has a double dividend effect.However, a certain amount of economic losses will be generated in the process of collection and recycling, which makes the capital collected higher than that of recycling.The loss aversion effect will offset the cost generated in the implementation process of the carbon tax policy, thus reducing the negative effects of the carbon tax policy on the economy and strengthening the positive benefits on the environment.
Second, combined with other interventions, the environmental improvement function of the carbon tax policy can be promoted.The results of this paper show that the reduction of unit energy consumption, the optimization of energy structure, and the loss aversion effect are conducive to the implementation of the carbon tax policy.Among them, the loss aversion effect can further enhance the efficacy of emission reduction based on the emission reduction of the carbon tax policy.The reduction of unit energy consumption and the optimization of energy structure will reduce the negative effect of the carbon tax policy on the economic market, and then help the implementation of China's carbon tax policy.In combination with other environmental measures, the economic and environmental dividends of the carbon tax policy are not impossible to achieve a win-win situ ation.
Third, the environmental spillover effect of the carbon tax policy has significant positive benefits.The improvement of environmental conditions has a great impact on the public health.Although the carbon tax policy is based on CO 2 as the object of taxation, it is aimed at carbon-containing energy.The carbon tax policy reduces the consumption of carbon-containing energy but also reduces the emission of other pollution from such energy.With the improvement of scientific and technological levels, the health benefits of the carbon tax policy will also be further improved.In addition, some measures to enhance the environmental benefits of the carbon tax policy can also improve public health.
Based on the internal mechanism of the carbon tax policy on the economic, environmental, and health, the following recommendations should be taken seriously by policymakers.
First, policymakers should comprehensively evaluate the advantages and disadvantages of the carbon tax policy from the perspective of system science.In the research results of the paper, it is found that theoretically speaking, there are three kinds of winner's curse phenomenon in the carbon tax policy.Compared with environmental dividends and social welfare, the external spillover effect of the carbon tax policy is more important to the sustainable development of society.Identifying the defects in the external spillover effect is particularly important for the accuracy of decision-making and the scientificity of policy.In addition, by enhancing the advantages of the spillover effect, the development of social health can be further promoted.Second, while using the carbon tax policy to improve the living environment, other auxiliary policies should be adopted to supplement and strengthen the defects and advantages of the carbon tax policy.The research results of this paper show that although the carbon tax policy has the advantage of energy conservation and emission reduction, it also has some negative effects, such as increasing market economic losses.Therefore, using the tax revenue to make up for the economic losses of the affected population is a necessary measure.In addition, other measures or policies can be used to further reduce the negative economic effects of the carbon tax policy and strengthen the positive environmental benefits.For China, if it wants to introduce a carbon tax at present, it must be combined with other auxiliary means, otherwise, the negative benefits of the carbon tax will exceed the positive.
Third, incorporate more micro-assessment indicators and formulate reasonable emission reduction targets.Although the paper carefully analyzes the economic, environmental, and health impacts of the carbon tax policy, the actual effectiveness of the carbon tax policy for smaller industries is still not clear.For small and energyintensive enterprises, the implementation of a carbon tax policy will undoubtedly increase the operating costs of enterprises.Therefore, it is necessary to accurately assess the micro impact of the carbon tax policy, and the authors will continue to explore the deep impact of the carbon tax policy in future work.
Fourth, strengthen the public's perception of loss and weaken the perception of the benefit of the carbon tax policy.The study found that the loss aversion effect can reduce market capital losses by about 1% and improve public health by about 2%.Therefore, the loss aversion effect is beneficial to carbon tax policy.Government departments can strengthen public loss aversion through media publicity, to enhance the actual utility of the carbon tax policy and help realize the double dividend of the carbon tax policy without increasing the cost.4) describe in detail the main variables in the primary industry.The relevant variables for the secondary and tertiary industries are formally identical to them.Equations (4)-( 6) are derived from the research of Zhang and Lu [13].Equation (7) is the calculation formula for SO 2 emissions.Equations (8-10) respectively draw on the research of Wu et al [23], Xin-gang and Jin [27], and Lin et al [28].

Appendix B
Figure B1 shows that over time, the unit energy consumption of the three industries will gradually decrease, and the proportion of energy consumption will gradually be optimized.It should be noted that although the unit energy consumption of the three major industries is gradually decreasing, the unit energy consumption of the secondary industry is still higher than that of the primary and tertiary industries.Figure B1 (b) shows more significant changes in coal and clean energy consumption ratios and relatively small changes in oil and gas consumption ratios.Therefore, reducing the unit energy consumption of the secondary industry and the proportion of coal consumption, and increasing the proportion of clean energy are more conducive to slowing down the negative economic effects of the carbon tax policy.
The figure is a visual graph based on the carbon intensity ratios of the three major industries in table 3, which mainly illustrates the gap between the carbon intensity of the primary and tertiary industries and that of the secondary industry.This figure explains the Simpson paradox of the carbon tax policy.

Figure 1 .
Figure 1.Cycle flow chart of the carbon tax policy based on the loss aversion effect.

Figure 2 .
Figure 2. Cycle simulation flow diagram of the carbon tax policy based on the loss aversion effect.

Figure 3 .
Figure 3.The ratio of the difference in GDP, CO 2 emissions, and the population affected by SO 2 under different scenarios.

Figure 4 .
Figure 4. Simulation results of industrial upgrading and economic loss in scenarios 1-3.

( 10 )
Atmospheric altitude = 1076, Unit: meter Since the equation forms of some variables in the simulation flow diagram are similar, ten representative variables are selected for description in this paper.Equations (1)-(

Figure B1 .
Figure B1.Unit consumption and of energy consumption.

Table 2 .
Error and degree of incidence between simulated and actual values of main state variables.
4.1.3.Impact of the carbon tax policy on three major industries Due to the different unit energy consumption of each industry (figure

Table 6 .
The ratio of the population affected by SO 2 in scenarios 4-6 to scenario 1 (%).