In order to test the environmental impact of developed countries’ climate assistance to developing countries, this paper selects total greenhouse gas emissions of carbon dioxide equivalent and energy consumption emissions of carbon dioxide equivalent of countries as a climate environmental indicator from 1990 to 2017. It also analyzes the climate-related development finance assistance behavior of developed countries to developing countries since 2000, which was published by the OECD as a quasi-natural experiment, and utilizes the DID (differences-in-differences) method to identify the causal effect. From the perspective of recipient countries, this paper regards developing countries (regions) receiving aid as intervention groups, and other developing countries (regions) not receiving aid as control groups. By comparing the samples in the intervention groups and control groups, we can evaluate the net effect of the “South-North cooperation” aid behavior. The specific model is set as follows: C a r b o n i t = α 0 + α 1 C R F i t + α 2 ∑ C o n t r o l i t + μ i + δ t + ε i t (1)

In the formula above, C a r b o n i t refers to total greenhouse gas emissions thousand tons of carbon dioxide equivalent ( C O 2 ) and energy consumption emissions thousand tons of carbon dioxide equivalent ( e n e r g y C O 2 ) in each country and region. C R F i t indicates whether country i is a recipient target in year t. This paper defines it by setting a dummy variable. If a country has received climate-related development finance, it is assigned a value of 1, which is regarded as an treatment group. Otherwise it is 0, and is regarded as a non-treatment group (control group). This definition could distinguish recipient countries from non-recipient countries, and lay a foundation for setting treatment group and non-treatment group, so as to meet the application conditions of DID method. The treatment group is the object affected by the policy, and the non-treatment group is the object not affected by the policy. Through the DID method, that is, after two differences between the treatment group and the non-treatment group, the “net effect” of policy implementation could be effectively identified, so that the policy could be evaluated more scientifically. Control refers to controlling other factors that may affect carbon emissions, while μ i refers to the national fixed effect,   δ t refers to the time fixed effect ,  and   ε i t refers to the random factors involved. α 1 is our major concern here.

In order to further analyze the internal reasons for the positive impact of climate assistance on carbon emissions reduction, this paper discusses the transmission mechanism by analyzing the relationship between climate assistance and energy structure change from the perspective of energy structure. The specific model is constructed as follows: e n e r g y i t = β 0 + β 1 C R F i t + β 2 ∑ C o n t r o l i t + μ i + δ t + ε i t (2) c o a l p o w e r i t = γ 0 + γ 1 r e n e w a b l e s i t + γ 2 ∑ C o n t r o l i t + μ i + δ t + ε i t (3) c o a l p o w e r i t = θ 0 + θ 1 r e n e w a b l e s i t × C R F i t + θ 2 e n e r g y i t + θ 3 ∑ C o n t r o l i t + μ i + δ t + ε i t (4)

Among them, formula (2) represents the impact of climate assistance on energy changes in recipient countries. e n e r g y i t represents proportion of coal power generation in total power generation ( c o a l p o w e r i t )   and renewable energy power generation ( r e n e w a b l e s i t ) . Formula (3) shows the impact of renewable energy development on coal power consumption in developing countries, which can reflect the degree of substitution of the former for the latter. Formula (4) shows the impact of renewable energy development in developing countries on coal power consumption under the background of climate assistance.

In order to ensure the comprehensiveness of the regression results, the significance level of the results is also reported here. This paper uses t test for analysis. In addition, to make the research results more logical, this paper presents the subsequent research procedure in the form of a diagram. The specific results are shown in Figure 3.

Explained variable: the total greenhouse gas carbon emissions equivalent value ( C O 2 ), energy consumption carbon emissions equivalent value ( e n e r g y C O 2 ). The above two variables are measured by carbon emissions level, and the unit is thousand tons of carbon dioxide equivalent. The reason why the total greenhouse gas carbon emissions equivalent value is selected as the research object is that non carbon dioxide emissions may distort the results of climate change policies. Limiting carbon dioxide emissions in climate change policy analysis will greatly underestimate the actual impact of policies. Therefore, it is necessary to take non-carbon dioxide emissions as the research object (Nong et al., 2021). In addition, because the total carbon emissions can’t fully reflect a country’s contribution to carbon emissions reduction, this paper makes a supplementary analysis from carbon emissions per capita and carbon emissions per unit of GDP in benchmark regression.

Core explanatory variables: In this paper, we select the climate assistance behavior in 2000 as the core explanatory variable, and the value of the countries receiving assistance is 1 ¹ , otherwise it is 0. In addition, the amount of climate-related development finance, the amount of a climate-related development finance with adaptation (adapt), and the amount of a climate-related development finance with mitigation (mitigate) are also regarded as the core explanatory variables in the subsequent test.

Control variable: Considering the great influence of economic factors on carbon emissions, this paper mainly identifies and controls carbon emissions from the perspectives of economic development, consumption, and investment and trade activities, respectively, using GDP (GDP) (Wang, 2011) and its square (GDP ² ) (Acaravci and Ozturk, 2010), household consumption expenditure (priconsum) (Nonhebel and Moll, 2008), government consumption expenditure (pubconsum), total capital (grosscapital), fixed capital (grossfixedcapit), as well as export (export) and import (import) (Ferdousi and Qamruzzaman, 2017) to express the correlation. In addition, this paper believes that the degree of urbanization has a direct impact on regional energy consumption (Liu, 2009). Therefore, this paper uses the proportion of urban population in the total population to measure the degree of urbanization (urban) and adds it into the control variables.

This paper addresses developing countries, and the original data of climate-related development finance comes from the open database of the OECD ² . The data related to carbon emissions are mainly from Wind database. The data of proportion of coal power generation in total and renewable energy power production capacity are mainly from the World Bank (WB). The rest of the data comes from the public data collected by the United Nations Statistics Division (UNSD). In this study, all the variables involved were analyzed using descriptive statistics. The mean of a variable represents the average level of all countries under this variable, and standard deviation represents dispersion degree of data of all countries under this variable. Since many variables need to be placed in the same model for analysis, a large numerical gap between variables may impact the stability of regression results. The mean value in a reasonable range shows that the relationship between variables is stable, which is conducive to improving the scientificity of regression results. Standard deviation in a reasonable range indicates that there is no abnormal value and the data is reliable. It can be seen that the values of all the variables were in a reasonable range, which shows that the variables selected in this study were effective. The specific results are shown in Table 1.

This paper estimated model (1) to evaluate the effect of the assistance behavior of climate-related funding on the carbon emissions of recipients. The results are shown in Table 2. As can be seen from columns (1)—(2), the assistance behavior of climate-related development finance had an inhibitory effect on the total greenhouse gas carbon emissions equivalent value and energy consumption carbon emissions equivalent value of recipient countries, and above results passed the 1% significance level test (the p-values here are all 0.002), respectively. Specifically, there was a unit of assistance behavior of climate-related development finance, the total greenhouse gas carbon emissions equivalent value in recipient countries will be reduced by 68%, while energy consumption carbon emissions equivalent value will be reduced by 60%. This shows that assistance behavior of climate-related development finance can effectively reduce the carbon emissions level of recipient countries and improve their climate governance status. Because the total carbon emissions can’t fully reflect a country’s contribution to carbon emission reduction, this paper makes a supplementary analysis from carbon emissions per capita and carbon emissions per unit of GDP. The specific results are listed in column (3)—(6). It can be seen that climate assistance behavior also helps to reduce carbon emissions per capita and carbon emissions per unit of GDP in recipient countries, which demonstrates that the climate assistance behavior of developed countries to developing countries is effective in reducing local carbon emissions.

The reason for this result may be that the economic development level of developing countries is generally low and their financial strength is weak, particularly given the limited financial support for addressing the challenges of climate change. The financial support provided by developed countries eliminates, to some extent, the financing gap of climate finance in developing countries. This provides more opportunities for exploring feasible ways to reduce carbon dioxide and, thus, directly promotes the realization of carbon emission reduction in these countries. Additional data analysis revealed that the number of developing countries receiving assistance expanded from 93 in 2000 to 145 in 2017, and the support for climate-related development finance for developing countries increased from 398 million dollars in 2000 to 56 billion dollars in 2017. This demonstrates that it is the support of external capital inflow that promotes the realization of carbon emission reduction effects in developing countries.

It is necessary for developed countries to provide climate assistance to developing countries. Most developing countries mainly rely on the development of fossil energy such as coal to promote economic development (Akram et al., 2020). This model of sacrificing the environment for economic growth is an important reason that restricts them to achieving good results in climate governance. Therefore, promoting the process of climate assistance to developing countries can provide them with sufficient funds from the outside, so that the funds can flow into the fields of environmental governance, renewable energy development, education and scientific research, fundamentally change their economic development model, and then achieve sustainable development. In addition, climate governance is a problem that all countries in the world need to face. Therefore, climate assistance from developed countries to developing countries is a new attempt to jointly address the challenge of climate change, which is conducive to give full play to the financial and technological advantages of developed countries, and mobilize more developing countries to participate in the process of climate governance, so as to promote the faster realization of the global carbon emission reduction target. Therefore, assistance of climate-related fund is necessary and effective at present.

Heterogeneity analysis of island countries and non-island countries. The essence of developed countries’ assistance behavior in regard to climate-related development finance to developing countries is that they help developing countries jointly cope with the challenges brought about by climate change. Thoughtful analysis shows that the assistance behavior of climate-related development finance effectively reduced carbon emissions in developing countries, but whether this effectiveness fully meets the needs of developing countries requires further discussion. Based on this, this paper divides developing countries into island countries and non-island countries for analysis. The reason why developing countries are divided into island countries and non-island countries here is that this paper attempts to discuss whether climate assistance focuses on such countries threatened by sea-level rise, that is, island countries. Under the influence of climate change, island countries are more vulnerable to the threat of sea level rise and food crisis (Organization, 2008; Bijay et al., 2013). Therefore, from the theoretical perspective, climate assistance should focus more on such countries to help them improve their ability to deal with climate change. The island countries here are mainly composed of Small Island Developing States (SIDS) and other island countries. ³

Through the comparison, the results (see Table 3) show that compared with island countries, assistance behavior of climate-related development finance had a positive impact on carbon emissions reduction of non-island countries. There was one unit of assistance behavior of climate-related development finance, and the total carbon dioxide equivalent of greenhouse gas emissions in non-island countries will be reduced by 87%, while energy consumption carbon emissions equivalent value will be reduced by 83%. And the above results passed the 1% significance level test (the p-values here are 0.004 and 0.003, respectively). Assistance behavior of climate-related development finance also had an inhibitory effect on the carbon emissions reduction of island countries, but the effect is not significant, indicating that the climate governance effect of island countries that are more urgent to deal with the challenge of climate change is not ideal under the background of climate assistance. This result also reflects that climate assistance should focus on island countries more affected by climate change in the future, that is, there is a deviation in the current direction of climate assistance behavior. The foresight report on small island developing States issued by the United Nations Environment Programme (UNEP) in 2014 pointed out that with global warming and the increase in the scale and frequency of natural disasters, 52 small island countries around the world face trillions of dollars of economic losses every year. Extreme climate promotes the rise of sea level, which directly increases the threat of flood and coastal erosion to urban tourism and the damage to infrastructure. The deviation of climate-related development finance direction to island countries makes such countries more passive in climate governance.

This paper combs the island countries and non-island countries involved, and finds that in terms of quantity, the number of island countries is 54, while the number of non-island countries is 119. This makes it easier for climate finance to focus on non-island countries. In addition, developed countries may have assistance preferences in the process of climate assistance, which is closely related to factors such as history, geopolitics, economic and trade relations. Due to relevant factors, island countries are generally not as concerned as other countries. Therefore, island countries may receive less attention in the climate-related development finance. This paper also combs the assistance financing and finds that the total amount of climate financing received by island countries was 38.96 billion dollars from 2000 to 2017. And the total amount of climate financing received by non-island countries was 308.75 billion dollars from 2000 to 2017. This also confirms the speculation that a larger number of non-island countries receive higher attention.

The positive impact of climate assistance on carbon emissions reduction of island countries is not significant. In order to make a more detailed analysis of the situation in the group of island countries and discuss the importance of climate-related development finance inflow, this paper takes island countries as the research object, and divides them into strong climate assistance island countries and weak climate assistance island countries according to the degree of assistance amount to analyze the carbon emissions reduction effect of climate assistance. Through the analysis, it is found that climate assistance has a more significant impact on the carbon emissions reduction effect of island countries receiving more assistance. The results of total carbon emission and energy carbon emissions passed the significance level test of 5 and 10% respectively (p-values are 0.040 and 0.053, respectively). However, climate assistance has no positive effect on carbon emissions reduction in island countries with less assistance. Therefore, it can be considered that within the group of island countries, the inflow of assistance funds is indeed an important factor affecting the realization of the carbon emissions reduction targets of island countries, so it is necessary to increase the support for climate assistance to island countries. The specific results are shown in Supplementary Appendix Table A1 of the appendix.

The purpose of climate assistance is to help developing countries in need of better climate governance. Facing the threat of climate change, island countries are more affected by the impact of sea level rise, food crisis and extreme weather on domestic economic development. Therefore, the flow direction and assistance effect of climate assistance fund should focus more on such countries. At present, the effect of climate assistance to island countries is not ideal. On the one hand, it shows that the flow of assistance funds is not inclined to the focus of such countries. On the other hand, it also shows that island countries need more investment in climate governance to produce the target effect. Study has shown that climate-related development finance flowing to developing countries accounts for only 4% of the reported total (Donner et al., 2016), which could be inferred that less funds flow to island countries. Therefore, it is worth discussing whether the direction of climate assistance needs to be adjusted in the future.

Heterogeneity analysis under different economic development levels. The differences in the economic development levels among countries indicate the different needs for aid recipients. Countries with low economic development levels may require more attention and assistance coping with the series of challenges related to climate change. Assistance is an important component of financing for such countries (Bourguignon and Platteau, 2021). Based on this, and in accordance with the list of the least developed countries released by the United Nations in 2014, this paper categorized developing countries into countries with a high degree of economic development and countries with a low degree of economic development, ⁴ and then evaluated the assistance effect of recipient countries with different levels of economic development. The specific results are in Table 4.

It can be seen that compared with countries with low economic development level, climate assistance behavior had a significant inhibitory effect on carbon emissions of countries with high economic development level. Specifically, there was one unit of assistance behavior of climate-related development finance, and the total carbon dioxide equivalent of greenhouse gas emissions in countries with high economic development level will be reduced by 81%, while the carbon dioxide equivalent of energy consumption will be reduced by 70%. And the above results passed the 1% significance level test (the p-values here are all 0.001). However, the effect of climate assistance behavior on carbon emissions reduction in low economic development countries is not ideal. This result shows that climate-related development finance mainly flow to countries with high levels of economic development and insufficient investment in countries with low levels of economic development, indicating that the current direction of climate assistance has deviated. For countries with low economic development level, their coping ability is weak in the face of extreme weather and climate. In 2016, the World bank issued a relevant report pointing out that poverty makes relevant regions unable to resist natural disasters, and at the same time, the emergence of extreme climate will in turn exacerbate the problem of global poverty. It can be seen that it is necessary to pay attention to climate assistance to countries with low economic development level.

As the number of low-economy developing countries in this paper is only 48, the rest are high-economy developing countries. It can be considered that in terms of quantity, developing countries with high economy are more likely to get attention. To confirm this prediction, this paper further analyzed the amount of assistance funds received by these two types of countries, and found that low-economic developing countries received a total of 72.20 billion dollars in climate assistance funds from 2000 to 2017, while high-economic developing countries received 273.41 billion dollars. It is the high economic developing countries that have received more climate funding and attention, so their carbon emissions reduction effect is more significant than that of low economic developing countries.

For countries with low economic development, it is easier to drive economic development by sacrificing the environment and using cheaper fuels such as coal. At the same time, due to the low level of economic development, the funds invested in the field of climate governance are also very limited. Therefore, the support of the climate-related development finance for the sustainable development of such countries needs to be greater. At present, the impact of climate assistance on carbon emissions reduction in countries with low economic development has not achieved positive results. It can be considered that the attention of climate-related development finance to such countries needs to be further improved, the existing direction of climate assistance should be adjusted, and the number of climate-related development finance for countries with low economic development should be increased, so as to improve the climate governance capacity of such economically vulnerable countries.

Climate assistance has a positive impact on the development of renewable energy in developing countries, so the way to achieve this result is worthy of in-depth discussion. As the development of renewable energy requires high technology, the level of science and technology could be regarded as the key to the development of renewable energy (Geng and Ji, 2016). And after sorting out the key areas of climate assistance, this paper found that climate-related development finance of developed countries pay special attention to the education field of developing countries (this field is included in department of education, health and infrastructure). Education is the basic support for a country’s scientific research and development, which helps to improve the quality of human capital and promote the realization of scientific and technological innovation (Dakhli and De, 2004; Wu, 2006; Xu et al., 2006). The inflow of climate-related development finance into the field of education can be regarded as support for the scientific and technological development of the recipient country. Therefore, the research on the current situation of science and technology development in recipient countries under climate assistance will help this paper further understand the reasons for the rapid development of renewable energy.

Based on this background, from the perspective of science and technology, this paper attempts to deeply explore the impact of climate-related development finance on the scientific and technological level of developing countries, so as to better identify the key factors promoting the positive transformation of energy structure. This paper selects the number of scientific researchers in developing countries as the proxy variable of scientific and technological level. In addition, because the impact of different types of climate-related development finance may be different, this paper analyzes the impact of the development of total climate-related development finance (totalfund), adaptive climate-related development finance (adapt) and mitigation climate-related development finance (mitigate) on technology development respectively. As shown in Table 6.

From columns (1)—(2), it can be seen that climate-related development finance helps to promote the increase of the number of scientific researchers in recipient countries, indicating that climate-related development finance has a positive impact on promoting the improvement of scientific and technological level of recipient countries (the p-values here are 0.013 and 0.065, respectively). Columns (3)—(6) show impact of the adaptation and mitigation climate-related development finance on the number of scientific researchers in the recipient country. Whether from the regression coefficient or significant results, the development of the adaptation climate-related development finance has a more positive impact on the increase of the number of scientific researchers in the recipient country (the p-values here are 0.034 and 0.095, respectively). In other words, compared with the mitigation climate-related development finance, the development of adaptive climate-related development finance plays a greater role in promoting the technological progress of recipient countries. In conclusion, the increase of climate-related development finance can indeed have a positive impact on the improvement of scientific and technological level in developing countries, and technological progress plays an important supporting role in the development of renewable energy. Therefore, it can be considered that the impact of climate assistance on renewable energy in developing countries is mainly achieved by promoting the improvement of local technological level.

It is necessary to promote inflow of climate-related development finance into field of education in developing countries, and the development of adaptive climate-related development finance is indispensable. To fundamentally promote the transformation of the economic development model of developing countries and gradually reduce their dependence on fossil energy dominated by coal, it needs to vigorously improve the utilization efficiency of energy (Sarkar and Singh, 2010) and the substitution of renewable energy, which all need to be based on the improvement of technical level. The sustainable development of cities needs to balance the relationship between economy and environment, and technological innovation is the key to achieve the balance between the two. On the premise of ensuring the level of economic development, the improvement of technical level can improve energy utilization efficiency, promote the reduction of fossil energy consumption, lay a technical foundation for the development of renewable energy, and finally achieve the goal of environmental protection. Therefore, to promote the improvement of urban climate and environment in developing countries and reduce the frequency of extreme weather, so as to maintain the sustainable development model, we need to fundamentally improve the technical level. The development of education can provide a strong foundation for the training of scientific researchers and the improvement of technical level in a country. Therefore, paying attention to the investment in education in developing countries and guiding the climate-related development finance to flow more into the field of education is the fundamental policy to promote the sustainable transformation and development of developing countries in the long term.

Through the above analysis, it can be seen that different types of climate-related development finance have different results in supporting local technology development. Compared with the mitigation climate-related development finance, the effect of the adaptive climate-related development finance is more significant. In view of this result, this paper makes a comparative analysis on the development trend of these two types of climate-related development finance, trying to clarify the relationship between this result and the development of assistance amount, that is, does the positive effect of adaptive climate assistance also mean that this type of assistance amount is higher? So as to clarify whether there is a deviation from the current assistance direction. The specific results are shown in Figure 4.

It can be seen that the mitigation climate-related development finance has been paid attention to since 2000, and has shown a rapid upward trend. By 2017, the amount of assistance in this field has exceeded 38 billion US dollars. The development of adaptive climate assistance started late and the development speed is relatively slow. In 2017, the amount of assistance received in this field was about 23 billion dollars. Combined with the results in Table 6, it is not difficult to find that the field of adaptive climate has received less attention, but it has a better positive impact on the technological development of developing countries. Therefore, whether more attention needs to be paid to adaptive climate is worth discussing.

In view of the large differences in the scale of climate-related development finance flowing into the two fields of climate adaptation and climate mitigation, this paper analyzes the carbon emission reduction results and the development impact of renewable energy brought by the two types of climate-related development finance, and tries to compare the effectiveness of the results brought by the current focus on the fields of climate adaptation and climate mitigation to better evaluate the development status of the two types of climate assistance. The specific results are shown in Table 7.

Columns (1)—(4) show the impact of these two types of climate-related development finance on carbon emissions of recipient countries. It can be seen that the development of these two types of climate-related development finance helps to reduce carbon emissions of recipient countries, but the emissions reduction effect brought by adaptive climate assistance is more significant, and the result passed the 10% significance level test (the p-values here are 0.073 and 0.065, respectively). From columns (5)—(6), it can be seen that these two types of climate-related development finance have a positive role in promoting the development of renewable energy in recipient countries (the p-values here are 0.000), but by comparing the regression coefficient, it is found that the promotion effect of adaptive climate assistance is better. Combined with Figure 4, this paper believes that under the background of more mitigation climate-related development finance, its impact on carbon emissions reduction and renewable energy development is not as good as that brought by the adaptation climate-related development finance. In the face of extreme weather in cities, it is particularly important to adapt to the climate. In order to cope with possible floods, the construction of urban coastline and the improvement of urban infrastructure are an important part of adapting to the climate. At the same time, in order to reduce the adverse effects of extreme drought, rational allocation of urban water and agricultural irrigation water is also a typical way to adapt to climate change. Better adaptation to the climate is conducive to the sustainable development of all cities. So the current attention to adaptive climate assistance should be further improved (Pauw, 2021).

The support of climate-related development finance is the basis for promoting climate governance and the development of specific recipient sectors in recipient countries. Therefore, it is very necessary to study the preference and direction of fund inflow. From the perspective of different types of climate assistance, we have carried out a detailed analysis of the trend of fund inflow and its impact on carbon emissions reduction. And does the fund have a certain preference and focus on the inflow of different sectors? Based on this, this paper analyzes the current focus of climate-related development finance on key sectors. Due to the large difference in the amount of climate assistance received by recipient departments, this paper presents it through two figures. The specific results are shown in Figure 5, Figure 6 and Figure 7.

Through comparison, it can be seen that the four sectors “Education, health and infrastructure,” “Transportation, energy and business services,” “Various industries” and “Environmental protection and others” receive more climate-related development finance, and the development of climate-related development finance is faster, especially the two sectors “Education, health and infrastructure” and “Transportation, energy and business services” receive more attention. The two sectors “Budget support and commodity aid” and “Emergency, post disaster construction and disaster prevention” received less attention. It can be seen that climate assistance is more focused on environmental related sectors at present, such as transportation and energy. Therefore, there is no deviation in the current direction of climate assistance in support of specific sectors.