Conflict or Coordination? The interactions between climate change mitigation and adaptation: Evidence from China

4 As two important strategies to reduce adverse climate effects, mitigation and adaptation actions can 5 interact, resulting in synergies or trade-offs. Using data from 30 Chinese provinces from 2008 to 2017, 6 this study employs a panel vector autoregression (PVAR) model to study the interactive relationships 7 between mitigation and adaptation. Moreover, based on the coupling coordination model, this paper 8 investigates the coordination degree of mitigation and adaptation in China. The results show that 1) there 9 is Granger causality between mitigation and adaptation, and the positive impact of mitigation on 10 adaptation is greater than the negative impact of adaptation on mitigation. Therefore, an integrated 11 approach that considers these interactions can help enhance synergy and create a win-win situation. 2) 12 The dynamic relationship between mitigation and adaptation in China has reached a barely balanced 13 stage, and there are large regional differences. 3) Compared with the mitigation evaluation value, the 14 adaptation evaluation value has a more positive effect on promoting an increase in the coordination 15 degree. These findings can contribute to the formulation of effective regional sustainable development 16 strategies. 17


20
Climate change has had a visible effect on the natural and human environment, and inevitably, 21 decisions must be made to mitigate greenhouse gases and adapt climate change to cope with the rapidly

39
Although the majority of interactions are deemed positive (Berry et al., 2015), there is the possibility 40 of maladaptation (the "problem of increasing risks from adaptation") or malmitigation (i.e., increasing 41 risks from mitigation). Therefore, it is necessary to enhance the synergies between mitigation and 42 adaptation actions to expand shared interests and weaken conflicts. Against this background, there has 43 been a growing body of literature addressing the interaction between mitigation and adaptation from 44 different perspectives, e.g., the feasibility of simultaneously implementing mitigation and adaptation 45 strategies (Locatelli et al. 2011;Wilbanks et al. 2007;Hulme et al. 2009), examples of synergies between 46 mitigation and adaptation measures in different sectors (Sharifi, 2021;Berry et al. 2015), core drivers 47 contributing to or hindering synergies (Landauer et al. 2015;Landauer et al. 2018) and so on.

48
Despite growing interest in the linkage between mitigation and adaptation, the majority of extant 49 works only discuss mutual relations without empirically testing them. Most generally focus on 50 "relationship discovery" rather than "relation mining", which means there is a lack of empirical research 51 investigating the level of integration between adaptation and mitigation and how this integration affects 52 outcomes (Grafakos et al. 2020;Grafakos et al. 2019

87
According to the existing literature, the ratio of the added value of secondary industry to regional GDP 88 is adopted as an indicator in this article. The lower the IS value is, the more reasonable the industrial 89 structure, indicating that the economy is inclined toward developing in the direction of "low pollution 90 and low energy consumption". The energy structure (ES) is a key factor in reducing greenhouse gas 91 emissions. This study adopts the proportion of thermal power generation in electricity generation, and a 92 lower value of ES means a more sustainable energy structure. Energy intensity (EI) refers to the quantity 93 of energy required per unit output or activity, so this paper adopts the energy consumption per 10,000 94 yuan of real GDP as the indicator. Low EI values are a proxy for energy efficiency improvements and Indicators that have a negative contribution to the AEV and MEV, where i refers to the ith sample (i = 1, 2, …, n), j refers to the jth indicator (j =1, 2, …, m), Max is the 125 maximum value of a given indicator, and Min is the minimum value of a given indicator.

126
To reduce interference from subjective selection factors, the entropy method is used to calculate the 127 indicator weight according to its variability.

128
Equations 3 and 4 show the methods for calculating the MEV and the AEV, respectively.

133
where N, E, F, H and S refer to dimensionless data normalized by the maximum and minimum methods.

137
Descriptive statistics for each variable are presented in Table 2. In addition, we visualize the raw 138 data in Fig. 1. Fig. 1 documents at least three facts. First, substantial regional inequality exists in 139 mitigation and adaptive capacity. Second, the regional growth rates of these two variables are quite

217
First, the shocks to mitigation and adaptation both have a positive and significant impact on them 218 (Fig 2a and Fig 2d), which means that both are progressive and self-reinforcing over time. However, 219 this positive effect tends to decrease with time. The results are further evidence that a development 220 model that relies solely on adaptation or mitigation is not sustainable. In addressing climate change risks, 221 policy makers should strike a balance between adaptation and mitigation and carry out overall planning 222 to achieve coordinated development.

223
Second, the response of mitigation to adaptation is weakly negative and registers a decreasing trend 224 (Fig. 2b). There are several reasons behind this sign: in the first place, there appears to be a crowding-225 out effect of investment between adaptation and mitigation, given a limited investment budget.

226
Generally, the crowding-out effect of adaptation on mitigation is much larger than that of mitigation on 227 adaptation. That is, under budget shortages, more financial resources tend to be concentrated on

233
Focusing next on the responses of adaptation (Fig. 2c), we note that the response of adaptation to 234 mitigation is positive. This result is consistent with theoretical predictions from previous literature 235 (Ayers and Huq 2009;Landauer et al. 2015). High mitigation capacity means that there is limited room 236 for improvement, and more resources, especially financial resources, can be focused on adaptation 237 activities. Second, in the long run, mitigation activities can reduce the concentration of greenhouse gases 238 in the atmosphere, which will decrease the intensity and frequency of events such as meteorological 239 disasters. In addition, studies have shown that established mitigation policies rather than local climate 240 risk profiles drive cities to join municipal adaptation networks. Cities committed to actual progress on 241 mitigation policy (i.e., with a monitoring system in place) are more likely to adopt adaptation policies 242 (Lee et al., 2020).

244
We next conduct a study of the variance decompositions to complement the impulse response 245 analysis; this can reflect the relative cumulative contribution of each of the variables in the system.

246
Through 200 iterations of the Monte Carlo simulation, the variance decomposition of the two variables 247 for 30 prediction periods can be acquired. As is reported in

249
Overall, there is a positive relationship between each variable and its predicted values. The results

250
in Table 5 report that the interpretability of the MEV to its predicted values accounts for 91% in the 10 th 251 period, and it remains at the 90% level in the 30 th period, indicating that the MEV follows a process of 252 continuous accumulation. In other words, the predicted values of the MEV are to a large extent 253 determined by its current values. By contrast, the error term decomposition results for the AEV show 254 that its own interpretability is lower than that of the MEV to its predicted values.

255
In addition, the explanatory power of each variable toward the other variables gradually increased.

256
In the 30th forecast period, the impact of the MEV on the AEV (44.1%) was greater than the impact of 257 the AEV on the MEV (10%). This is indicative of that mitigation take a more significant impact on

286
These results confirm that China has achieved basic coordination between mitigation and 287 adaptation systems, while there are large regional differences in terms of coordination development. As 288 shown in Fig. 4, the coordination degree of most provinces was between 0.5 and 0.6. Only a handful of 289 provinces with both higher development stages and environmental quality scored higher than 0.6, such To determine the reasons for the non-ideal coordinated development between mitigation and 302 adaptation, Fig. 5 shows the spatial distribution of mitigation system evaluation and adaptation system 303 evaluation values (mean value). Most of the provinces showed unbalanced development between the 304 mitigation and adaptation systems. All 30 provinces performed better for the mitigation system than for the ratio of the added value of second industry to the