Does climate affect regional economic development in China?

Abstract

The climate significantly varies from eastern to western in China. The purpose of this paper is to analyze how climate affects regional economy in a variety of provinces of China. Regions are divided into three distinct regions based on geographic proximity. We choose 19 provinces, six in the east, five in the middle, and eight in the west. Utilizing data spanning between 1997 and 2019, and the ARDL approach, the result shows that the impact of climate on GDP varies across regions. Precipitation contributes to positive economic growth in the high-income provinces gathering in eastern China, for instance, Guangdong, Zhejiang, and Jiangsu. The impact of temperature mainly occurs in under-developed regions in western China. Overall, this paper analyzes how climate impacts differently in different region of China, suggesting policymakers to implement appropriate strategies in different provinces to limit the negative impacts of climate on economy and achieve sustainable economic development in the long run.

Keywords:

• Climate
• GDP
• China
• ARDL

1. Introduction

China’s diverse geographical location and natural conditions have resulted in significant climate and economic variations across different regions, mainly due to the unique west-to-east international transaction flow. Some areas still need to be explored and developed in the western regions. In contrast, in middle and eastern areas, most provinces have achieved a certain level of development, albeit with varying conducive conditions for economic growth. Recognizing the potential of climate in shaping economic development has become a crucial concern, prompting the evaluation of climate as a vital step in addressing the issue of imbalanced economic development. It is undeniable that climate conditions have long influenced production activities since early history (Pei et al., 2016). In historical China, land utilization and agriculture are vital for human survival and development. In modern economic development, production activities also affected by climate (Kalkuhl & Wenz, 2020). The effects of climate on economic performance differ across regions with different geographical locations. How is the correlation between climate and economy from east to west of China? Less studies analyzed the question in regional China by province and this field lacks deeply empirical tests results. This paper compares the differences in precipitation and temperature condition in most provinces of China, analyzing the effect of climate on regional economy from 1997 to 2018. And policy suggestion gives direction to government on how to get better utilization of climate condition in different provinces to achieve future healthy development. To address the issues, we assess annual distribution of rainfall and temperature in conjunction with national economic output. The uniqueness of the study can be seen in the comparison of different effect of climate from east to west, in individual provinces and diverse regions within a country. This paper analyzes the climate impact on economy to different provinces in China from micro eyesight, especially the different situation in each province. Kahn et al. (2021) found a similar conclusion, but without individual analysis to samples district. Paper of Tang et al. (2023) mentions the economic impacts of climate change and policies are disproportionately distributed across regions. We analyze how the impacts distribute differently in different region of China. We provide a scientific basis for the adjustment of local economic activities, contributing to the scientific basis for optimal regional adaptation to climate. This paper helps policymakers to implement appropriate strategies in the provinces to limit the negative impacts of climate on overall economic, social, and agricultural activities, especially in underdeveloped areas of productivity, to achieve balance and sustainable economic development between regions in the whole country.

In the eastern region of China, extreme climate change events, such as coastal flooding, storm surges, coastal erosion, and seawater intrusion, are more prevalent. Conversely, in the western region, there is a higher occurrence of mudslides and landslides. These observations raise concerns as they can disrupt market equilibrium, even leading to socio-economic crises. If climate change mitigation measures are not implemented, China could experience economic losses ranging from 0.5 percent to 2.3 percent of its GDP as early as 2030. This climate-economic relationship influences the scale and magnitude of climate change’s impact on markets in the coming century. Thus, comprehending this relationship is essential for accurately predicting losses caused by anticipated climate change and making informed decisions assessing mitigation strategies’ benefits and costs. However, the economic impact on their performance may differ due to different climates in different provinces. Regional heterogeneity poses challenges to accurately estimating the climate impact on economy. Aiming at quantifying the economic effects of weather conditions on various economic outcomes across different provinces, the empirical analysis conducted in this study provides valuable insights for estimating the costs associated with climate change and identifying effective strategies for mitigating its impacts.

Despite regional disparities in national development, a specific subset of provinces emerges as the primary contributors to China’s economy. Geographically and administratively, China can be categorized into three distinct regions: eastern, middle, and western regions.¹ Notably, China’s eastern region exhibits higher economic development, openness, and urbanization than the middle and western regions. This regional classification is frequently employed in scholarly investigations of China’s interregional disparities (Xu et al., 2019).

At present, the top 8 provinces ranked by average GDP amount from 1997 to 2018, which reached more than the average level of USD200 billion in China, namely Guangdong (USD625 billion), Jiangsu (USD566 billion dollar), Zhejiang (USD367 billion), Henan (USD306 billion), Shandong (USD238 billion), Hubei (USD220 billion) Fujian (USD217 billion) and Beijing (USD205 billion). Among these top 8 provinces and cities, five are in the eastern parts of China, three are in the middle parts of China, and no province is in the western provinces.

Guangdong ranks at the top of GDP in the whole country. Figure 1 shows that Guangdong enjoys the highest temperature and precipitation in the east region. Wherever the economic contribution is mainly promoted by eastern coastal provinces in the whole country. Economic rapid development concentrates on the eastern provinces. Taking up approximately 40% of the total population and 10.5% of the land area, east regions host 84.4% of the FDI amount and 52.1% GDP amount, according to the Foreign Investment Statistics of the Ministry of Commerce (NBS, 2021). Since 1978, the Chinese government has implemented policies to promote the development of the open eastern provinces. However, notable disparities exist in the government’s approach to fostering development in eastern and western provinces. The prevailing warm climate in coastal regions enhances the quality of life for residents and provides favorable temperature and rainfall conditions for manufacturing production.

Figure 1. Ranking of average temperature of each province in 1997 – 2018 in China (in °C).

Source: Chinese National Bureau of Statistics of 1997–2018.

Moreover, the favorable physical conditions in these areas meet the needs of the labor force. China spans over sixty longitudes across five time zones, encompassing an east-west distance of approximately 5200 kilometers. Certain regions experience year-round snow coverage and shallow temperatures, rendering local economic activities challenging to pursue. Consequently, this study examines the potential impact of climatic factors on economically relevant outcomes. The findings demonstrate that temperature and precipitation exert significant and statistically significant influences on various economic outcomes, highlighting the complex and multifaceted relationship between weather conditions and the economy (Dell et al., 2014). This paper aims to explore the influence of climate as a significant determinant of economic outcomes in China, particularly in the transition from coastal to inland regions.

Existing literature on economic growth has primarily focused on specific regions of China, such as the Pearl River Delta and the Yangtze River Delta, thereby neglecting comprehensive comparisons across the three major parts of China. Considering this gap, this study examines climate, foreign direct investment (FDI), exports, and GDP relationships across the different provinces of China. By analyzing data from 1997 to 2018, a period characterized by rapid development and China’s integration into the global economy through policies of openness, revolution, and WTO membership, this study empirically reveals the nuanced relationship between climate conditions and provincial GDP using appropriate econometric techniques. We use the Autoregressive Distributed Lag (ARDL) model to investigate the short-term and long-term effects. This study addresses the research gap by utilizing provincial-level climate factors, employing econometric techniques, and incorporating the latest data for different provincial regions.

The organization of the paper: Section 2 is a literature review. Section 3 introduces the data sources and methodology employed for analyzing the results. Section 4 presents the empirical findings in detail, utilizing panel econometric methodologies. Finally, Section 5 concludes the paper by presenting practical policy recommendations for fostering future economic growth in different provinces of China.

2. Literature review

2.1. Climate impact on GDP

It is widely acknowledged that the climate-economy relationship plays a crucial role in market impacts (Kompas et al., 2018). Due to the climate-economy relationship, it is easier to project damage associated with anticipated climate change. This relationship reveals that warmer climates stimulate economic growth in colder countries while impeding growth in countries with hot environments. In Turkey, temperature increment negatively affects agricultural GDP (Kilicarslan & Dumrul, 2017). From a policy-making perspective, the climate-economy relationship can be utilized to assess the benefits and costs associated with climate change mitigation efforts. Developed and developing countries have experienced impeded agricultural and industrial output growth due to the ongoing global warming trend. Scientific models estimate that the optimal global temperature for GDP growth is approximately 13 degrees Celsius. Without mitigating climate change, these models predict a potential global income loss of 23% by 2100 (Burke et al., 2015; Newell et al., 2021).

2.2. Temperature impact

Temperature and precipitation changes have demonstrated substantial repercussions on economic activity. Research estimates indicate that a 5°C warming could potentially lead to GDP reduction ranging from 0% to 20% (Burke et al., 2015; Dell et al., 2014). Global warming affects hot regions more profoundly than cooler regions, significantly negatively affecting Gross Regional Product (GRP). To elucidate the marginal effect of temperature changes, it is observed that across various specifications, temperature growth has damages to marginal effect on GRP, especially in the case of high average temperatures. At a significance level of 10% or higher, when the temperature is 10°C, there is no marginal effect; at 25°C, the marginal effect is significant, ranging from 2.6% to 3.7%. Thus, with the temperature rise of 1 Celsius degree, GRP decreases by approximately 3% in a hot region. An optimal annual average temperature level can be determined by non-linear panel models, which lie between 5°C and 10°C.

Furthermore, alterations in annual mean temperatures exert a non-linear influence on economic output (Kalkuhl & Wenz, 2020). Temperature increases generally boost GRP in regions with annual mean temperatures below 5°C and diminish GRP in hot regions. Long time temperature increase result in a continuing economic output decline in hot regions. For instance, a 1°C temperature rise in a region with an annual mean temperature of 25°C corresponds to a 3.5% reduction in GRP within that region. This economic loss can be attributed to production disruptions caused by average temperatures higher than 27°C. Additionally, in the Caribbean and Central America, economically substantial effects of the yearly average temperature on total output and sectoral production are revealed by linear model estimation (Hsiang, 2010). However, at the international level, increasing temperatures have not hindered economic growth over the past 50 years (Sequeira et al., 2018).

In the developed region, GDP has no statistically significant relation with climate. Industrialized nations have better temperature and climate adaptation capabilities. Its economic sectors are less vulnerable to weather and climate and exhibit lower susceptibility to climate change impacts (Newell et al., 2021). The GDP growth climate can impact rich and non-agricultural production countries. Additionally, the findings from Burke et al. (2015) state that the aggregate macroeconomic productivity of wealthy countries is not affected by temperature, while the productivity of poor countries is reported to be linearly responsive to temperature (Adam & Drakos, 2022).

Furthermore, for economic growth in financially stable countries, temperature growth has limited effects, but it does negatively impact output levels and growth rates. Another noteworthy finding is the non-linear relationship between temperature and economic growth in Sub-Saharan countries, characterized by a “Laffer curve” shape, resembling an inverted U. Economic growth tends to increase when temperatures below 24.9°C but decline for temperatures exceeding that threshold (Paul Alagidede & F, 2016). Over the past 50 years, increasing temperatures have not impeded economic growth in heterogeneous dynamic models at the international level. Temperature rises will show an adverse outcome in lower-income countries.

2.3. Precipitation impact

The investigation reveals that in underdeveloped countries, a strong correlation is noticed between precipitation and the economy. Similarly, the correlation is more substantial in regions with extensive cropland. Precipitation affects Turkey’s agricultural GDP positively (Kilicarslan & Dumrul, 2017). Precipitation and GDP are positively correlated, and a 100 mm rise in rainfall leads to a 0.86% increase in GDP. The benefits of precipitation to the economy concentrates in developed districts (Duan et al., 2022).

However, in lower-income countries with colder climates, a negative effect on the economy is observed with precipitation increments. FMOLS and DOLS models can be employed to validate the negative impact of temperature and rainfall increases on economic growth. The results indicate a decrease in GDP by 0.25% and 0.024% (for FMOLS) and 0.28% and 0.005% (for DOLS) for each unit increase in temperature and rainfall, respectively (Meyghani et al., 2022). However, the impact of precipitation fluctuations on Gross Regional Product (GRP) is generally minimal across all model versions. A study utilizing regional panel regressions estimated rainfall distribution and its macroeconomic effects in 1554 regions worldwide over 40 years. Extreme daily rainfall occurrences caused by human activity have an unfavorable effect on world economies (Kotz et al., 2022). Furthermore, the delayed consequence of climate change on the economy has far-reaching consequences. For example, a global panel sample demonstrates that temperature fluctuation can cause inflation four years later (Mukherjee & Ouattara, 2021). Additionally, natural disasters can have a five-year lasting effect on the economy following their occurrence (Atsalakis et al., 2021).

3. Methodology

An annual time series covering 1997 to 2018 is utilized here with GDP, FDI, and Export data collected from China’s National Bureau of Statistics (NBS) and Bureau statistics of different provinces. The climate data is collected from CMDSC (China Meteorological Data Service Centre). This paper uses the average of 22 years precipitation and temperature data to measure the climate effect. ARDL cointegration test was developed first by Charemza and Deadman (1992) and later improved by Pesaran et al. (2001). It can test whether an economic system is in equilibrium. ARDL Model is superior to other cointegration tests, compared with Engle and Granger’s cointegration test 1987 and Johansen and Juselius’s cointegration test 1990. The main advantage can be that ARDL considers a small sample size. In the realm of econometrics, Pesaran et al. (2001) devises the ARDL bounds testing technique to examine enduring associations between variables featuring diverse orders of integration (either I(0) or I(1)) within a limited sample size. This methodology precludes the presence of any variable integrated at an order exceeding I(1). Once these criteria are met, the subsequent error correction model assesses the cointegration relationship. Next, ARDL and ECM models can be analyzed.

This section delineates the model for exploring the influence of climate factors, namely precipitation, temperature, FDI, and exports, on economic development. Economic development is regarded as the dependent variable in the model, as measured by each province’s real gross domestic product (GDP). Typically, the characteristics of GDP development adhere to a well-established framework in prior literature on economic development, while the literature recognizes the independent variables as robust determinants (Adam & Drakos, 2022; Husain & Javed, 2019; Kong et al., 2021). In selecting variables for the study, we followed the following methods by Kilicarslan and Dumrul (2017), Pata and Isik (2021), and Dell et al. (2015). The model gives the following quotation.

(1) $GD{P}_t={\beta }_0+{\beta }_{1}FD{I}_t\hspace{0.17em}+\hspace{0.17em}{\beta }_2\mathrm{E}\mathrm{X}\mathrm{P}{\mathrm{T}}_{\mathrm{t}}\hspace{0.17em}+\hspace{0.17em}{\beta }_3\mathrm{W}\mathrm{E}\mathrm{A}\mathrm{T}\mathrm{H}\mathrm{E}{\mathrm{R}}_t\hspace{0.17em}+\epsilon \mathrm{t}$(1)

where: T refers to Time period., β₀, β₁, β₂, β₃ refers to positive or negative evaluated coefficients. GDP refers to the GDP amount in a certain province. FDI refers to the foreign direct investment amount shown in the model at time t in a certain province. EXPT refers to the export amount at time t in a certain province. WEATHER refers to the precipitation amount annually or average annual temperature at time t in a certain province. ε_t refers to a random variable of the estimated regression. Previous empirical studies have been mentioned for the above variables’ selection (Ahmad et al., 2022). Different provinces have different conditions, China is a country with large territory area. For β₃, the weather can be positive or negative owing to local weather conditions. Weather includes precipitation and temperature factors from the local annual precipitation or temperature measurement, the coefficients in the model might be significant or not.

Before conducting a cointegration test, it is imperative to ascertain the stationarity of a time series. The augmented Dickey-Fuller test revealed that all variables exhibited I(0) or I(1) integration orders.

(2) $\mathrm{\Delta }{X}_t=\phi X_{t-1}+{\sum }_{i=1}^m\delta \mathrm{\Delta }{X}_{t-i}+{\epsilon }_t$(2)

The difference operator is denoted by ∆, the intercept is represented by α, and t is utilized as a time index. The coefficient ϕ denotes the process root, specifically emphasizing testing, while γ signifies the coefficient on a time trend. The autoregressive model with m lags is denoted by m, and εt represents the random error. In the initial stage, a joint integration test is conducted within the framework of the UECM, which is formulated as follows:

(3) $\mathrm{\Delta }{Y}_t={\alpha }_0+{\sum }_{i=1}^m{\beta }_i\mathrm{\Delta }{Y}_{t=i}+{\sum }_{i=0}^n{\theta }_i\mathrm{\Delta }{X}_{\tau =i}+{\lambda }_1{Y}_{t=1}+{\lambda }_2{X}_{t=1}+{\eta }_t$(3)

The coefficients λ₁ and λ₂ represent the long-run relationship, whereas β and θ explain the short-run relationship. The operator ∆ represents the first differences, and m and n indicate the lag periods, recognizing that these lag periods may not necessarily be equal (m ≠ n). The random error term η exhibits zero means with constant variance and lacks self-continuous connections. The ARDL methodology enables the determination of the optimal number of lag periods. This is typically unattainable in other conventional cointegration tests. Consequently, issues related to residual correlations and subjectivity can be alleviated by optimizing the lag periods.

The common assumptions are here: Null hypothesis (H₀) states no cointegration, i.e., λ₁ = λ₂ = 0, while the alternative hypothesis (H₁) suggests the existence of cointegration, i.e., λ₁ ≠ λ₂ ≠ 0. Comparing the estimated F-value against critical limits before rejecting the null hypothesis, two critical limits are considered: lower (LCB) and upper (UCB) critical bound. Reject null hypothesis when F-value exceeds UCB. Accept null hypothesis when computed F-value below LCB. The result is inconclusive when F-value falls between UCB and LCB (Pesaran et al., 2001). When the variables are cointegrated, then estimate the equation using the formula below:

(4) $Y_t={\alpha }_0+{\sum }_{i=1}^p{\mathrm{\partial }}_i{Y}_{t-1}+{\sum }_{i=0}^q{\delta }_i{X}_{t-1}+{\epsilon }_t$(4)

p and q indicate the deceleration periods, and ε denotes the limit of random errors. Typically, we choose a maximum of two deceleration periods. Thirdly, the error correction model (ECM) is used:

(5) $\mathrm{\Delta }{Y}_t=c+{\sum }_{i=1}^p{\vartheta }_i\mathrm{\Delta }{Y}_{t-i}+{\sum }_{i=1}^q{\delta }_i\mathrm{\Delta }{X}_{t-i}+\psi EC{M}_{t-1}+U_t$(5)

The error correction factor, denoted by the parameter ψ, measures the rate at which the disequilibrium is rectified in the short term, moving towards the long-term equilibrium.

4. Result and discussion

This paper invites FDI and export together as independent variables for empirical tests. All five variables, GDP, precipitation, temperature, FDI, and export of this study, are subjected to Augmented Dickey and Fuller (ADF) test for stationary. For empirical results, we separate precipitation and temperature impacts into two models. The joint impact of temperature and precipitation may be significantly different from that of temperature alone. The results are represented in Table 1 and based on the Augmented Dickey and Fuller (ADF) test suggests a mixed result.

Table 1. Unit root test—ADF

	West Region				West Region				East Region
	Level		1^st Difference		Level		1^st Difference		Level		1^st Difference
	C	C & T	C	C & T	C	C & T	C	C & T	C	C & T	C	C & T
	Sichuan				Shaanxi				Beijing
GDP	2.62	−0.67	−6.17***	−4.19**	1.45	−1.55	−3.20*	−3.32*	9.53	−1.17	0.26	−4.05**
FDI	−0.78	−3.09	−2.69*	−2.62	0.13	−1.37	−5.89***	−5.47**	−0.71	−4.49**	−3.32**	−3.53**
EXP	1.84	−0.99	−3.52**	−4.58***	1.05	−2.99	−4.87***	−5.2***	−3.16	−2.38	−2.84**	−2.70
RAIN	−1.55	−1.22	−3.23**	−3.64*	−4.10**	−3.99**	−4.98***	−4.83***	−1.67	−4.79***	−4.89***	−4.52***
	ChongQing								Shanghai
GDP	−0.36	−0.61	−2.55*	−3.33*					7.62	0.18	0.54	−3.8*
FDI	−1.33	−2.12	−6.24***	−6.09***					0.03	−2.4	−2.98**	−2.78
EXP	2.69	−2.52	−2.72*	−4.73**					−0.89	−1.10	−3.89***	−3.91**
RAIN	−6.12***	−5.96***	−6.17***	−7.14***					−2.49	−2.27	−3.77**	−3.81**
					Middle Region
	Guizhou				Shanxi				Guangdong
GDP	1.52	−2.42	−2.71***	−3.42*	1.28	−2.66	−2.98**	−3.36*	8.43	−0.24	−0.06	−4.11**
FDI	−0.68	−2.15	−4.03***	−4.91***	−0.86	−3.03	−3.78**	−3.69**	−0.83	−2.04	−4.56***	−4.45**
EXP	−1.11	−1.98	−3.99***	−3.76**	−0.61	−3.73**	−5.90***	−5.83***	−0.34	−1.72	−3.46 **	−3.36*
RAIN	−4.76***	−4.64***	−9.43***	−9.24***	−1.61	−2.69	−4.37**	−4.44**	−4.92***	−4.98**	−7.54***	−7.36***
	Tibet				Jilin				Jiangsu
GDP	−1.20	−1.61	−4.14**	−4.34***	−0.59	−2.61	−1.69	−1.42	7.64	−1.42	−0.71	−3.85**
FDI	−2.06	−3.01	−7.12***	−6.95***	1.10	−7.61***	−4.39***	−4.03**	−1.31	−1.6	−2.85**	−2.91
EXP	−2.40	−2.64	−3.11**	−3.07	−1.24	−1.80	−4.34***	−4.33**	0.05	−2.28	−4.18 ***	−4.09**
RAIN	−3.26**	−3.11	−6.34***	−6.33***	−7.46***	−4.04**	−3.57**	−3.47*	−3.66**	−4.32**	−4.77***	−4.59**
	Qinghai				AnHui				Zhejiang
GDP	6.94	−0.15	−0.14	−4.82**	3.98	0.83	0.15	−4.58**	6.74	−0.54	−1.12	−3.70**
FDI	−2.44	−0.92	−3.70	−4.82***	0.10	−1.87	−1.63	−1.97	0.43	−5.86	−4.29***	−4.09**
EXP	−1.67	−3.78**	−3.53**	−3.39*	0.72	−1.83	−3.23**	−3.41**	0.74	−2.76	−3.79***	−3.88***
RAIN	−4.78**	−4.66***	−9.29***	−9.36***	−3.78***	−4.33**	−6.72***	−6.49***	−3.25**	−3.29*	−5.39***	−5.24***
	Xinjiang				Henan				Fujian
GDP	3.72	−1.06	−1.65	−4.55**	6.61	0.53	−0.76	−4.16*	3.84	1.49	3.83**	−3.73**
FDI	−2.89*	−2.49	−3.38**	−3.43**	6.18	−1.23	−1.39	−2.94	−2.06	−2.01	−3.88**	−3.82**
EXP	−1.18	−1.95	−4.59**	−4.52**	0.64	−2.83	−2.85**	−2.89	0.05	−2.59	−3.13**	−3.06
RAIN	−4.38**	−4.28**	−6.62***	−6.43***	−4.24***	−4.22**	−8.56***	−8.36***	−5.39***	−4.24**	−5.46***	−5.48***
	Guangxi				Hubei
GDP	7.24	−0.86	−0.43	−3.34**	3.29	0.54	0.29	−3.91**
FDI	−2.16	−2.93	−4.94**	−4.84**	0.15	−3.22	−3.07**	−2.92
EXP	1.70	−1.35	−3.73**	−3.74**	−1.15	−2.03	−1.52	−4.76***
RAIN	−5.32***	−5.23***	−8.43***	−5.35***	−4.43***	−4.33**	−4.66***	−4.59**

Note: TEMP is omitted to conserve space. Available upon request. C is constant without a trend, and C & T refers to constant with a trend. Asterisks *, **, and *** denote significance at 10%, 5%, and 1% critical values, respectively.

As shown in Table 1, The variables under examination exhibit stationarity at levels I(0) and I(1). The null hypothesis cannot be refused when the test statistics for these assessments fail to reach statistical significance. We separate the 19 provinces in China into the East, Middle, and West three districts to do detailed analyses. In the eastern districts, GDP, precipitation, temperature, FDI, and export are all I(0) or I(1) suitable for the ARDL test. ARDL bounds test results are presented in Table 2. The empirical findings indicate the rejection of the null hypothesis of no cointegration at a significance level of 1%. This is supported by the F test statistic, which surpasses the critical values, providing evidence of a co-integrating relationship. The ARDL bounds test is employed to examine the existence of a cointegration relationship between GDP, precipitation, temperature, FDI, and export across different provinces. The results reveal that all provinces’ values exceed the 5% upper bound. This suggests that the long-run equilibrium relationship exists.

Table 2. Bound cointegration test

West Region	WEAT=TEMP	WEAT=RAIN	Middle Region	WEAT=TEMP	WEAT=RAIN	East Region	WEAT= TEMP	WEAT=RAIN
Sichuan	62.12***	76.57***	Shanxi	14.26***	5.91***	Beijing	8.54***	8.28***
Chongqing	6.43***	7.57**	Jilin	3.72**	3.28**	Shanghai	11.04***	7.73**
Guizhou	8.43***	22.02***	Anhui	12.55***	13.79***	Guangdong	9.77***	23.59***
Tibet	17.98***	66.05***	Henan	25.76***	39.41***	Jiangsu	10.11***	35.25***
Shaanxi	7.96***	20.29***	Hubei	4.89**	23.04***	Zhejiang	11.62***	34.05***
Qinghai	10.33***	22.24***				Fujian	9.43***	6.39***
Xinjiang	13.50**	7.19***
Guangxi	95.78***	75.41***
Critical Value
1%	3. 59		5%	4.31		10%	5.97

Note: Asterisks *, **, and *** denote significance at 10%, 5%, and 1% critical values, respectively.

The ECM findings for the west, middle, and east regions are presented in Table 3, with levels of significance denoted as ***, **, and *, corresponding to the 1%, 5%, and 10% thresholds, respectively. The operator Δ represents the difference operation. Notably, the highly significant and negative coefficient of the Error Correction Term (ECT) at a 1% level signifies the speed of adjustment to a long-term equilibrium relationship. Using ARDL framework, as presented in Table 4, the long-run analysis provides insight into the enduring relationship between the variables.

Table 3. Error correction model

Panel I: Error Correction Model	Panel II: Diagnostic Test^a
	AC	HET	NORM	STAB
East provinces:
Beijing: ∆GDP = − 0.14ECT(−1)*** − 1.51 ∆GDP(−1)** − 38.84∆FDI − 78.91∆FDI(−1)** + 1.07∆EXP*** + 0.74∆EXP(−1)** − 57.51**	17.16	1.32	0.51	0.15
Shanghai: ∆GDP = − 0.07ECT(−1)*** − 2.13∆FDI − 4.69∆FDI(−1)*** −0.13∆RAIN***	0.39	0.75	2.43	0.38
GuangDong: ∆GDP = − 0.13ECT(−1)*** + 0.26∆GDP(−1)*** − 2.8∆FDI*** + 2.72∆FDI(−1)*** + 0.38∆EXP***-0.49∆EXP(−1)*** + 0.09∆RAIN*** − 0.42∆RAIN (−1)***	4.71	0.29	4.71	1.19
Jiangsu: ∆GDP = − 0.12ECT(−1)*** − 0.34 ∆GDP(−1)** + 0.53∆FDI + 1.23∆FDI(−1)** + 0.49∆EXP*** −0.33∆EXP(−1)*** + 0.12∆RAIN*** − 0.28∆RAIN(−1)***	4.46	0.59	7.94	0.69
Zhejiang: ∆GDP = − 0.14ECT(−1)*** + 8.43∆FDI*** − 11.94∆FDI(−1)** + 0.33∆EXP*** − 0.52∆EXP(−1)*** + 0.22∆RAIN*** − 0.51∆RAIN(−1)***	0.71	0.73	0.29	0.79
Fujian: ∆GDP = − 0.12ECT(−1)*** − 0.17∆FDI − 2.16∆FDI(−1)* + 0.75∆EXP*** + 0.01∆RAIN − 0.09∆RAIN(−1)** − 431.46***	1.38	0.55	0.38	1.52
Middle provinces:
ShanXi: ∆GDP = − 0.25ECT(−1)*** − 2.97∆FDI	1.18	1.94	2.87	1.93
Jilin: ∆GDP = −0.26ECT(−1)*** + 0.46∆GDP(−1)*** + 0.07RAIN**	1.20	1.13	0.67	1.19
Anhui: ∆GDP = − 0.93ECT(−1)*** − 0.61∆GDP(−1)** + 5.94∆FDI*** + 19.14∆FDI(−1)*** − 3.43 ∆EXP*** + 1.03∆EXP(−1)** − 0.04RAIN + 0.09∆RAIN(−1)***	0.99	3.05	0.73	0.10
Henan: ∆GDP = − 0.28ECT(−1)*** + 1.50∆FDI − 0.08∆RAIN − 159.38***	6.02	4.55	0.32	0.45
Hubei: ∆GDP = − 0.06ECT(−1)*** + 0.29∆GDP(−1)** + 0.60∆EXP + 1350.64***	6.97	1.37	0.29	0.42
West Provinces:
Sichuan: ∆GDP = − 0.23ECT(−1)*** − 0.41∆GDP(−1)*** + 6.59∆FDI*** + 0.64∆EXP*** −0.18∆RAIN*** + 0.13∆RAIN(−1)***	2.01	1.12	15.67	0.04
ChongQing: ∆GDP = − 0.06ECT(−1)*** + 2.24∆FDI** − 2.16∆FDI(−1)** + 0.16∆EXP** − 0.56∆EXP(−1)*** + 0.192	17.91	1.70	2.82	0.84
GuiZhou: ∆GDP = − 0.23ECT(−1)*** − 3.85∆RAIN** + 6.93∆RAIN(−1)** + 1.38***	2.10	0.92	3.92	1.69
Tibet: ∆GDP = − 0.01ECT(−1)*** + 1.22∆RAIN + 0.0001∆RAIN(−1)***	1.04	2.27	0.96	0.11
ShaanXi: ∆GDP = −0.09ECT(−1)*** − 432.63***	5.70	1.18	0.81	1.86
QingHai: ∆GDP = −0.19ECT(−1)*** −5.86∆FDI**+9.12∆FDI (−1)*-1.52	0.79	1.49	1.22	0.16
XinJiang: ∆GDP = − 0.09ECT(−1)*** + 0.51∆GDP(−1)*** − 11.85∆FDI**-39.60∆FDI (−1)*** + 2.14∆EXP***-0.06∆RAIN*** − 0.08∆RAIN(−1)***	2.74	0.88	0.03	0.90
GuangXi: ∆GDP = − 0.47ECT(−1)*** + 15.39∆FDI*** − 88.43***	5.35	2.42	1.28	0.25

Table 4. Long-run estimates of ARDL [DV: GDP]

West

Sichuan

Chongqing

Guizhou

Tibet

Shaanxi

Qinghai

Xinjiang

Guangxi

Constant

−1.76 [−7.43]*

−2.67 [−2.27]*

2.39 [1.89]**

1.58 [3.73]**

0.15 [11.49]***

−8.84 [1.26]

3.97 [0.86]

1.32 [0.53]

3.49 [0.13]

2.26 [2.32]**

3.97** [2.12]

3.19 [2.70]

7.80 [2.17]**

−3.72 [−2.24]**

1.87 [1.85]**

−7.78 [−0.59]

FDI

−9.31 [−2.36]*

−10.04 [−1.49]

8.27 [1.23]

4.33 [2.19]**

0.07 [0.85]

−1.47 [0.32]

−0.6 [−0.13]

11.02 [6.73]***

6.77 [0.05]

−1.89 [−1.71]

−3.99*** [−4.07]

−2.56 [−2.73]*

−1.69 [0.68]

3.75 [2.14]**

−2.81 [−0.35]

−3.77 [−1.04]

EXP

5.47 [4.09]*

7.67 [4.31]*

1.55 [0.88]

0.67 [0.65]

0.002 [0.86]

1.07 [0.57]

−0.03 [−0.16]

−4.57 [−5.48]***

1.28 [1.27]

3.65 [2.70]*

−3.75 [−0.93]

−3.03 [−1.13]

5.33 [0.21]

−5.99 [−1.42]

6.37 [1.29]***

6.47 [2.37]**

TEMP

-

1.67 [2.37]*

-

−1.44 [−0.00]

-

5.52 [1.26]

-

−1.9 [−0.49]

-

−1.51 [−2.53]**

-

−2.78 [−2.89]*

-

4.69 [2.28]**

-

1.73 [1.02]

RAIN

2.36 [2.97]*

-

−0.26 [−0.33]

-

−0.001 [−1.33]

-

−0.03 [−0.77]*

-

0.77 [0.62]

-

−0.05 [−0.35]*

-

−1.71 [−1.83]*

-

−0.02 [−0.38]

-

Middle

Shanxi

Jilin

Anhui

Henan

Hubei

Constant

−1.36 [−0.94]

7.17 [1.64]

−1.89 [−1.35]

3.27 [2.55]**

7.17 [0.59]

−4.14 [−0.77]

5.38 [1.35]

8.12 [1.66]

2.00 [2.12]**

−1.34 [−0.75]

FDI

9.63 [4.46]***

6.34 [8.48]***

3.57 [1.12]

2.47 [2.03]***

6.08 [2.11]**

3.98 [1.23]

3.36 [4.41]***

2.10 [3.09]***

1.42 [1.74]

−9.35 [−0.25]

EXP

−1.42 [0.09]*

0.47 [0.17]

1.48 [0.94]

2.57 [4.48]***

7.88 [5.25]***

9.03 [5.67]***

−4.58 [1.48]

0.01 [0.00]

6.66 [1.24]

3.88 [1.09]

TEMP

-

7.17 [1.64]

-

−4.75 [−1.91]**

-

4.79 [1.12]

-

−5.76 [−1.58]

-

−1.68 [−0.75]

RAIN

0.89 [2.64]***

-

0.75 [1.11]

-

0.16 [1.25]

-

−0.35 [−0.52]*

-

−0.11 [−2.35]**

-

East

Beijing

Shanghai

Jiangsu

Guangdong

Zhejiang

Fujian

Constant

4.01 [0.79]

−1.38 [0.08]

2.22 [0.19]

1.28 [1.35]

−2.03 [−2.57]**

−1.66 [−0.79]

−1.22 [−0.53]

5.32 [1.04]

−1.07 [−3.64]***

6.62 [0.37]

3.33 [1.85]**

9.15 [0.41]

FDI

1.03 [0.21]

3.72 [1.10]***

2.62 [2.52]**

−5.39 [0.31]

−1.43 [−3.19]***

−4.98 [0.76]

−6.75 [−2.56]**

−77.32 [−1.09]

1.98 [3.19]**

2.67 [0.68]

−3.66 [−2.59]**

−1.60 [−3.00]***

EXP

−2.39 [−3.17]**

−2.96 [−6.81]***

2.55 [1.06]

−0.17 [−0.26]

6.72 [6.56]***

7.73 [0.00]***

4.29 [4.41]***

3.70 [2.01]**

−7.51 [−0.53]

1.37 [0.09]*

1.63 [1.79]**

TEMP

-

−7.21 [−5.88]***

-

−4.36 [−1.39]

-

1.72 [0.94]

-

1.72 [0.94]

-

−4.11 [0.30]

-

−3.84 [−0.36]

RAIN

−3.09 [−3.87]**

-

0.27 [0.23]

-

4.56 [4.94]***

-

4.82 [3.15]***

-

7.81 [0.00]***

-

−1.57 [−1.31]

-

Note: Asterisks *, **, and *** denote significance at 10%, 5%, and 1% critical values, respectively.

In under-wealthy regions, the aggregate macroeconomic productivity is reported to be significant and linearly responsive to temperature, including positive and negative responses (Burke et al., 2015). We noticed that the benefits brought by temperature occurs in Sichuan and Xinjiang, in the west region, whose GDP ranked 16 and 19 in the entire 19 provinces we investigated, much lower than the average level of the whole country. The negative impact of temperature occurs in Shaanxi in the west, Jilin, Qinghai in the middle, whose economic level does not rank top in the whole country, and Beijing in the east. In general, temperature significantly influences the economic growth in underdeveloped China. In other provinces, the climate has little influence on the economy. The finding is consistent with Newell et al. (2021), claiming regions characterized by advanced industrialization and higher levels of development exhibit a greater adaptive capacity to climate change. Economic sectors that are less susceptible to weather and climate fluctuations. As for Beijing, the negative impact of temperature generally remains. As the political and cultural center of China, Beijing with highly dense population and prosperous economy is highly sensitive to climate change. Specifically, policymaker currently encourages the improvement of climate resilience of metropolitan, to strengthen the prevention and response of extreme weather risks, firmly establish safety-oriented policy, improve the awareness of weather forecasting, and build the first line of defense for meteorological disaster prevention and mitigation. Like Beijing, the capital of China with large population, is necessary to improve its climate resilience ability.

From Figure 2, Guangdong ranks first in the whole country in economical amount, which enjoys the highest temperature and precipitation in the whole country and positively significantly correlates with the economy. Why does Guangdong enjoy the benefits of precipitation? Considering the economic structure has been continuously upgraded in Guangdong, enterprises with high water consumption and low output value have been eliminated or transformed. In Guangdong, the application of new water-saving technologies is implemented, resulting in continuous improvement in the efficiency of water consumption. Guangdong has made elimination and transformation for enterprises with high water consumption and low output value, encouraging the application of new water-saving technologies, resulting in continuous improvement in efficiency of water consumption by enterprises.

Figure 2. Ranking of average GDP of each province in the year 1997 – 2018 in China (in billion yuan).

Source: Chinese National Bureau of Statistics of 1997–2018

In contrast, Guangdong province has emerged as a frontrunner in implementing China’s most stringent water resources management system. This comprehensive system entails the rigorous oversight of the “three red lines,” encompassing measures to control total water consumption, enhance water use efficiency, and limit pollution within designated water function zones. Notably, Guangdong province has effectively instituted assessments at the provincial, municipal, and county levels, ensuring the enforcement of this rigorous water resources management system while also cascading the responsibility for water conservation to local jurisdictions. Provincial authorities and 78% of provincial institutions were built as water-saving units. The concept of water conservation continued to grow in popularity.

Policy should be promoted to the middle and west. The primary consideration of middle and western provinces relies on how to increase the precipitation benefits and get high-quality usage of the water resources. An observable pattern is evident in the arid and semi-arid regions of Northwest China during the period from 1979 to 2019, whereby there has been a substantial increase in average annual precipitation. Projections based on the impact of global warming indicate a further anticipated rise in precipitation levels in Northwest China from 2015 to 2100 (Zhang et al., 2022). The government could only do something manually, considering global climate change. In order to maximize the benefits of precipitation usage, the policy must constantly optimize the structure and improve the efficiency of rainfall use, shift the way from sloppy to economical and intensive, and bring positive economic benefits of precipitation. Firstly, it entails conducting comprehensive assessments to unlock the water resources potential of the Northwest Arid Zone, establishing a solid groundwork for water storage, implementing water transfer mechanisms to augment supplies, and exploring strategies to enhance water availability and optimize the overall water security framework. Secondly, for water storage, it is imperative to reinforce the development of crucial control water conservancy projects within mountainous regions, augmenting water resources’ storage and safeguarding capabilities. This comprehensive approach will effectively tackle regional and seasonal water deficits and address structural constraints, ensuring a sustainable water supply. Thirdly, as for water conservation, the strategy will remain on stimulating technological innovation, enhancing the efficiency of water resource production, optimizing the industrial water utilization framework, and robustly advocating for efficient water conservation techniques in the agricultural sector. Fourthly, for water transfer, the policy utilizes the concept of “spatially balanced” water management, where we will expedite research efforts about inter-basin and trans-regional water transfer. These initiatives aim to surmount the prevailing impediments arising from resource-based water scarcity. Last, proactive measures will be undertaken to undertake extensive investigations on artificially influenced weather phenomena for water augmentation. Furthermore, policy tends to implement artificial water augmentation techniques within mountainous regions, thereby augmenting mountainous watersheds’ water storage and supply capabilities. For temperature, utilizing the benefits of “Laffer curve”, balancing the average temperature below 24.9 Celsius degree to keep the qualified condition economy activities (Paul Alagidede & Frimpong, 2016).

The economic development of Xinjiang and Sichuan provinces benefits from increasing temperatures. Despite their respective economic sizes of 67 billion and 13 billion dollars, lower than the national average of 195 billion dollars, these provinces serve as prominent tourist destinations during high temperatures, particularly during summer vacations, leading to increased local consumption. As economic growth progresses and living standards improve, high-temperature consumption gradually evolves and displays a diversified pattern. Xinjiang is a hot region in China. It has taken advantage of its unique climatic resources to set up an automotive exposure test site to test the thermal safety performance of cars. In addition, there are seven categories of auto parts, chemical materials, plastic products, rubber products, and 25,000 other kinds of parts and materials. Get here to test exposure to the sun. These productions have brought hundreds of millions of dollars of thermal economic effects to Xinjiang. Crops planting in Sichuan and Xinjiang promotes economic growth and gives full play to the positive impact of temperature on crops, and policy should promote the development of local agriculture. The results are consistent with what we have discussed in the temperature impact in literature review. Empirical findings indicate that hot temperatures have statistically significant adverse effects on the GDP levels of poor countries and agricultural sectors (Adam & Drakos, 2022; Sequeira et al., 2018). In contrast, rich countries and non-agricultural production appear to be less affected.

5. Conclusion

We contribute to existing research in the following way: Employing ARDL model, we establish a solid theoretical foundation for selecting regression model. Utilizing extensive data set encompassing economic across sub-provincial regions with nearly nationwide coverage, we conduct empirical examinations of the relationship between climate and economy across different time spans. Our discoveries can enhance the formal depiction of climate effects in individual provinces and diverse regions within a country. Precipitation contributes to economic growth in developed provinces in the east, and the negative influence of precipitation significantly occurs in the west region. Positive effects of temperature tend to occur in western regions.

Many studies focus on projecting assessments of climate-related losses in China, as these evaluations are essential in assessing climate policies. For instance, employing a non-linear framework incorporating the historical interactions between climate and the economy, estimations indicate that China’s average climate-induced damage could reach 4.23 percent of its GDP by the year 2100 (Duan et al., 2022). Before studying the losses caused by climate, we analyze the impact of climate on regional economy, so as to lay a foundation for policy to make better use of climate characteristics and avoid greater losses.

We reassess this empirical issue province by province. This study points out that precipitation contributes to economic growth in developed provinces in China, which reach a GDP of more than USD300 billion on average. The advantages of precipitation predominantly manifest in developed provinces located in eastern region of China. Conversely, the positive effects of temperature tend to occur in western regions. The positive impact of temperature on economy is more likely to be observed in western provinces with lower annual temperatures, less precipitation, and GDP below the average level. The negative effect of precipitation occurs in provinces where the average temperature is below the national average of 14.8°C. Therefore, the reduction to economy caused by precipitation happens in the sample of cooler provinces. Finally, with mixed results in different provinces, the provinces with non-significant results of climate factors have greater capacity to adapt to the climate.

This paper also shows the new results that economic ranking and local precipitation rankings of provinces are generally consistent. The provinces with positive impact of precipitation all have economies of over 300 billion dollars and precipitation exceeds the average level of the whole country; The provinces with economy below the national average also have precipitation below the national average. To summarize, there is a great deal of similarity in how provinces rank in terms of temperature, precipitation, and economy concerning the national average. Developed regions exhibit less susceptibility to temperature fluctuations, whereas underdeveloped areas are more susceptible.

Precipitation is more correlated with the economy in developed regions. The relationship between temperature and economic levels follows an inverted U-shaped pattern, with none of the provinces in China reaching a peak average temperature of 24.9 Celsius degrees, according to Figure 1, thereby maintaining a positive correlation between temperature and economy. The negative influence of precipitation on the economy significantly occurs in the west region, where precipitation is below the national average level, seriously lacking water compared with the east. To avoid damages less precipitation brings, the policy maker should establish policies, strategies, or programs to make good use of climate characteristics in east.

Our study has important policy implications. Policymakers must keep in mind to utilize scientific water management for eastern regions. The 20th Party Congress report of China highlights the objective of fostering developmental framework in the western region while concurrently facilitating the integrated management of water resources, water environment, and water ecology. In order to guarantee sustainable and eco-friendly progress within the Northwest Arid Zone, the problem should be systematically addressed.

Besides the water policy suggestion above, local economy development implements an adaptation-based strategy with a targeted approach to climate. In particular, the western region is rich in energy resources. It has many traditional high-carbon economies. Economic transformation to green and low carbon is an inevitable choice. The western region vigorously develops wind power and photovoltaic industries, which drives regional economic development while positively contributing to reducing global greenhouse gas emissions and controlling the rate of warming. By implementing appropriate policies, the government aims to get high-quality usage of climate conditions in the western and middle regions. We reassess the empirical issue taking into account a neglected important issue: the global warming trend. Under the warming background, it is necessary for policy maker to get utilization of different characteristics of precipitation and temperature in different provinces.

These policies are intended to foster the economic development, attract foreign direct investments, export, and technological advancements, enhance meteorological capabilities, promote a robust open economy, and foster increased interaction and cooperation between different regions. The government actively encourages west and middle regions to establish strong connections with prominent eastern districts, thus fostering a new developmental framework and facilitating high-quality progress.

Given that this analysis has showcased the influence of precipitation and temperature on GDP in both modeling and sampling, it implies the need for additional research to enhance our comprehension of the climate-economy connection, especially when examining specific economic sectors in detail. Future study could involve annual precipitation and temperature data to investigate variances in sensitivity across seasons or months of the year, as well as exploring potential non-linear effects associated with daily temperature fluctuations. Also, the limitation of the study focuses on other climate measurements, for instance, solar radiation, light length, humidity, or sea level. Therefore, readers can better understand the natural climate and economy region variation and give suggestions to better policy to government of China.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1. Provinces under Eastern region are Shanghai, Beijing, Guangdong, Jiangsu, Zhejiang, Fujian. Provinces under Middle region are Jilin, Shanxi, Henan, Hubei, and Anhui. Provinces under Western region are Guangxi, Qinghai, Sichuan, Guizhou, Chongqing, Shanxi, Xinjiang, and Tibet.