


































Advances in Politics and Economics 
ISSN 2576-1382 (Print) ISSN 2576-1390 (Online) 

Vol. 5, No. 4, 2022 

www.scholink.org/ojs/index.php/ape 

20 
 

Original Paper 

Fiscal Decentralization and Natural Disaster Relief in China 

Wang Bo
1
 

1
 School of Economics and Management, Northwest University, Xi’an, Shaanxi Province, China  

 

Received: September 14, 2022     Accepted: October 3, 2022    Online Published: October 8, 2022 

doi:10.22158/ape.v5n4p20             URL: http://dx.doi.org/10.22158/ape.v5n4p20 

 

Abstract 

This paper analyzes the impact of fiscal decentralization on the level of government relief for natural 

disasters using 30 provincial-level data from 1997 to 2017 in China. The results show that fiscal 

decentralization positively impacts local government expenditure on natural disaster relief. Although it 

is different from the existing conclusion that fiscal decentralization inhibits the supply of local 

non-economic public goods, this conclusion can be explained by the literature on disaster prevention 

and control; the incentive of decentralization to increase government revenue increases the opportunity 

cost of disaster prevention and control investment, which makes the pre-disaster funds insufficient, and 

the post-disaster financial expenditure of natural disaster relief increase accordingly. 

Keywords 

Fiscal decentralization, Non-economic public goods, Natural disaster relief 

 

1. Introduction 

In recent years, Andreas and Timo (2008) and Jia et al. (2014) noted that China’s decentralization 

reforms had pushed local governments to favor economic public goods services to promote regional 

economic growth. In addition, China’s tax-sharing reform has created significant incentives for local 

governments to rely on land finance, which further skews local governments toward economic public 

goods services (Ding et al., 2014; Hui Li, 2017). However, as a non-economic public good, the public 

goods of disaster prevention, mitigation, and relief have achieved different results under China’s fiscal 

decentralization system. It has been nearly 30 years since China implemented the tax-sharing reform in 

1994. Apart from the earthquake in Wenchuan County in 2008, the financial expenditure on natural 

disaster relief has reached an all-time high; from 2000 to 2017, the overall Natural disasters in China 

fiscal spending on bailouts showed an upward trend (Figure 1). This makes it necessary to consider 

whether fiscal decentralization impacts public goods of disaster prevention, mitigation, and relief 

differently than general public goods. 



www.scholink.org/ojs/index.php/ape                 Advances in Politics and Economics               Vol. 5, No. 4, 2022 

21 
Published by SCHOLINK INC. 

 

Figure 1. Spending on Natural Disasters in China from 2000 to 2017 

 

It is now widely accepted that decentralized government is more effective in disaster prevention and 

mitigation than centralized government (Rumbach, 2016) because natural disasters are limited to a 

particular geographical area and rarely affect the whole country. Therefore, for local governments that 

have complete control of regional information, knowledge, and resources, decentralization helps 

prevent and mitigate regional disasters. In addition, numerous studies have shown that countries with 

greater decentralization have higher budgets for disaster prevention and mitigation (Fahlevi et al., 2019) 

and fewer deaths from disasters (Skidmore & Toya, 2013), and the impact is more pronounced in 

developing countries (Escaleras & Register, 2012). In short, according to the existing research, there is 

relatively little literature on the financial expenditure of natural disaster relief. Much research has been 

carried out on fiscal decentralization and natural disaster prevention. It is generally believed that fiscal 

decentralization is conducive to reducing the impact of natural disasters on the economy and society. 

Under the specific condition of government aid standards, the scale of government aid to natural 

disasters is directly related to the impact of economic and social disasters. What is the effect of China’s 

fiscal decentralization on local government disaster prevention, and how does that affect natural 

disaster relief, which is an important topic discussed in this paper? 

 

2. Theoretical Basis and Research Hypothesis 

How local governments use and control the fiscal revenue of decentralization depends on the specific 

characteristics of fiscal decentralization. For China’s tax-sharing reform in 1994, the following two 

specific features: the central government’s fiscal revenue has been increased, giving the central 

government more room for policy adjustment. The central government can use the transfer payment to 

regulate the equalization of regional public services. On the other hand, the core measure of tax-sharing 

reform is the “Tax-sharing system” change. The tax revenue is divided into three parts: state, local, and 



www.scholink.org/ojs/index.php/ape                 Advances in Politics and Economics               Vol. 5, No. 4, 2022 

22 
Published by SCHOLINK INC. 

shared tax. The state tax is obtained by the central government, local governments obtain local taxes, 

and shared taxes are divided between the central and local governments in a particular proportion. At 

the same time, the fiscal responsibility of local governments has not been changed, and local 

governments still bear nearly 80% of public goods expenditures in all regions. However, this heavy 

expenditure responsibility is the political promotion incentive that decides the official promotion by the 

performance appraisal. This fiscal decentralization system incentivizes local governments to develop 

the economy and generate tax revenue. As for the public goods of disaster prevention, reduction, and 

relief, although they are non-economic public goods, they play an essential role in supporting regional 

economic growth. Specifically, China’s financial investment in disaster prevention, mitigation, and 

public relief goods include financial investment in disaster prevention and control and financial 

investment in disaster relief. Pre-disaster disaster prevention and control funding are critical to 

improving the region’s resilience to natural disasters, reducing disaster losses, and achieving 

sustainable economic development (Cohen & Werker, 2008). The financial input of disaster relief after 

the disaster benefits the economic recovery of the disaster-stricken areas. However, under the 

decentralized system, local governments prefer economic public goods services, increasing the 

opportunity cost of disaster prevention and control investment correspondingly increase, thus reducing 

the investment of disaster prevention and control funds. As Burby (2006) puts it, the local government 

paradox is that local officials see natural disasters as a secondary issue and are reluctant to invest in 

disaster mitigation. In addition, most natural disaster relief funds for local governments in China come 

from special transfer payments from the central government, meaning that the state fully ensures 

economic losses caused by natural disasters. Therefore, at the local government level, there is a 

trade-off between the supply of economic public goods and the optimal input of public goods for 

disaster prevention, mitigation, and relief. The change in financial expenditure for natural disaster relief 

will be the result of this trade-off. 

In order to simplify the study, we take a single local government as an example to explore the local 

government’s financial expenditure on the trade-off problem. Local governments provide economic 

and non-economic public goods for their jurisdictions according to the tax and transfer payment 

revenue they receive. Suppose  is an economic public goods expenditure,  is a non-economic 

public goods expenditure, and  is the ratio of economic public goods expenditure to non-economic 

public goods expenditure ( ): 

                           (1) 

Assuming that the central government uniformly determines the national, local, and shared tax rates, 

the statutory tax rate is  ( ). If the regional economic output is  ( ) and the level of 

decentralization of local government revenue is , then the local government revenue sharing is 

. In order to simplify the study, local government debt and other funds outside the system are 

not taken into account; the total fiscal revenue  received by local governments consists of two funds, 



www.scholink.org/ojs/index.php/ape                 Advances in Politics and Economics               Vol. 5, No. 4, 2022 

23 
Published by SCHOLINK INC. 

tax-sharing revenue ( ) and central transfer payment (including general transfer payment  and 

unique transfer payment ), that is: 

 (2) 

Assuming that the local government investment in disaster prevention and control is , it is mainly 

used to invest in the modification of disaster prevention works, special expenses of disaster forecast 

and detection, propaganda of disaster prevention and reduction, personnel training, etc. Suppose the 

ratio of disaster loss to local economic output is . In that case, the ratio of disaster loss ( ) and the 

capital input for disaster prevention and control ( ) shows a reverse decreasing relationship (Zhuo & 

Duan, 2012), that is, , and satisfy , . Suppose that the 

probability of a regional natural disaster is  and that the probability of a disaster occurring in one 

region is independent of the probability of a disaster occurring in another region since most natural 

disasters are inevitable but predictable (Nagasaka, 2008). Therefore, this paper assumes that the 

probability value  of a natural disaster is an exogenous variable and a known constant at the same 

time. Assuming that the initial social capital stock of an area is , the social capital stock affected by 

natural disaster risk  can be expressed as: 

                        (3) 

Suppose that the proportion of natural disaster relief expenditure to non-economic public goods 

expenditure is . The proportion of special funds for disaster prevention and control to total disaster 

prevention and control funds is , the proportion of the special expenditure of economic public goods 

to the total expenditure of economic public goods is  the proportion of the special expenditure of 

other non-economic public goods to the total expenditure of non-economic public goods is , then the 

financial expenditure of natural disaster relief is , and there are: 

                      (4) 

Barro’s (1990) endogenous economic growth model uses a production function that includes both 

private capital and public expenditure in its inputs and is therefore widely adopted (Davoodi & Zou 

(1998)). In this paper, we construct the regional output function with reference to its expression as 

follows: 

                              (5) 

Where  is the economic output of the region, and  is the other factors (such as technical factors) 

that affect the economic output of the region other than economic public goods expenditure and capital 

stock,  reflects the degree of economic public goods expenditure affecting the economic output of 

the region ( ),  reflects the degree of capital stock affecting the economic output of the region 

( ). On the premise of other constant factors, the increase of social capital stock or economic 

public goods expenditure level will promote the local economic output level. 



www.scholink.org/ojs/index.php/ape                 Advances in Politics and Economics               Vol. 5, No. 4, 2022 

24 
Published by SCHOLINK INC. 

For local governments, the goal is to increase economic output and non-economic public goods 

expenditure such as education, health care, and social security to maximize the utility of local 

governments. With reference to Cai & Treisman (2005), the utility function of local government is a 

utility maximization problem as follows: 

 

                      (6) 

Among them,  denotes the importance of regional economic output affecting the utility of local 

government ( ). From the equations (1)(2)(4) and (6): 

                          (7) 

According to the practical significance, the local government’s economic public goods expenditure  

and non-economic public goods expenditure  should be greater than zero, so the following 

inequality holds: 

                           (8) 

Combined (5)(6) and (7) can transform local government utility maximization problems into: 

                (9) 

Formula (9) is the derivative of the investment in disaster prevention and control fund , and partial 

, so that the optimal utility function of local government should satisfy the following condition 

formula: 

                       (10) 

According to Formula (10), when the utility function of local government can be optimized, the 

financial expenditure of natural disaster relief  can be further expressed as: 

                 (11) 

From equations (4) and (8), we know that  and . Therefore, from 

formula (11), we know that the financial expenditure of natural disaster relief  increases with the 

ratio  of economic public goods expenditure to non-economic public goods expenditure. Secondly, 

after considering the input of funds for disaster prevention and control, the more the proportion of 

special funds for disaster prevention and control ( ), the less the proportion of funds for disaster 

prevention and control in general transfer payment is, the financial expenditure of natural disaster relief 

( ) increased. Finally, the degree parameter , which reflects the influence of economic public goods 

expenditure on regional economic output, is positively correlated with the financial expenditure of 

natural disaster relief, that is, the greater the impact of economic public goods expenditure on regional 

economic output, the greater the impact of economic public goods expenditure on regional economic 



www.scholink.org/ojs/index.php/ape                 Advances in Politics and Economics               Vol. 5, No. 4, 2022 

25 
Published by SCHOLINK INC. 

output, so economic public goods fiscal expenditure is more conducive to regional economic growth, 

thus bringing more fiscal revenue for local governments, the overall level of public goods fiscal 

expenditure of local governments. 

Overall, China’s tax-sharing reform makes local governments favor economic public goods supply. 

This financial expenditure bias is further reflected in the public goods of disaster prevention, mitigation, 

and relief. In terms of disaster risk response, while the central government is in the best position for 

political centralization, it is understood in advance for local governments. It takes into account future 

disaster relief actions by the central government; any ex-post adjustment of transfer payments by the 

central government will further distort the regional incentive mechanism and lead to insufficient 

ex-ante investment to further stimulate local governments, the final performance of the disaster 

prevention and control funds to reduce and increase the financial expenditure of natural disaster relief. 

Therefore, we get the following theoretical hypothesis: 

H1: There is a positive correlation between the degree of the financial income distribution ( ) and 

the local government expenditure on natural disaster relief ( ); 

H2: fiscal decentralization reduces local government funding for disaster prevention by affecting local 

government preferences for economic public goods service; this, in turn, leads to an increase in local 

financial expenditure on natural disaster relief ( ). 

 

3. Model Settings, Variable Descriptions, and Data Sources 

3.1 Model Settings 

To test H1, we constructed the following econometric model: 

 

Where  stands for region and  for year.  is the intercept term,  is the region fixed effect,  

is the year fixed effect,  is the residual term. The disease  is the explanatory variable of 

this paper, which represents the financial expenditure of the local government in the natural disaster 

relief of  area in the  year.  is the explanatory variable, representing the share of local 

government revenue in  area in the  year.  was the control variable,  and  were 

the coefficients. 

In order to test H2, we add in the interaction terms to verify that the local government economic public 

goods service bias affects the financial input of disaster prevention and control, and then affects the 

financial expenditure of local natural disaster relief, and construct the model as follows: 

 

The interactive item  is the explanatory variable, in which  is the disaster prevention 



www.scholink.org/ojs/index.php/ape                 Advances in Politics and Economics               Vol. 5, No. 4, 2022 

26 
Published by SCHOLINK INC. 

fund input of local government in  area in  year. 

3.2 Variable Description 

The Financial Expenditure Index of natural disaster relief ( ) is the explained variable in this 

paper. To eliminate the influence of outliers and make the data more stable to weaken the collinearity 

and heteroscedasticity of the model, we take the logarithmic values in the empirical process. 

Fiscal decentralization ( ) is the core explanatory variable in this paper. Based on the theoretical 

analysis, this paper analyzes the impact of fiscal decentralization on the financial expenditure of natural 

disaster relief from the perspective of fiscal decentralization. Therefore, this paper adopts fiscal revenue 

distribution ( ) as the explanatory variable and fiscal expenditure distribution ( ) as the substitute 

variable for the robustness test. 

 

 

This paper uses the ratio of economic public goods expenditure to total public goods expenditure ( ) 

to measure the change in local government’s public goods supply preference. The so-called economic 

public goods service is a public product that can bring direct economic effects to the region’s 

development, including transportation, energy, communication, and other infrastructure. A fundamental 

fact of public investment (ps) in disaster prevention and control is difficult to measure the number of 

funds invested in disaster prevention and control in a region. Therefore, to measure public investment 

performance in disaster prevention and control, existing studies have used indicators such as 

disaster-affected populations or disaster economic loss (Escaleras & Register, 2012; Skidmore & Toya, 

2013). In this paper, the ratio of direct economic loss due to disasters to the GDP of the whole province 

is used to replace the public investment (ps) in disaster prevention and control, and the ratio of 

disaster-affected population to the total population of the whole province (szrk) is used for the 

robustness test. 

Through reviewing the literature, we find that both the density and the urban are considered to be the 

factors that affect the probability of death from natural disasters, and the consumption index will affect 

the level of disaster relief. Given the above three factors will directly affect the results of this study, so 

as a control variable in the empirical model. In addition, Seth (2008) study shows that transportation 

systems are an essential factor in local natural disaster population loss. Therefore, this paper takes the 

area of per capita road (pertraffic), regional passenger traffic level (pass), and other indicators as 

control variables. 

3.3 Data Sources 

Due to the following data limitations, the decision of this study cycle. Since 1997, when Chongqing 

became a municipality, the country has been divided into 34 province-level divisions (four 



www.scholink.org/ojs/index.php/ape                 Advances in Politics and Economics               Vol. 5, No. 4, 2022 

27 
Published by SCHOLINK INC. 

municipalities, 23 provinces, five autonomous regions, and two special administrative regions) and 

long-term stability. Second, China began to implement the tax-sharing system reform in 1994, and by 

1997, has experienced several years of adjustment and improvement, and the system tended to mature. 

Third, China’s natural disaster relief expenditure of the latest figures until 2017. Therefore, we chose 

data for the period 1997-2017. In addition, we exclude Tibet, Taiwan, Hong Kong and Macau, and 

select provincial panel data from 30 regions across China from 1997-2017, taking into account data 

availability and comparability issues. The empirical analysis was conducted using STATA 16.0. The 

data of each variable are derived from the Chinese statistical yearbook, Chinese Financial Statistical 

Yearbook, Chinese Civil Statistical Yearbook, and Wind database. 

 

4. Empirical Results and Their Explanations 

Table 1 shows the regression results of empirical model 1 and empirical model 2. The R 2 of Model 1 

and Model 2 are above 0.5, which shows that the fitting effect of the model is better. From the 

regression results of Model 1, it can be seen that fiscal decentralization has a significant positive effect 

on the growth rate of fiscal expenditure on natural disaster relief, which is significant at the 10% level. 

It shows that the greater the degree of decentralization of fiscal revenue, the greater the degree of local 

government expenditure on natural disaster relief. So suppose that H 1 is supported empirically. From 

the regression results of Model 2, fiscal decentralization has a significant positive impact on the 

proportion of local government economic public goods expenditure, and the statistical results are 

significant at 5% significance level. It shows that fiscal decentralization makes local governments have 

obvious economic expenditure bias. From Model 2(2), it can be seen that the interaction between the 

proportion of economic public goods expenditure and the proportion of disaster economic loss has a 

significant positive effect on the growth rate of financial expenditure of natural disaster relief and the 

statistical results were significant at the 1% significance level. It shows that the economical expenditure 

of local government influences the financial expenditure of natural disaster relief by affecting the 

intensity of disaster prevention and control. 

 

Table 1. Results of Basic Model Regression 

 
Model 1 

Model 2 

 (1) (2) 

 lndiff exp lndiff 

fdr 1.5359* 10.3073**  

 (0.8089) (4.0050)  

ps*exp   0.3054*** 

   (0.0370) 

urban 2.1473*** -14.3907*** 2.3635*** 



www.scholink.org/ojs/index.php/ape                 Advances in Politics and Economics               Vol. 5, No. 4, 2022 

28 
Published by SCHOLINK INC. 

 (0.3320) (1.7989) (0.3317) 

lnconsume 0.4255 8.0352** 0.4578 

 (0.5789) (3.1523) (0.5820) 

lndensity 0.1133** 0.2651 0.0857* 

 (0.0496) (0.2696) (0.0498) 

lnpass -0.2277* 0.1767 -0.2740** 

 (0.1283) (0.6955) (0.1277) 

ps 7.2061***   

 (0.8648)   

pertraffic -0.0497***   

 (0.0143)   

constant 7.1858** -24.3660 8.0726*** 

 (3.0204) (16.4386) (3.0431) 

R-squared 0.553 0.696 0.543 

N 630 630 630 

Time fixed effect Yes Yes Yes 

Area fixed effect Yes Yes Yes 

Note. standard deviation in parentheses, * 0.1 significant * * 0.05 significant * * 0.01 significant. 

 

In order to investigate the robustness of the models, for models 1 and 2, we replace the fiscal revenue 

decentralization ( ) with the fiscal expenditure decentralization ( ) and replace the proportion of 

direct disaster economic loss with the proportion of the disaster-affected population, the other variables 

were left unchanged for regression. The results are shown in Table 2. From the table, we can see that 

the sign of the main variables in Model 1 and Model 2 is changed, and the significance level is 

enhanced, which shows that our empirical model is robust. 

 

Table 2. Robustness Test 

 
Model 1 

Model 2 

 (1) (2) 

 lndiff exp lndiff 

fds 3.1711*** 48.0050***  

 (0.7766) (4.9711)  

szrk*exp   0.0252*** 

   (0.0059) 

urban 0.4652* -12.5055*** 0.5864** 

 (0.2658) (1.6825) (0.2895) 



www.scholink.org/ojs/index.php/ape                 Advances in Politics and Economics               Vol. 5, No. 4, 2022 

29 
Published by SCHOLINK INC. 

lnconsume 0.5925 4.0966 0.4879 

 (0.4452) (2.9703) (0.4763) 

lndensity 0.0778** -0.0624 0.0724* 

 (0.0379) (0.2533) (0.0409) 

lnpass -0.0702 -0.3019 -0.0720 

 (0.0980) (0.6484) (0.1052) 

ps 6.7932***   

 (0.6576)   

pertraffic -0.0145   

 (0.0103)   

constant 3.6482 -30.4747** 6.4882*** 

 (2.3073) (15.3487) (2.4960) 

R-squared 0.625 0.736 0.557 

N 630 630 630 

Time fixed effect Yes Yes Yes 

Area fixed effect Yes Yes Yes 

Note. standard deviation in parentheses, * 0.1 significant * * 0.05 significant * * 0.01 significant. 

 

5. Conclusion 

China’s tax-sharing reform in 1994 made local governments responsible for most of the expenditures, 

including public goods for disaster prevention and relief. The theoretical and empirical evidence shows 

that there is a positive correlation between the share of fiscal revenue in China and the expenditure of 

local governments on natural disaster relief, local Governments’ limited share of fiscal revenue and 

China’s special incentives for political promotion have led to a bias towards economic public goods 

service, which has further led to a reduction in local government funding for disaster prevention and 

control, accordingly, local governments are urged to invest more money in natural disaster relief to help 

the economic recovery of the affected areas. 

 

References 

Andreas, K., & Timo, V. (2008). Fiscal federalism and the composition of public investment in Europe. 

European Journal of Political Economy, 24(3), 562-570. 

https://doi.org/10.1016/j.ejpoleco.2008.06.002 

Barro, R. J. (1990). Government Spending in a Simple Model of Endogenous Growth. Journal of 

Political Economy, University of Chicago Press, 98(5), 103-126. https://doi.org/10.1086/261726 

Burby, & Raymond, J. (2006). Hurricane Katrina and the Paradoxes of Government Disaster Policy: 

Bringing About Wise Government Decisions of Hazardous Areas. Annals of the American 

Academy of Political and Social Sciences, 604, 171-191. 



www.scholink.org/ojs/index.php/ape                 Advances in Politics and Economics               Vol. 5, No. 4, 2022 

30 
Published by SCHOLINK INC. 

https://doi.org/10.1177/0002716205284676 

Cai, Hongbin, & Daniel T. (2005). Does Competition for Capital Discipline Governments? 

Decentralization, Globalization, and Public Policy. American Economic Review, 95(3), 817-830. 

https://doi.org/10.1257/0002828054201314 

Cohen, C., & Werker, E. D. (2008). The Political Economy of “Natural” Disasters. Journal of Conflict 

Resolution, 52, 795-819. https://doi.org/10.1177/0022002708322157 

Davoodi, H., & Zou, Hengfu. (1998). Fiscal Decentralization and Economic Growth: A Cross-Country 

Study. Journal of Urban Economics, 43(2), 244-257. https://doi.org/10.1006/juec.1997.2042 

Ding, C., Yi, N., & Erik, L. (2014). Spending preferences of local officials with off-budget land 

revenues of Chinese cities. China Economic Review, 31, 265-276. 

https://doi.org/10.1016/j.chieco.2014.10.001 

Escaleras, M., & Register, C. A. (2012). Fiscal Decentralization and Natural Hazard Risks. Public 

Choice, 151, 165-183. https://doi.org/10.1007/s11127-010-9740-4 

Fahlevi, H., Indriani, M., & Oktari, R. S. (2019). Is the Indone-sian disaster response budget correlated 

with disaster risk? Jamba-Journal of Disaster Risk Studies. 

https://doi.org/10.4102/jamba.v11i1.759 

Hui, L. (2017). An Empirical Analysis of the Effects of Land-Transfer Revenues on Local 

Governments’ Spending Preferences in China. Lee Kuan Yew School of Public Policy Research 

Paper, 17-04. 

Jia, J. X., Guo, Q. W., & Zhang, J. (2014). Fiscal decentralization and local expenditure policy in China. 

China Economic Review, 28, 107-122. https://doi.org/10.1016/j.chieco.2014.01.002 

Nagasaka, T. (2008). New Problems in the Study of Disaster Prevention Based on Disaster Risk 

Governance. Quarterly Review, 27, 77-92. 

Rumbach A. (2016). Decentralization and small cities To-wards more effective urban disaster 

governance? Habitat International, 35-42. https://doi.org/10.1016/j.habitatint.2015.08.026 

Seth, D. G. (2008). Natural disaster risk analysis for critical infrastructure systems: An approach based 

on statistical learning theory. Reliability Engineering & System Safety, 855-860. 

https://doi.org/10.1016/j.ress.2008.09.003 

Skidmore, M., & Toya, H. (2013). Natural Disaster Impacts and Fiscal Decentralization. Land 

Economics, 89(1), 101-117. https://doi.org/10.3368/le.89.1.101 

Zhuo, Z., & Duan, S. (2012). Investment expenditure on disaster prevention and mitigation, disaster 

control and economic growth: an economic analysis and China’s empirical study. World of 

management, 04, 1-8. 


