Assessing the effect of panel threshold on relief expenditure,disaster losses and economic growth

In this paper,we build three regression model to analyze the relationship between relief expenditure,disaster losses and economic growth by using panel data from 31 provinces in china,year from 1990 to 2011. the basic hypothesis as: (1) we does not consider the panel regression variable relief expenditure and disaster loss; (2) we consider relief expenditure and disaster loss,but not threshold variable; (3) we consider relief expenditure and disaster loss as a threshold variable respectively. on this basis,the paper studies the relationship between relief expenditure,disaster loss and economic growth. results show that there is double threshold effect between relief expenditure,disaster loss and economic growth in china. the relief expenditure will have negative effect on economic growth in two stages,one is when the relief expenditure lowers than 5.05 million,and another is between 5.05 million and 16.26 million. also,the relief expenditure will promote economic growth when relief expenditure is higher than 16.26 million. in addition,when the provincial actual disaster loss is more than 56.09 million,disaster losses have a positive effect on economic growth,and this effect will be weaken gradually when the disaster loss is more than 99.48 million. in conclusion,sustained disaster losses will have a significant impact on the economy,so the government must take relevant measures to avoid the ―creative destruction. © 2014 pakistan journal of statistics