First, I set up the relationship between rainfall and political party foundings, and then modify the rainfall measure to generate a statistically independent measure for droughts.This modification will allow us to isolate the effect of economic strain from other underlying features of a region.
- Show the visual relationship between the level of rainfall in a district in the period leading up to the current election, and the number of political parties that are founded in a region
- Prove that the raw level of rainfall are not independent from one election period to the next within a district, as well as from neighboring districts from one election period to the next (plm model: Panel Linear Models)
- Create a measure that sums the number of years a district experiences either moderate droughts or floods during the interval starting from the year following the previous election up until the year of the current election (pglm model: Panel Generalized Linear Model)
Next, I analyze whether there are more new political parties when droughts or floods occur, by running a pglm regression model predicting the number of new political parties that are formed as a function of the number of years a district experiences droughts or flooding in the interval starting from the year following the previous election up until the year of the current election.
The results show that extreme weathers have significantly impact on the new polical parties founding. Besides,I found that farming party is more likely to be formed when a district experiences extreme weather.
3. Relationship between extreme weather in one district with political activity in another district.
In this part, I looked at how political activity stimulated by droughts or floods in one district might affect political activity in another district. (pglm model)
The results show that even when taking into account a district’s own droughts and floods, that district’s degree of political foundings will also depend on the number of years its neighboring districts experience years of droughts or flooding in the interval starting from the year following two elections ago, up until the year of the previous election—the election lead-up interval before the current one.
Extreme weather events like droughts or floods can erode the stability of political systems and wear away at the entrenched power bases of large, national-scale parties that have difficulty responding to the needs of affected regions. Thus, here I wonder whether experiencing droughts or floods relate to political concentration. I performed a pglm regression predicting the Herfindahl Index of a region as a function of the number of years of droughts or flooding that occur in a district in the interval leading up to the current election, and the number of years of droughts or flooding that occur in its neighboring districts in the interval leading up to the previous election.
The results show that political concentration are not independent on the number of years of droughts or flooding that occur in a district in the interval leading up to the current election, and the number of years of droughts or flooding that occur in its neighboring districts in the interval leading up to the previous election.(both variables are significant). But the current period's extreme weather situation in the district has negative impact on the concerntration, while the last period's situation in its neighbour districts has positive impact on the concerntration.
Political parties are formed to accomplish a variety of goals. Individual parties can also exist in the context of larger social and cultural trends, especially when regions influence each other as political organizing activity diffuses across regions over time. To understand the diffusion process more, I analyze whether the new parties that appear in a district are the same parties that have appeared in neighboring districts in the past, or if it the process of political organization, rather than the content of a specific political party, that is diffusing.(pglm model)
The results show that extreme weather situation variables are not significantly effecting the diffusion of political parties in neighbour district. While, the extreme weather events of neighbour during last period do have positive impact on non-contest new party formed. It may illustrate that diffusion of political organizing is dependent on extreme weather situation in a district and its neighbours.