As part of my University research project, I have decided to study the impact of Immigration on house prices but find that the immigration Variable that should be statistically significant according to literature appears to be insignificant. I don't know whether this was down to the method I used to produce a the immigration variable or because of other factors.
Variable that should appear significant is insignificant
- Thread starter karan_089
- Start date
- Tags regression analysis regression coefficent significance significance level time series analysis
Hello Obh,
I am testing the impact of immigration on house prices in London and have collected data on the following variables:
As per the literature the model I have to:
I have tried to proxy the data as close as to the literature but was unable to access special license data from the ONS. For variables like Immigration and Regional GDP I had to use interpolation as this data was only available of an annual basis- my supervisor suggested that I should not use panel data and rather time series because of the complexities in the investigation were beyond what was required. Also one of my peers had told me that Stationary not being present could be a reason significant variables are appearing insignificant.
I have attached the data via a google drive link as it has only allowed me to upload a PDF.
https://drive.google.com/file/d/1R8AwiKVvfvzvQxD7OOtlYw9SYndEJik7/view?usp=sharing
thanks
I am testing the impact of immigration on house prices in London and have collected data on the following variables:
- House price- HPI
- Immigration - Long term immigration flow
- Interest rates - Bank of England base rate
- Market capitalization- S&P equity index
- Unemployment rate- ONS Unemployment rate
- Regional GDP- ONS regional gdp estimates
- Total crime- Crime data Via Home office
- Benefit Claim- ONS Claimant Count
- Population- ONS London population estimate.
As per the literature the model I have to:
- take the difference of log of HPI
- calculate the % change in immigration by doing (IMGquater 2- IMG quarter 1)/ POPULATION quater 1
- all other variables enter with a lag of one.
I have tried to proxy the data as close as to the literature but was unable to access special license data from the ONS. For variables like Immigration and Regional GDP I had to use interpolation as this data was only available of an annual basis- my supervisor suggested that I should not use panel data and rather time series because of the complexities in the investigation were beyond what was required. Also one of my peers had told me that Stationary not being present could be a reason significant variables are appearing insignificant.
I have attached the data via a google drive link as it has only allowed me to upload a PDF.
https://drive.google.com/file/d/1R8AwiKVvfvzvQxD7OOtlYw9SYndEJik7/view?usp=sharing
thanks
