Which variables should I include/exclude in my regression analysis for GDP?

Question

I am wanting to establish the impact mobile money has on Kenyan economic growth. I am building my regression model and have collected data for GDP, consumption, government spending, investment, net exports, mobile money agents, accounts and volume of transfers as my independent variables, labour and capital acting as moderating. My regression results are showing high levels of multicollinearity and an almost perfect R-squared. I am trying to figure out what variables to exclude from my model, but am unsure how to go about this as literature suggests strong explanatory power of the independent variables for GDP. I wondered if anyone was able to share some information about variable selection for a linear regression?