I have a time series data on shipping accident and offshore drilling accident economic losses as X1 and X2 respectively, as independent variables and Y = GDP of the maritime sub-sector as dependent variable.
The ideal way to do this would be with a general equilibrium model, but that's quite costly. Second-best would be to use an input-output model. Take a look at this article on the impact of occupational injury on the economy. Your regression would have far too much noise to be credible. The Impact of Occupational Injury Reduction on the U.S. Economy, E Zaloshnja, T Miller, G Waehrer, American Journal of Industrial Medicine, 49:9, 719-727, 2006.