Policy chance analysis - time series count data


I need some help to determine my approach.

For a school project I would like to examine the effect of a policy change in the Danish registration law for certain cars, happening in june 2007. I have a dataset with the number of registrations pr. month from jan-97 to dec-15.

If I put the data in a graph it is easy to see the policy change had a huge effect. But my question is:

How should I approach performing a statistical analysis of the change?