For most people, the chance they are a victim of theft (VOT) in year 1 is correlated to the chance they are a VOT in year 2. So the probability that they are a VOT at least once in those two years is NOT simply (1 - (1 - 2%)^2). That formula only works when the two events are independent, like two coin flips.
As an obviously extreme example, imagine a world where 98% the people live in zero-crime areas, and the rest live in places where they are robbed annually.
In such a world, the percentage of people who were a VOT in a single year would be 2%, and it would not rise as you broadened to multiple years. (The same 2% of people would be targeted over and over.)
This is all just a roundabout way of stating the unfortunate fact that some people live in bad areas.
As an obviously extreme example, imagine a world where 98% the people live in zero-crime areas, and the rest live in places where they are robbed annually.
In such a world, the percentage of people who were a VOT in a single year would be 2%, and it would not rise as you broadened to multiple years. (The same 2% of people would be targeted over and over.)
This is all just a roundabout way of stating the unfortunate fact that some people live in bad areas.
I'm sorry to hear about your experience.