Part of the survey response you are seeing is that while crime itself may be down, citizens are uncomfortable and feel less safe (I provided an example that you can come and see for yourself any time of day) and I believe that helps to explain the discrepancy in the data you are seeing. How do you account for increases in crime when someone runs a red light (that's a crime) that simply is not reported? What am I going to do, call the local police? Are they going to actually do anything?
Data is only just one piece of the puzzle. Arguably not the most important in many circumstances. If the data is telling you crime is down, but people feel as though it is not, that doesn't necessarily mean "right-wing propaganda", and for someone who is inclined as you are to look at data you would think you wouldn't immediately draw such a conclusion and instead you would want to question your data collection process, the methods you are using, whether or not specific groups have stories they want to tell.
> You're not really talking about actual crime--you're talking about your own perception/feelings.
> By all measurements and all sources of data available, violent crime has been absolutely plummeting since around 1991. Property crime is plummeting in direct parallel.
> Yet, nearly every year, a large majority of people (~66% or so) who are polled believe that crime is higher this year than it was last year.
Maybe the majority is right and your data is wrong?