People always have a very hard time believing crime is decreasing where they live. For example, I'm sure I could convince you that crime is decreasing in Russia

People always have a very hard time believing crime is decreasing where they live. For example, I'm sure I could convince you that crime is decreasing in Russia, Poland, Scandinavia, Germany, Canada or even New York but I'll never ever get you to believe it is decreasing in San Francisco! :) Similarly, a Londoner would never believe their city's crime rate is decreasing but easily accepts that San Francisco's is. There are a lot of criminologists in the United States and they will all agree on that violent crime is much less than what it used to be, with some parts of the country (Colorado, for example...) bucking the trend. The idea that they would all have been fooled by manipulated data is absurd. For example, yes it is possible that some police departments has a tendency to under-report crime. Criminologists aren't dumb and they know about such tendencies. So to counter that they measure crime rates using other statistics too. One example is victim surveys in which they ask X number of people if they have been the victim of a crime. If the number of respondents answering yes decreases then it is plausible that the crime rate has decreased. Another is hospital surveys in which they ask hospitals how many patients they have treated for being punched in the face. If the hospitals report that fewer patients have been treated for getting punched in the face than before, then it is plausible that the number of people getting punched in the face has decreased. Putting enough of these statistics together lets you conclude that, yes, violent crime definitely has decreased. For a better explanation you have to ask a real criminologist, which I'm not.