Edit: Ok, figured out what the issue is.
The crime rates I referenced â actual reports to police â are much smaller, oddly enough, than the crime rates you reference, which are peopleâs self-reporting of crime to government surveyers.
So in 1973, the FBI estimated that police received reports of violent crimes that were 417.40 per 100,000. What the graph you posted represents is that in the NCVS survey of 90,000 random households that people told researchers that crime rates were actually 10 times higher than that. In 1973, the NCVS reported the violent crime rate was 4,770 per 100,000.
So what we learned in 1973, if that graph is to be believed, is not that people mistakenly thought the rate of violent crime was much higher than it was, but rather that most of the violent crimes in the country were going unreported and, if anything, official counts prior to 1973 were drastically underestimating the level of violence in America.
But the basic claim still holdsâŚas both the UCR and NCVS counts of violent crime have declined in the U.S., so has support for the death penalty.
Original post:
No, sorry, not false at all The issue seems to be how we measure crime and, oddly enough, may be opposite of your claim that:
âsupport for the death penalty is frequently based on popular perceptions of reality rather than actual facts.â
There are two ways we tend to measure crime rates: 1) ask police how many people they actually arrested last year for a given crime; and 2) survey households randomly and ask them whether or not they were the victim of a crime.
In the case of violent crime from 1960-1991, the number of violent crimes the FBI reported that were actually reported to police through the UCR increased four-fold, from 160.9 per 100,000 in 1960 to 758.2 per 100,000 in 1991.
See: http://www.albany.edu/sourcebook/pdf/t31062004.pdf
The chart above that you posted appears to be based on the NCVS surveyâŚit begins in 1973, which is also when the NCVS began. The NCVS randomly calls 90,000 households and asks respondents questions about crime victimization.
There are many differences in methodology including the fact that the NCVS doesnât include crimes of violence against children under the age of 12 in its crime estimates while the UCR does.
Thereâs a fascinating look at the differences here: https://www2.fbi.gov/ucr/cius2009/about/crime_measures.html