Writing about the whole race and realignment issue, Ross says the key thing to keep in mind is that "Publius is right: 'Nixon’s law and order message was lost on no one.' It was lost on no one because violent crime went up three hundred and sixty-seven percent between 1960 and 1980." Silly Publius, silly liberals.
But wait, Nixon was elected in 1968, not 1980. Meanwhile, the population grew between 1960 and 1980. And the overall violent crime index is unreliable thanks to the vagaries of classification standards. But lets look at the homicide rate which works for comparisons over time. We see that the murder rate went up a good deal from 5.1 per 100,000 people in 1960 to 6.9 per 100,000 in 1968, at which point Nixon's "law and order" message captured the hearts of America's silent majority. But as you'll see from the above chart, there's no clear further correlation between either the absolute level of crime or the trend in crime and partisan or incumbent party success. Nixon wasn't punished in 1972 for the fact that, despite his law and order platform, crime kept going up. Crime was way higher in 1984 than it was in 1968. And so on and so forth.
Along these lines, if a move to the right was really the consequence of rising crime rates, one would expect the most conservative groups in the electorate to be those most afflicted by violent crime -- low-income African-Americans. But of course that's not how it works at all. Thus while "crime" and "law and order" were obviously successful electoral themes for the Republican Party in the 1970s and 80s there's little indication that their utility as messages was tightly linked to the objective facts about crime. After all, if the appeal of "crime" messaging was really about crime, its effectiveness should have diminished in years (1972, 1988, and to some extent 1984) when GOP leadership failed to address the issue and/or its effectiveness should have been correlated with the audience's risk of being victimized by crime, but neither is true.