Misattributed quotes can still be valid, and such is the case with the often repeated, “There are three kinds of lies: lies, damned lies, and statistics,” typically associated with Mark Twain (possibly the first person to attribute the saying to the wrong person).
Statistics are a powerful kind of lie because data allow people to state factually true statistics that still mislead or distort the topic being addressed.
Republicans and conservatives have used the statistical lie often as a dog whistle for their racist base. Two of those issues are fatal police shootings and black-on-black crime.
Let’s look at how these statistical lies work.
Republicans and conservatives are apt to note the raw numbers on fatal police shootings broken down by race:
Yes, police shoot and kill more white people per year than Black people, but a statistical fact of this data is that there are about 5 times as many white people in the U.S. as there are Black people; therefore, for this data set to be equitable, about 5 times more white people would be killed than Black people (note that the difference is only about twice as many).
Thus, a better statistic is the rate of fatal police killings by race:
Therefore, fatal police shootings are racially imbalanced (Black people shot and killed at about 2.5 times higher rate than white people), if not racist.
In the case of fatal police shootings, then, the raw data are both accurate and misleading when trying to understand racial inequity.
A much more insidious use of statistics is the overuse of black-on-black crime in media, public, and political discourse.
Black-on-black crime rates are extremely high, often at a 90%+ rate.
But there is almost no media, public, or political rhetoric around the white-on-white crime rate, which is about statistically the same (high 80% rate). 
Crime rates are almost entirely within races in the U.S. (see p. 13 from the U.S. Department of Justice ) because the country is still strongly racially and economically stratified.
While highlighting the very high black-on-black crime rate is factually correct, omitting that most crime is intraracial makes that emphasis misleading, and another dog whistle for racists.
But Republicans aren’t stopping there; consider the Lt. Governor of Texas who has now blamed Black Texans for being unvaccinated and causing the newest Covid spike:
However, as you may suspect, there are problems with this claim:
Once again, Republicans are using the statistical lie as a dog whistle for racist constituents.
Many racial groups are under-vaccinated, and there certainly is a significant issue with vaccine hesitancy and resistance among Black Americans, but the sheer numbers in Texas make Patrick’s careless and racist claim more than preposterous.
Further, raw data on low vaccination rates among races also ignore causes for those rates. Black Americans are disproportionately poor and live in areas were vaccine access has been weak or even suppressed.
There is ample evidence that political leaders have always cherry-picked statistics and data to promote agendas, but there is also ample evidence that Republicans target statistics as part of their larger strategy to court their racist base.
Patrick’s most recent egregious use of the statistical lie is further proof that Trump did not create the Republican Party as a party of lies, but he certainly helped the strategy gain momentum.
 See data here: