Original post by Thomas Sowell.
Highlights that Mark Perry parsed out from the article:
1. Empirical studies that follow income brackets over time repeatedly reach opposite conclusions from studies that follow individuals over time. But people in the media, in politics and even in academia cite statistics about income brackets as if they are discussing what happens to actual human beings over time.
2. All too often when liberals cite statistics, they forget that correlation is not causation. For example, the New York Times crusaded for government-provided prenatal care, citing the fact that black mothers had prenatal care less often than white mothers — and that there were higher rates of infant mortality among blacks. But American women of Chinese, Japanese and Filipino ancestry also had less prenatal care than whites — and lower rates of infant mortality than either blacks or whites.
3. When statistics showed that black applicants for conventional mortgage loans were turned down at twice the rate for white applicants, the media went ballistic crying racial discrimination. But whites were turned down almost twice as often as Asian Americans — and no one thinks that is racial discrimination.
Bottom Line: Facts are not liberals’ strong suit. Rhetoric is.
Remember that if statistics can be interpreted more than one way, the person doing the interpreting is going to highlight the statistics that fit their agenda/confirmation bias. If the data doesn’t support your claim…discount the data as bad OR scream louder than your opponent. Bottom line, think critically about what you read and hear.