An excellent discussion of the misuse of statistics in current political discourse. This book is essential reading for anyone exposed to current political rhetoric which, of course, is everyone. As an added bonus the book is short. This means anyone who reads it will quickly be able to recognize how ubiquitous misleading statements are. Indeed, misleading statements seem to be the rule rather than the exception these days. After reading this book you will likely think "I knew the situation was bad but it is even worse than I imagined!" Indeed, one even gets to the point of looking at a statement by the media or a politician and thinking "what is the worst case scenario that would make the statement technically true when the reality is close to the opposite of the impression trying to be conveyed."

Sowell is excellent at providing examples to show common problems in the presentation of statistics. For instance, crime statistics are often presented as "black people make up 13% of the population but are many times more than 13% of those stopped/arrested/incarcerated. The shows racism at work.". Sowell compares this to concluding that since NBA referees call fouls on black players much more than 13% of the time the reason must be racial prejudice. Sowell shows that it is not nearly that simple. For example the average age of the populations matters since younger people are more likely to engage in crime than older ones. The amount of crime actually being committed matters as well. Sowell mentions, as an example, experiments with radar cameras showing the black people, in fact, are more likely to speed.

Sowell's writing is not only concise but also very clear and well organized. For example, to head off any ambiguity he clearly distinguishes three types of discrimination and labels them separately. One type of discrimination is good: discernment between individuals based on their individual characteristics and not race. One type is the bad kind: prejudicial attitudes toward a group based on false beliefs about them. A third type is in between: attitudes towards a group based on things statistically true about some but not all members. Sowell uses the example of criminal background checks to show how not allowing the good type of discrimination leads to some unexpected consequences. Employers will then use the statistical type of discrimination and end up hiring fewer members of the group than if background checks are allowed. Think disparate impact laws can be added to prevent that? Employers will just locate to places where there are fewer members of the group. Similarly, Sowell shows how the presence of a particularly problematic subset of a group can cause businesses to increase the price of their product which is then misinterpreted as a "tax" on the all members of the group including those who do not engage in the behavior. The business may even decide not to do business in certain areas at all. This is highly timely given all the businesses shutting down recently due to increased shoplifting and "smash and grab" attacks.

This books contains bits of history which caused me to rethink others books I have read. For example, black incomes did rise after the Civil War until 1900 faster than those of whites. This is not something you are going to hear about in books which treat post-Reconstruction as a virtual reset to slavery. For all its statistical detail The Republic for Which it Stands did not discuss it. Neither did The Black Tax which explicitly counts the period from Reconstruction until the Civil Rights era as equivalent to another 100 years of slavery when figuring how big "the black tax" has been over the years. Sowell also mentions that school desegregation was already happening in the North, was welcome and was being enforced by courts until southern black immigrants who were more prone to crime arrived.

Sowell also mentions the rather remarkable phenomena of survey data being used as a key component in decrying wealth inequality. For example asking lower paid worker how many hours they work compared to higher paid workers and treating lower paid workers' self reporting as fact, which then turned out to be false. Although Sowell does not mention it, self reporting statistics are often key in claims about disparate rates of arrest and incarceration during the war of drugs.

Beyond racial disparities there is also a good discussion of how statistics regarding income disparities involving "the bottom 20%", "the top 20%" or the "the top 1%" are highly misleading given how fluid who is in these groups is.

I could only find a few minor flaws with the book. For example when comparing the United States to other nations when calling Marxists' claims into question Sowell says that other nations are not doing as well as the United States despite all the capitalists there. Any Marxist worth his salt would, of course, instantly point out that perhaps the United States is doing so well because of its capitalists exploitation of the globe. Similarly, attacking the Soviet Union as a Marxist state is a straw-man. Just because the Soviet Union said they were Marxists does not mean they actually were. Finally, Sowell's belief is that the rise in crime starting in the 1960's was due to leftist ideologies being more tolerant of undesirable types of behavior. This is possible but the consensus is that the exact causes are not well known other than there being an explosion of young people, those most likely to be involved in crime, due to the baby boom.

Brand	N/A
Rating	4.9 (4177 ratings)
Price	$16.99
Category	Accounting

Discrimination and Disparities