> here's an even bigger problem to worry about: the government believes statisticians
I think this is an oversimplification, and a dangerous one. These statistics are biased, one-sided and misleading. That is indeed a property of some or even most statistics, just as "being wrong" is a property of some or even most information. Yet no-one would think to proclaim that the problem with government is that it believes information.
Real statistics, like real science, is a force for good. Bad statistics, like pseudoscience, is a force for bad - but it can only fool the ignorant. The solution isn't to ban statistics. It's to educate decision makers so they can tell the difference.
Any social program in the government uses stastical methods to determine utility, impact and ROI.
Why? It works. Would you prefer a more "faith-based" approach?
Perhaps in our day and age of more direct interactions (ie, email, SMS, twitter) it might be possible to augment statistical methods with direct feedback.
Agreed. I see lots of people arguing that statistics is a lie, a bad thing, a tool of evil, etc. Statistics is a tool - just like English language. I don't see people condemning English as a tool of evil because you can lie in it.
I think this is an oversimplification, and a dangerous one. These statistics are biased, one-sided and misleading. That is indeed a property of some or even most statistics, just as "being wrong" is a property of some or even most information. Yet no-one would think to proclaim that the problem with government is that it believes information.
Real statistics, like real science, is a force for good. Bad statistics, like pseudoscience, is a force for bad - but it can only fool the ignorant. The solution isn't to ban statistics. It's to educate decision makers so they can tell the difference.