Certainly you've heard a government official frame an issue he didn't like as "bad for the economy"? In the US, it's basically the first page in the playbook for both sides of the aisle because very few issues are as sacred as a healthy economy. But if hurting the economy just means bringing down the GDP, then there are countless reasons why the underlying issue might still be beneficial to the welfare of its citizens.

Want to boost your country's GDP? Here's some tips. Become terminally ill. Get involved in a costly divorce. Try gambling or drinking habitually. Take a job where you commute long hours wasting gas in traffic. Hell, open a strip mine - every resource extracted is a net positive on the national ledger. Bonus points if you can coerce your fellow citizens into compulsory spending - a nasty coal plant triggering asthma in the local population is a great example.

Even worse, a focus on GDP directly contributes to income inequality because as a measure it is blind to the distribution. If the top 1% gain more than every one else loses in a year, GDP still rises. For this reason Rowe calls GDP a "statistical laundry operation that hides the suffering at the bottom".

Measuring and tracking GDP is a splendid idea, but using it as even a proxy for national welfare is insanity. I encourage anyone interested to read the linked article, which is as true today as when Rowe testified before the US senate 8 years ago. Also at the end he takes a first pass at laying out principles around new metrics to fix the situation.

It's not merely used in policy making: GDP as a metric basically emerged for policy making in the contexts of resource allocation during the Great Depression and war planning during WWII. (And arguably during prior wars before that: http://www.google.com/search?q=invention+of+gdp .)