Why some policy changes should be slow and measured: a case study

This is Joseph

Gary Ray nails it:

The problem with minimum wage is when it rises too fast. When it rises above a reasonable level of annual profit of a small business, you've got trouble. Using Borderlands numbers, a 20% rise over 3 years would require a nearly 7% increase in sales. Most retail businesses are not growing at 7% a year. I don't need back of the napkin math to figure this out. The Department of Commerce numbers show average retail sales growth median to be 5.04%, with 2014 at 3.17%. So minimum wage yes, but not so quickly.

I think that fast changes accelerate the costs of policies by increasing the penalty from disruption.  I think that this is an unpopular idea in the "age of disruption" (see Uber) but slower phasing in of policy can often permit industries and people to adapt.  It's also notable that bad industry practice, like the MSRP, was involved in this outcome by making bookstores less flexible in the face of changing costs.

Noah Smith has more on the specific situation, and his own comment on the externalities involved in fast changes. 

Does this mean a minimum wage increase is bad policy?  No, but I think it would be technically better to phase it in slowly, like Canada, rather than a large change all at once.