This is a bit premature — we won’t be doing Markov switching until the end of March.
But, the March 2 issue of The Wall Street Journal contained this graphic to accompany an article entitled “States Fumble Revenue Forecasts”:
What I want you to get out of this is the pattern of errors: long periods of underestimates during expansions, punctuated by recessions with overestimates.
Markov-switching is a process that can explain this, although neither it (or much else) will be much help in solving this problem. We just have to live with it, and recognize that it will happen (that’s not something that government or business owners are very good at).