Paul Krugman tears apart a new post from John Taylor in which he attributes high unemployment to the falloff of investment, noting the sharp decline of investment as a share of GDP. Taylor's remedy for this problem is to spur investment by cutting business taxes. 

Krugman shows that the bulk of the falloff was due to a falloff in residential investment. In other words, housing construction plummeted following the collapse of the housing bubble. Actually, Taylor's case is even even weaker than the picture Krugman shows. There was also a bubble in non-residential real estate. The falloff in non-residential construction is due to the fact that there was enormous over-construction in most categories of non-residential construction. Tax breaks are not going to persuade builders to put up another office building or mall in a glutted market.

Investment in equipment and software is down by less than 0.8 percentage points as a share of GDP from its pre-bubble peak. That is not bad given the falloff in demand. I have also included a line that subtracts vehicle leasing from investment. The issue here is that a leased vehicle will count as investment by the car leasing company, whereas a purchased vehicle will count as consumption by the consumer. There was a big surge in car leasing in the 90s which explained about 0.3 percentage points of the increase in the investment share of GDP over the course of the decade. (The calculation here just subtracts the current lease expenditures. In principle we would want to pick up the value of cars purchased for leasing. The expenditures on leased cars likely lag the purchases by a year or so. They will also not be exactly the same, but they should give the right general size of this effect.)

Click to Enlarge
Source: Bureau of Economic Analysis.