Stephen Roach's article bashing QE 3 and expansionary monetary policy in general is a very odd duck. One expects an article on that theme to simply deny that there's a demand problem and say we need to focus more on structural reforms blah blah blah.
But actually Roach thinks we're in "a protracted balance-sheet recession" (i.e., debt-induced crisis of demand), so I really don't understand what his problem is. It's precisely from a pure balance-sheet view that monetary expansion makes the most sense. A large and nontrivial share of debts are nominally denominated, so in terms of how indebted the economy is it's the nominal size of GDP that matters. Anything the Federal Reserve does to push nominal GDP upward is per se a deleveraging measure. Personally, I think these balance-sheet issues are often overemphasized, but to the extent that you buy a balance-sheet story, you need to get much more enthusiastic about monetary easing.
What's even stranger is where Roach thinks this balance-sheet problem arises, namely that excessive debt loads have "turned a generation of America’s consumers into zombies – the economic walking dead."
Here I've indexed the four major components of domestic demand—personal consumption, government purchases and investment, residential investment, and nonresidential investment—to their precrash peaks:
What you see is that it's actually personal consumption that's done the best here. And why not? People still eat, people still go to the doctor, we're all slinging around smartphones and iPads, if the fridge breaks you still buy a new one, I'm very committed to having pants on all the time, et cetera. By contrast, residential investment is a disaster. Americans own the biggest houses in the world, houses last a long time, and population growth is slowing thanks to an immigration crackdown, so it turns out to be possible to keep housing the population without building new houses—twentysomethings just live at home and unemployed people move in with their sisters. Nonresidential investment is also in bad shape, and if you peer into the details, this turns out to be happening despite strong equipment and software investment. Basically we're not building malls and offices.
Then there's the government. For a while, we did the smart thing had government purchases pick up some of the slack. But lately, this has been edging downward even though it's really an ideal time for some public-sector investment in durable infrastructure. And all signs are that the decline will get even more severe next year.