Bank for International Settlements 84th Annual Report, page 11 —
Second, as growing evidence suggests, balance sheet recessions are less responsive to traditional demand management measures (Chapter V). One reason is that banks need to repair their balance sheets. As long as asset quality is poor and capital meagre, banks will tend to restrict overall credit supply and, more importantly, misallocate it. As they lick their wounds, they will naturally retrench. But they will keep on lending to derelict borrowers (to avoid recognising losses) while cutting back on credit or making it dearer for those in better shape. A second, even more important, reason is that overly indebted agents will wish to pay down debt and save more. Give them an additional unit of income, as fiscal policy would do, and they will save it, not spend it. Encourage them to borrow more by reducing interest rates, as monetary policy would do, and they will refuse to oblige.
Note the contradictory logic of the two reasons. The first says that banks will only lend to bad borrowers and willing good borrowers can’t get credit. The second says people getting extra income from fiscal or monetary stimulus will only use it to pay down debt, reducing demand. But what about those good borrowers who can’t get credit due to the first reason? And what about those people who are paying down debt, thus helping banks get into better shape and thus, er, lend?
What starts out as an argument for weak multipliers doesn’t add up, and it’s not made any easier to follow by the apparent decision not to directly address their main critic on this point, Paul Krugman.