When should policy makers start to withdraw the stimulus? 

Christine Romer in the Economist digs up an historical analogy that Paul Krugman is fond of: the way FDR precipitated a mini-Depression in 1937-8 by going for a balanced budget too soon, and also tightening monetary conditions. She warns against holding back on stimulus until we are returning to full employment.  Paul Samuelson too argues for the stimulus to be sustained.  Perhaps the green shoots we are seeing are only, um, shooting because of the expectation that the stimulus will last.

Similar reasoning helps explain why Wolfgang Munchau is so furious about the Berlin Hairshirt. A constitutional commitment to a tight fiscal policy is crazy.  What if Germany does not resume growth at a healthy pace? Munchau writes:

"This general level of debt-aversion is bizarre. Many ordinary Germans regard debt as morally objectionable, even if it is put to proper use. They see the financial crisis primarily as a moral crisis of Anglo-Saxon capitalism. The balanced budget constitutional law is therefore not about economics. It is a moral crusade, and it is the last thing, Germany, the eurozone and the world need right now."

This is one reason for our aversion to the Conservatives' Office for Budget Responsibility.   In certain circumstances - say, when the economy is growing, there is a surplus and a large debt to repay - having a Victorian Uncle tutting over the Chancellor's shoulder about debt repayment may make sense.  But what about during another downturn? Fiscal policy is a matter for judgement, not dogma.  You may hear more reasons for this aversion on Sunday in Westminster Hour, for which I was interviewed. 

One reason to slam on the fiscal brakes might be long-term interest rates getting out of control (see earlier discussion about Niall Ferguson's views here).  When the economy is at capacity, public and private growth start to compete against one another.   But there seems to be a reasonable consensus behind the idea that the recent rise in rates is benevolent, reflecting higher future growth.  Paul Krugman, obviously, agrees.  However, David Smith discusses Mervyn King's views and hints that rates might rise faster than people currently expect.

One day we will have to deal with the structural fiscal deficit - we have a big paper coming out on the subject soon. Public spending cuts will surely play a part.  But I am amazed at the unbalanced hostility that the FT shows towards the public sector in this info-graphic. Phrases like "unsustainable leviathan" beg the question.  Few analysts point out how the loss of revenues and a collapse in nominal GDP are the immediate cause of the debt explosion.  We will.  

The other big issue is the future of banking regulation.  There is a growing consensus behind the idea of a wise systemic regulator solving problems in future.  No more will regulators just look at individual companies: instead, they will analyse the whole universe of financial entities, and work out when it looks likely to blow up the world, again. 

Hmm.  I'm glad someone is kicking back against this idea, or at least asking it questions.  John Plender:

"Heavy reliance, then, is being placed on the ability of this new regulatory approach to prevent a future global financial crisis. Yet there is no agreement on how it might work in practice and good reason to question whether it will live up to its advance billing. One very senior former member of the global central banking establishment even says “not only is macro-prudential regulation rubbish, but it gives rubbish a bad name

Capital requirements that track the cycle, together with overall leverage ratios, are thus central planks of the macro-prudential approach. Yet the attempt to counter pro-cyclicality raises huge questions, most notably on the issue of whether regulators should have discretion to change capital requirements in the course of the cycle or whether the capital regime should be subject to pre-determined rules. In a perfect world, giving discretion to regulators to calibrate capital requirements according to the state of the economy makes sense. In the real world, regulators would face the ever-difficult problem of defining where they were in the economic cycle.”.

Precisely.   An excellent letter today makes a similar point. 

"Virtually nobody within the commercial banks themselves recognised the danger of what they were putting on their balance sheets. How could one expect a regulator, whose information is not nearly as complete, to ascertain a problem? Not a single central banker in the last episode noticed the crisis before it was already manifest. "  Well said, Thomas Dempsey, Chadds Ford, PA, US

But Ric Mishkin disagrees.