"We've got to reverse QE and do so quickly" says Tim Worstall. I don't want to take a view on that. What interests me instead is that reverse QE might be a misnomer, because central banks selling bonds doesn't have the same effects as them buying bonds. "Reverse QE" is not simply QE with the opposite sign.

What I mean is that the first round of QE in particular worked (pdf) through two mechanisms. One was a portfolio rebalancing effect. The Bank's buying of gilts forced yields down, and this helped reduced yields on equities and corporate bonds as investors rebalanced their portfolios towards those assets. The other was a signalling effect. QE signalled to everybody that the Bank intended for interest rates to stay low for a long time. This reduced gilt yields simply by reducing the expected path of short rates.

With QE, these two effects worked together. With reverse QE, however, they'll work in opposite directions. Yes, central bank selling of bonds will tend to depress their prices and raise yields. But the signalling effect will work against this. Investors might interpret reverse QE as an alternative to rate rises - because if policy is being tightened via the portfolio rebalancing effect, it won't need to be tightened so much via higher short rates.

The net impact of reverse QE is therefore unclear.

What we have here is an asymmetry. Policy works differently in one direction than it does in the other.

Which brings me to a more interesting question. How large is this set of asymmetric policies? I suspect it Is quite big.

Fiscal policy at the zero bound might be another example. If policy tightens, multipliers (pdf) might be large because there might be little reliable monetary offset; how true this is depends upon your view of the reliability of QE*. But if policy loosens, multipliers might be smaller because there is a monetary offset: rates can rise.

Brexit is another example. Having left the EU, we might not be able to rejoin it on the terms we currently enjoy: we might, for example, have to join the euro.

Another set of examples concern hysteresis effects. Simon says that austerity might have reduced trend growth by creating an innovations gap. Reversing austerity might not close this gap. It's possible that memories of years of slow growth post-2008 will continue to depress animal spirits even under more sensible fiscal policy.

The Lawson boom of the late 80s might be an example of this. It led to higher inflation in part perhaps because the Thatcher recession had destroyed skills and industrial capacity. Reversing tight policy did not therefore undo the damage of that policy.

Other examples centre on cultural effects. To take a long-term historical example, freeing slaves did not fully reverse the damage of slavery, as its adverse effects are still with us.

I fear that inequality falls into this category. The IFS said this week that inequality has fallen. This might not, however, swiftly reverse the damage done by increased inequality in the 80s and 90s - for example, the reduction in trust and in productivity.

Perhaps top tax rates are another example. The cut in these in the 80s led to (or at least was associated with) a sense among the rich that they are entitled to their pre-tax incomes. It reduced tax morale - the rich's willingness to comply with tax laws. And this in turn means that raising tax rates would lead to more tax-dodging, emigration or retirement than would otherwise have been the case. Reversing tax cuts might not therefore raise very much revenue.

Maybe you can think of other, better, examples from other fields.

My point here is that we shouldn't assume that policy-making is a simple hydraulic process of pulling and pushing levers. Pulling a lever doesn't always simply reverse the effect of pushing it. Bill Phillips Moniac model is perhaps sometimes a bad analogy. Instead, I prefer the wisdom which I associate with Joseph Schumpeter (although I can't track down the source): "If a man has been hit by a truck, you do not restore him to health simply by reversing the truck."

* I think reliability is a key issue. Maybe there was a particular amount of QE the Bank of England might have done to offset fiscal austerity. But given the Bank's inability to know this quantity precisely in 2010-11, a full offset was very difficult.