Everyone knows that central banks, like the Bank of England, create money. We often talk about this as central banks printing more money, but in reality when central banks want to create money they do so by creating a kind of electronic money held by commercial banks called ‘reserves’. Quantitative Easing is when a central bank creates reserves to buy its own government’s debt.
Central banks do this in a recession when the short term interest rate set by the central bank is so low that it does not want to reduce it further (the lower bound for interest rates), but it still wants to stimulate the economy. They cannot reduce short term interest rates any further, but can they reduce long term interest rates (the interest rate on assets held for many years)? The idea is that if the central bank buys government debt, there is less government debt for the private sector to buy, so the private sector is willing to accept a slightly lower interest rate on government debt. That in turn will put downward pressure on other longer term interest rates, stimulating the economy.
There is an extensive literature on how effective this Quantitative Easing (QE) is in stimulating the economy, but that is not my concern here. All I need to assume is that it has some limited effect in stimulating the economy. The question is whether this is something the central bank should be doing during recessionary periods.  For example, in a recent report by the Lords Economic Affairs Committee, the chairman said the Bank of England had become addicted to QE.
The problem with a great deal of public discourse on QE is that it takes QE as an alternative to fiscal stimulus, rather than a complement to it. There is a good reason for this confusion. During the aftermath of the Global Financial Crisis (GFC) many governments failed to enact fiscal stimulus and instead enacted austerity, saying that monetary policy could do the job. As interest rates were stuck at their lower bound, unconventional monetary was used, and QE was a key component of that. In the event the prolonged recession showed clearly that QE could not take the place of fiscal stimulus. But the problem was the lack of fiscal stimulus (and its replacement by the opposite, austerity), not QE itself.
The most common charge against QE is that it worsens inequality. As explained earlier, one aim of QE is to slightly lower long term interest rates. A fall in long term interest rates tends to raise asset prices, which includes the market value of existing government debt, stock prices and house prices. 
As the wealthy rather than the poor hold these assets, QE tends to raise the wealth of the wealthy, and so increases wealth inequality. However QE is not the most important influence on long term interest rates. Long term interest rates change mainly because expectations of future short term interest rates change. 
The best way to reduce inequality in a recession, therefore, is enact a large fiscal stimulus. This raises demand and inflation tends to rise. As a result, central banks will raise short term interest rates, which in turn will increase long term rates and lower asset prices. We can immediately see that, if you are concerned about wealth inequality, it’s the lack of fiscal stimulus that is the real problem, not QE.
Another, more primitive, argument against QE is that it causes inflation. This is primitive because it’s simple monetarism, which for most macroeconomists died a death in the 1980s. I could go through the theoretical reasons why it is nonsense, but it’s easier to note that the same people made the same claim last time there was a lot of QE (after the GFC) and there was no significant increase in domestically generated inflation.
Is it better to have a fiscal expansion combined with QE, rather than just a fiscal expansion? The only difference is that in the first case the central bank buys some government debt for some undefined period, and in the second case it doesn’t. The two options are often referred to by economists as money and bond financed fiscal expansion.  In very simple textbook models money financing involves greater stimulus, but often in those textbooks this is because the government is imagined to target the money supply, and that stopped happening in the early 1980s.
A more modern response would be that QE still has some impact in reducing long term interest rates , and that extra stimulus is worth having. On the other hand you could just increase the size of fiscal stimulus to compensate, and not do any QE at all. So the question remains whether there is any point in doing QE alongside fiscal expansion.
One significant positive point in favour of doing so is that money financing (fiscal stimulus plus QE) is cheaper than bond financing. Instead of the government paying interest on its debt, it gets paid to the central bank instead, and the central bank passes that money on to the Treasury. However there is currently a cost to QE in the UK and US, which is that the central bank pays its short term interest rate on the reserves held by commercial banks. Nevertheless, as short term interest rates are lower than the interest on longer term government debt, QE still reduces the borrowing cost of a fiscal stimulus. The OBR estimates that QE will save the government just under £18 billion in 2021/2.
However this has led some, including the OBR, to suggest that the financing costs of government debt are now much more sensitive to interest rates than they would be without QE. Without QE, if the Bank raised short term interest rates this would have a very limited immediate impact on the cost of UK government debt because of its long maturity. However with QE, if the Bank continued to pay its short term interest rate on reserves, the overall cost to the public sector would rise much more quickly.
This logic is based on a very unlikely assumption. It assumes that if short term interest rates rose the Bank would continue its practice of paying that rate on all its reserves. If nothing else that would be a large fiscal transfer from the Bank to commercial banks. As Karl Whelan sets out very clearly, it is far more likely the Bank would follow Japan and the ECB and adopt a tiering system, which would mean that most reserves would pay nothing. The scare about how vulnerable QE makes the government to higher rates is simply something the Bank can make disappear.
Occasionally you hear people say that QE (or monetary financing) is dangerous because it damages central bank independence. The idea is that governments will pressurise central banks to continue to issue new QE, after a recovery from recession is complete, because it makes its debt cheaper. Again we can discount that fear because we saw none of that after the GFC. 
I would add a final reason why expanding QE is a good idea in recessions. It prevents panics in the government debt market. If the market suddenly decides it doesn’t want to buy any debt, the Bank can step in. This happened at the start of the global COVID pandemic, and central banks did just that. More generally, it kills the argument that we cannot have fiscal stimulus in a recession because of the bond market, a point I made in one of my first posts. My own view, therefore, is that QE is worth doing, and it only has a bad name because in the past it has been used instead of fiscal stimulus.
Unfortunately there is a real danger that situation will happen again in the UK, as the Chancellor finds it more and more difficult to meet his deficit and debt targets and so asks for greater austerity. Once the latest COVID wave is over, our aim should be to use fiscal stimulus to expand the economy, as they are doing in the US. Instead it looks like we will have austerity trying to hit some arbitrary target for debt or the deficit.
There are many differences between austerity now and the austerity that began in 2010. In 2010 the Bank had a Governor who believed he could continue to effectively stabilise the economy with Quantitative Easing. Now central bankers are increasingly acknowledging the need for fiscal stimulus to end recessions. The US fiscal stimulus is forecast to return the US economy to where it would have been without any pandemic, and any failure of the UK economy to do the same may be largely down to our current Chancellor’s obsession with debt targets. Once again we will see QE without the necessary fiscal stimulus that should go with it.
 By recessionary periods, I mean the whole period between when output starts growing less than normal (or falling), and when output has completely recovered. This has nothing to do with the ‘technical’ definition of a recession (two periods of falling output), which I find can be misleading and unhelpful.
 The reason for this is simple. Ignoring any capital gains, the pay out from these assets are a fixed interest rate (or coupon) for most government debt, dividends for stock prices and rents (actual or implicit) for house prices. If interest rates fall and these payouts remain unchanged, all these assets become more attractive to hold (their pay out becomes more attractive relative to lower long term interest rates), so the price of these assets rise.
 If traders start expecting short term interest rates to go up in the future, and long term interest rates are low, it’s more attractive to short term assets rather than long term debt. If people are still going to hold long term debt, long term interest rates need to rise to compensate for higher expected future short term interest rates.
 There is some confusion about whether QE is monetary financing. This site, for example, says it’s not because QE is only a temporary creation of money, as central banks intend to sell the government debt it bought with the new money at some future date. However most examples of fiscal stimulus are temporary, so it seems perfectly sensible to compare money finance stimulus with debt financed stimulus, where the difference between the two is QE.
 A government that did this is what I call a government of central bank nightmares, and such a government is equally likely to just end central bank independence. I see no reason to believe that QE encourages governments to turn into this nightmare state.