Monetizing the deficit

Paul has been arguing against the need for cuts in government borrowing. I fear he’s making a simple subject unnecessarily complicated. We don’t need any “alternative economic model.“ A trivial identity is sufficient. This is simply that government borrowing can be financed either by issuing bonds or by printing money.
This implies that bond markets are not necessarily the binding constraint on government borrowing. If investors don’t want to buy as many bonds as the government issues, the government can print money instead; Ralph has been emphasizing this for some time.
Nor is this any mere theoretical possibility. It is what the Bank of England did under its quantitative easing policy. In 2009-10, it bought £198.3bn of the £227.6bn of gilts the government issued. In effect, then,  most of the government’s deficit was financed by printing money.
So, why can’t this be done on a larger scale, as Paul advocates, to remove the government‘s dependence on bond markets?
The answer is that it is potentially inflationary. Paul Krugman does the maths in reply to Jamie Galbraith (point 9 is most relevant) here - and see this reply to Krugman.
The key word here is “potentially”. It is not the case - as we’ve seen with QE - that printing money leads quickly and mechanically to higher inflation. But it is the case that, if the government does print enough money, inflation will ensue. After all, the price level is simply the inverse of the value of money. And if there's a big enough supply of money, its value must eventually fall.
But when is “enough” and when is "eventually"? My concern is that they could be sufficiently near to where we are now as to rule out large or persistent monetization of the deficit. I say so for three reasons:
1. Inflation has recently been surprisingly high. This suggests that the recession isn’t purely a story of weak demand and deflationary pressures - in which case printing money would be an easy solution. It might also be a story of a supply shock. This suggests that insofar as printing money does increase nominal demand, it might swiftly lead to higher prices rather than just greater economic activity.
2. A government that announces an intention to print money on a massive scale might well see its exchange rate fall, which might bring inflation via higher import prices.
3. If people expect  a large rise  in the money stock in future, we could get inflation  today - a point emphasized by John Cochrane here (pdf). 
You might reply: what’s wrong with a little inflation - assuming there is only a little? Simple. Higher inflation can be bad for economic activity. It might deter investment as lenders require higher real (not just nominal) interest rates to compensate for greater risks to their real spending power. And higher inflation often leads to increased uncertainty and hence lower capital spending and higher savings; the inflation of the 1970s, remember, was accompanied by weak activity and this wasn’t entirely perhaps an accident.
Now, I don’t say all this to dismiss entirely the notion of printing money. It might well work as a temporary expedient if not as a permanent remedy. This is an empirical matter; there is nothing odd or cranky about the proposal in theeory.
But let’s remember a key fact here. The bond market has not been demanding big cuts in borrowing; real gilt yields have been extraordinarily low for months. Osborne’s proposed cuts are a political judgment, not a necessary response to market pressure. The bond market is not - or at least not yet - a significant constraint upon government borrowing.

Comments

[Neil]

You too, are making the subject unnecessarily complicated.

If the amount of money increases, but the total value of goods in the economy remains static, it is a mathematical certainty that the purchasing power of each pound will diminish in proportion. That the markets may take some time to reflect this change though prices is irrelevant; the very moment the government prints new money, it diminishes the value of all the money already in existence.

In a world of fiat currency, "quantitative easing" is simply a stealth tax. New money appears in the government's accounts, everyone else's money becomes worth that much less. Wealth has been transferred from private hands to the public coffers. Simple as you like.

[Neil]

Of course, QE is not *necessarily* inflationary; the alternative is that investors mistake the increase in money supply with an increase in actual wealth, and invest accordingly in whichever market seems to be booming; so you get a bubble which will be sustained until people realise that the wealth it's built on is a mirage.

So, inflation or boom-and-bust?

[Left Outside]

I'd add neil, that the increase has to be expected to be permanent for your point to hold. Otherwise it will not be inflationary at all. An explicit price level target might work with QE. Sadly some things you have to complicate a little.

[Left Outside]

Oh and Chris, although inflation has still been surprisingly high, it is still way under trend. Likewise, although employment has held up surprisingly well demand is only going to get hit so disinfaltion is my defailt position at the mo.

[Luis Enrique]

Questions I should know the answer to already:

1. When the Bank of England buys government debt, does it keep the IOU on its balance sheet as "the government owes me £X" and do the government debt accounts include, "we owe the BoE £X"? Or does the BoE just tear up the govt's IOU, and the govt's debt is just written off. I'd really like to know the answer to this one!

2. Printing lots of money now might not be inflationary short-term, but is it necessarily storing up inflationary pressure at some point in the future? How easy is it to "sterilize" when the future arrives. Likewise with govt borrowing, the bond market might not be a constraint to borrowing, but in (say) 10-years time we're either going to have to pay or roll-over that debt, so when we're thinking about "how much can we afford to borrow" don't we need to be taking into account that future, not just today's interest rates?

["... mathematical certainty that the purchasing power of each pound will diminish in proportion" If the velocity of money is a constant. Which it aint.]

[Duncan]

Luis,

1/ It keeps them and is intedning to sell them back to the market at some point.

2/ Yes in ten years time we roll it over. But GDP has grown so debt/GDP is lower - we never actually pay any of this stuff off! But we grow.

I've posted a few, slightly related thoughts, below:

http://duncanseconomicblog.wordpress.com/2010/07/19/cheap-money-revisited/

[Luis Enrique]

Duncan, thanks

well, on 1., if we really want to inflate & stimulate the economy without worrying about debt, why doesn't the CB just give the govt money by buying new debt and then tear it up? One answer is that it would leave the CB without the means to reduce the money supply in the future (which it does by selling those bonds back) but in a deflationary environment in which the size of government debt is a worry, I don't quite understand why it ought not be considered.

I'd like to know more about the extent to which the CB effectively writes-off govt debt, and what determines how much it can get away with. Of course the CB has to increase the money supply all the time, as the economy grows - i.e. seignorage - which entails govt debt sitting there never to be paid back.

on 2. the relationship between issuing debt now and the path of GDP growth is what matters. Obviously if the future path of GDP growth is fixed, there will be some quantity of borrowing above which is too much. I do not accept that low-short run borrowing costs necessarily mean we can just borrow away and assume a rosy future. I find it a bit weird after a financial crisis, caused by "rash" behavior, to see people advocating borrowing without worrying about paying it back. Inflating debt away is not painless.

[Wiseman387]

' Nor is this any mere theoretical possibility. It is what the Bank of England did under its quantitative easing policy. In 2009-10, it bought £198.3bn of the £227.6bn of gilts the government issued. In effect, then, most of the government’s deficit was financed by printing money. '

That is not strictly true. During QE the BoE in their reverse auctions were deliberately buying different maturities to the ones the DMO were issuing to avoid that accusation. The institutions selling gilts to the Bank were not necessarily the same ones buying at DMO auctions.

Posted by: Luis Enrique | July 19, 2010 at 04:27 PM

'Questions I should know the answer to already:

1. When the Bank of England buys government debt, does it keep the IOU on its balance sheet as "the government owes me £X" and do the government debt accounts include, "we owe the BoE £X"? Or does the BoE just tear up the govt's IOU, and the govt's debt is just written off. I'd really like to know the answer to this one! '

The BoE hold the gilts on their balance sheet as assets. Moreover, the Treasury make full interest payments to the BoE for the gilts they hold just the same as for all other gilt holders. The BoE have been making large profits from these coupon payments, which are of course returned to the Treasury. In fact, the prices the BoE paid purchasing gilts during QE is showing a paper profit of 10 billion sterling. Many people were commenting during QE that the Bank were overpaying giving free money to the winners at the auctions. However, since then yields have fallen and prices have risen, hence the notional 10 billion profit. There is no theoretical reason why the BoE can't hold the gilts to maturity.

2. ' Printing lots of money now might not be inflationary short-term, but is it necessarily storing up inflationary pressure at some point in the future? How easy is it to "sterilize" when the future arrives. Likewise with govt borrowing, the bond market might not be a constraint to borrowing, but in (say) 10-years time we're either going to have to pay or roll-over that debt, so when we're thinking about "how much can we afford to borrow" don't we need to be taking into account that future, not just today's interest rates? '

Well that is just about good debt management. You cover all parts of the yield curve to minimise roll-over risks. Part of the euro periphery problems is so much of their debt is short-dated. It is cheaper for governments to issue short-dated debt and roll-over. Until one day a crisis hits and the market reprices your risk and you are left with lots of short-dated debt to roll-over at a much higher interest rate. See eurozone for what happens.

Roll-over risks are not a serious problem for the UK because the maturity is around 13 years. Ironically what makes us less of a sovereign risk exacerbated the banking crisis for UK banks. During the banking crisis they had no liquid short-dated government securities because the DMO had not been issuing them. Their balance sheets were full of illiquid mortgage-backed securities.

' why doesn't the CB just give the govt money by buying new debt and then tear it up? '

The government do not need anyone to give them money as they just credit accounts. The gilts are issued after the government spends. Why then do we have a gilts market ? The risk-free rate is important as a benchmark for pricing all other risk. Moreover, the bond market places a discipline on the government.

[Rich]

Inflation could be good for my gigantic mortgage though ...

[Joe Otten]

It seems to me that we have QE because it is "silly" to reduce the base rate below 0.5%.

So the way I would put this question - is it maybe silly to have base rates below, say 1.5%, in which case we could have more QE.