I posted this Twitter thread on the multiplier this morning:
The government wants to cut its spending. It says we cannot afford it. The claim is that if we have to borrow or increase taxes to spend then that must mean we can't afford things like more education or healthcare. But they are wrong. We should spend, now. A thread to explain…
First, trust me, everything I say in this thread is actually pretty straightforward. It's just stuff no politician ever talks about. That's because the truth I relate doesn't suit a political narrative that wants to cut the size of the state, come what may.
Second, also trust me that what I am going to discuss is not some weird, unknown, form of economics. This is actually fairly bog-standard stuff that's been known for a long time. It's just that politicians pretend they do not know it.
Third, there's technical name for what I am about to describe, which is the ‘multiplier'. The good news is that multiplication is, after adding up and subtraction, the easiest bit of maths, and vastly simpler than division, so nothing I am going to say is that hard.
So, as the saying goes, let's start at the very beginning. Let's assume a government spends and extra £1 into the economy. As assumptions go this is an easy one to make. The UK government spends more than £800 billion a year, so assuming it spends £1 more is not hard.
What the multiplier measures is how much new economic activity results from that additional £1 that the government spends. That's it. It's really not that hard to understand. That one sentence summarises what this whole thread is about.
It also hits the political economy of this on the head. If most right-of-centre politicians were to be believed every £1 that the government spends is paired into a black hole and we get not a penny's worth from it. They try to describe it as waste.
But that's not true, of course. The government does not pour money away. What it actually does is spend it for social purpose. What is more, its spending becomes someone else's income. And it adds to the sum of well being as a result. Understanding by how much is key.
The rule is simple. If the multiplier is 1, the £1 the government spends creates £1 of new economic activity. If that pound creates less than £1 of new income then the multiplier is less than 1. And if it creates more than £1 of new income the multiplier is bigger than 1.
So, a multiplier of less than 1 does not look good, and a multiplier of 1 only looks OK, whilst a multiplier of more than 1 is beginning to suggest something good is going on.
How does this work? An example will help. Suppose the government spends £1 extra on a teacher. That £1 is taxed, of course. The teacher probably gets about 67p of the income paid to them. Now, if they did nothing with that 67p the money spent would have achieved little.
This is possible, of course. The teacher could just put the money in their bank account and do nothing with it. That's what we call saving. But that means nothing happens with the cash: it becomes dead money.
And don't presume that saving is good because banks lend on the money deposited with them. We know that's not true now. The Bank of England admitted it in 2014. Cash savings are just money that's met a dead end. Banks lend newly created money, not savings.
So, if ending up in a savings account was to be the fate of the new government spending the multiplier would be very low. Income in the economy would hardly rise, at all.
Thankfully, that's not what usually happens when the government spends, for two reasons. The first is that few people save all their money. In fact, most people have almost no savings at all. That implies they spend most, if not all, of their income.
Suppose, then, that the teacher spends their extra 67p. What happens? The reality is that their spend does then become somebody else's income. The mechanism can be complicated, but if money is spent by one person it must always becomes the income of someone else.
On the way some more tax might be paid. If the money went to a shop then VAT might be paid. And even though the shop will have had to buy goods in some of what they get will also go to the shop staff. They're also taxed, of course, but then they'll also spend their net income.
That will also be true of the staff in the companies that made the goods that the shop bought in. So they too have a bit more to spend as well out of the extra £1 the government spent.
And so the process goes on, and on, with the amounts getting smaller at each stage, but still being significant so long as the process does not stop because someone saves instead.
So, two effects have been created. The first is that the 67p the teacher gets after tax goes on to be additional income for a lot more people than the teacher who first got the money.
Second, tax is paid at each stage of this process. The £1 of extra spend already creates 33p of additional tax to pay for itself when it is first paid to the teacher. But, at each stage as the 67p the teacher got changes hands the amount of tax paid, overall, increases.
In that case the multiplier effect as the money spent goes around and around in the economy does not just increase earnings; it also increases government revenues too.
That last point is really important and yet no one says it. The fact is that every time the government spends it increases its tax take. Politicians never say that, but it is a fact. It is in fact possible that government spending can entirely pay for itself out of new taxes paid.
Suppose the average UK overall tax rate is 33%, as it is, near enough, in the UK. Then suppose the multiplier is 3. In other words, the extra £1 spend creates £3 of new income. If that £3 of extra income is taxed at 33% the net cost to the government of spending is nothing.
The additional spend is £1 in this example. And as a result of that spend the additional tax due in this example is £1. The result is that the books balance, despite an increased spend. The spend pays for itself. That's not a miracle: that can happen.
The question is, how can the multiplier be 3? The answer is that this not only depends on the money continuing to move around the economy, although that is important. It also depends on what the extra money is spent on.
Some things, like defence spending, have low multipliers. Defence creates nothing that anyone wants to buy. So, whilst the spend on the people involved has some stimulus effect, because no one buys defence there is nothing for that extra government created money to be spent on.
The result is that money spent on defence might have a stimulus effect to the extent that the spend is re-spent by those who get it, the second critical aspect of the multiplier hardly exists with regard ti defence spending.
That second critical aspect is the impact the spend has on the rest of the economy. Contrast defence spending with health spending. Both have broadly similar direct impacts in that what is spent on people gets re-spent. But their impact on the economy is givers different.
Defence is the black hole that right wing politicians think all government spending is. It literally generates no value. Spend on the NHS though and you have healthier, happier, more optimistic people. They go to work too. And they spend more.
What is more, because healthy people are more productive they add more value when they are at work. So that increases the multiplier as well, of course.
The result is that the multiplier effect of health spending is way above 1. The King's Fund - a UK health think tank - thinks that the multiplier effect of health spending is 3.6. So for very £1 spent £3.60 of income is generated. And more than £1 of tax is paid.
The US Federal Reserve has estimated that the multiplier on education spending is 2.4. It's a little hard to tell. It depends for how long you think the impact lasts. A lifetime is a long time. So maybe it's higher in this case. Multiplier measurement can be difficult.
But one thing pretty universally agreed is that spending on investment brings high returns. That's unsurprising. That's precisely what good investments should do. It also makes it hard to explain why the government does not invest more.
Another agreement is that multipliers are higher when unemployment is higher. Again, that's unsurprising. And, as @d_blanchlower argues, we have so much spare capacity due to underemployment in our economy multipliers should usually be high.
Add it up then and a great deal of government spending largely or entirely pays for itself if there is even modest underemployment in the economy. That would be especially true if that spend was investment in something like a Green New Deal.
There's one more thing to add. That is that the multiplier also applies to taxes, and most especially tax cuts. Of course, a tax cut is just like a government spend. It appears to inject more money into the economy. But, who gets it has an enormous impact on its multiplier effect.
Suppose the tax cut goes to the least well off. They will spend it all. The multiplier effect is high. They keep the money circulating. But, give it to wealthiest and they can already spend all they want, and save what's left over each month. So they just save some more.
Remember, most of the wealthier people in any country are wealthy precisely because they can save out of what they earn. That means they have already got all they want and are unlikely to spend a great deal more.
So what happens if you give the wealthiest, or the companies that they tend to own, tax cuts? They simply save them. In which case the government's generosity does not move around the economy. It just ends up stuck in a savings account, and that is economically dead money.
That, incidentally, is why offering tax reliefs to those who can already afford to save is also a really daft idea, but we do it in the UK at a cost exceeding £55 billion a year. The multiplier effect of doing so is likely to be well below 1, which is crazy.
Crazy that is unless your aim is to increase inequality, which is what this tax relief to savers really does.
So, in conclusion, multipliers matter. Saying that I stress that money is not the only reason that governments should do things: it would be wrong to say it was. But when government money is being discussed then multipliers should be included in the debate.
That's because all government spending and all taxes have a multiplier. Their measurement is not perfect. But we know they vary enormously. Some things, like spending on the NHS and healthcare have really high multipliers.
Such things can actually, or nearly, pay for themselves as far as the government is concerned, and leave a lot of additional income remaining in the economy as well.
Other things, like defence spending and tax cuts for large companies and the wealthy have low multiplier effects: they don't provide much stimulus at all.
Not all government spending is equal then. Some is a waste of money. Boosting savings is pretty near top of that list. Some is amazing in terms of its impact in the economy. And it pays for itself. Health and investment usually does that.
And this matters. When the government says it can't afford education, healthcare and green investment for example, it's simply not telling the truth. As these things pretty much pay for themselves it's ignorance or dogmatic denial that motivates such claims.
That's also true on taxes: tax cuts for the wealthy really don't boost the economy, they just boost social division. So you have to ask, is that the aim of the who promote them?
The point is, knowing this matters. The next time a politician says we can't afford something ask them about the multiplier effect of the proposed spend. They'll squirm. Then patiently explain that much government expenditure pays for itself by boosting incomes.
After that ask them why they'd rather not boost incomes when they can, often costlessly. Hold them to account in other words. Make them explain. Because that's what democracy is all about. And that's what economics lets you do.
Thanks for reading this post.
You can share this post on social media of your choice by clicking these icons:
You can subscribe to this blog's daily email here.
And if you would like to support this blog you can, here:
As you recognise, spending on health and education in particular are justified not just by the cold economics, but also by their effect on individual and social wellbeing. A lifetime of impact indeed.
We do need some defence spending, and the will to use it when necessary. For want of a few thousand soldiers in Afghanistan to support their government, the country is lost to militant fundamentalist Islamists.
Andrew
As I carefully noted, money should never be the sole explanation for any government spending
The U.K. and USA have been shameful on Afghanistan in recent months
Richard
only “in recent months” – you are too kind!
An excellent explanation of the multiplier. I will be sharing this!
That took some writing…
Some threads are quick…that one took several sessions over a few weeks
The next o9nbe is sectoral balances….it is in progress
As someone in the public sector whose wages have been in steep decline since 2010, I can tell you that this rings true.
We can still save, but we are saving less, that’s for sure – half in fact what we were saving previously.
A good explanation.
In some senses, the “neoliberal, where will the money come from” crowd who are terrified that government spending is inflationary should approve of health spending over defence spending because of the “automatic” nature of the draining of the money out of the system through taxation. They should also oppose all tax incentives for pension savings.
But they don’t.
Thanks Richard, one of your best posts. It makes a good case for something which as it happens I did understand in general terms – I remember when I was a student one of my friends having a Damascene moment and enthusing to everyone he met about understanding the power of the multiplier; that enthusiasm for economics clearly stuck and he later became a Professor at LSE.
I am still somewhat unsure about why there are such huge disparities in the multiplier between different areas of government spending. Taking your example of military spending, surely a significant proportion goes in people’s income which generates taxation and more spending. Obviously that isn’t the case for some of the big ticket procurement like fighter jets from US manufacturers, but then the same is the case for much health procurement; MRI scanners are imported (despite the UK being pivotal in developing the technology) and as we have recently discovered even the low-cost high-volume supplies like gloves and masks are manufactured elsewhere.
Perhaps the more important question though is what the multiplier would be if the country invested in a decent social care system for the elderly and disabled. Would any balancing taxation be required? It comes down though to the question above, how do you work it out? Is it largely the proportion of spend going on salary (for which social care scores highly) or is it more complicated than that? Does it account for the long term benefits, which are perhaps easier to identify for better health or better education than for better care at the end of life, or better fighter jets.
Jonathan
Two big issues: imports are black holes, but then exports balance them (hopefully) and provide investment with strong multipliers, so they must be considered together
Why different multipliers? Because the effect of spending in itself is to trend towards a multiplier of 1 if there is no additional impact on productivity in the economy
So, defence has no additional impact. But health and social care both have the effect of releasing productive resources and so have high multipliers. That’s it really
Can I ask – does this multiplier effect apply to other spending e.g. by companies or individuals? For example if I buy a new car using savings?
To a degree, yes, of course
The phenomena is used almost entirely on public sector spending but the private sector also has them – and there are therefore better or worse cycles
But usually different methods of appraisal are used
I recommend the Fournier/Johansson paper from the OECD in 2017 titled:
“The effect of the size and mix of public spending on growth and inequality”
Some things absolutely should be cut from government spending but they are the sort of things Tory voters want to keep.
@ Ms / Mr Kasem: The paper you refer to may be one of the fopllowing:
http://dx.doi.org/10.1787/f99f6b36-en
• Fournier, J. and A. Johansson (2016), The Effect of the Size and the Mix of Public Spending on Growth and Inequality, OECD Economics Department Working Papers, No. 1344, OECD Publishing, Paris.
or http://dx.doi.org/10.1787/c57eaa14-en
• Akgun, O., B. Cournede, J. M. Fournier (2017), The Effects of the Tax Mix on Inequality and Growth, OECD Economics Department Working Papers, No. 1447, OECD Publishing, Paris.
Found on OECD page
https://www.oecd-ilibrary.org/economics/the-effects-of-the-tax-mix-on-inequality-and-growth_c57eaa14-en
I would guess that the multiplier effect is also significant with State pensions, particularly for the less well off where most of any increase will be spent on heating and food? And yet the UK has the lowest pension in the EU, less than half the average I believe.
Yes….
This has been studied by the OECD.
And?
Richard,
If I may say – that is one of your best blogs; trenchant. I have been thinking a great deal on this issue, and which you have now neatly summarised with the invaluably term “dead money”; how apposite.
Neoliberalism was designed to pass control of the economy solely to those who command ‘dead money’; and it has worked very, very well. There you have a term loaded with meaning, and it is worth using.
Thank you
Appreciated
In a sense there is within the above thread a swift answer to the question “How are you going to pay for it?”
The reply can then be, “We estimate that the multiplier will be at least 2.5.”
If the interviewer doesn’t know what the multiplier is then some considerable surprise can be expressed by the interviewee at the interviewer’s capability even to ask the spending and paying question.
And if he does know what the multiplier is, the interviewer would then have to explain to the audience…
That should nicely change the dynamics of the interview.
Agreed
The reason for writing it
Another marvellous contribution , Richard, which I will share as widely as makes sense. It might have been worth noting that neo libs (WHY “libs”!!!) still argue that the multiplier effect doesn’t exist! Anyone who didn’t understand why does now! Murphy the Subversive!
[…] does not have a multiplier effect in itself since it merely transfers the ownership of savings, which are the home of economically dead money, this 99% tax has not actually added anything to the cash pile of the government. Nor has it […]
Spot on – this is a really helpful explanation. The multiplier also functions helpfully in connection with localism – for example, the fact that a Credit Union, which lends back into its local community, boosts local economic activity, and incomes, by a measurable factor over and above that which an international bank operating in that community might. [Though I guess this effect is diluted by the fact that there may not be applicable very localised taxes]. The member-owned element further boosts this as there’s no haemorrhaging of profit to shareholders in other countries….but I think I’m going ‘off on one’, and drifting off topic. In any event…I’m keeping this article in my back pocket, as a useful explainer for the future.
Thanks
Puzzled why defence spending is viewed as having relatively little multiplier?
Surely even when you buy a weapon of mass destruction that money goes to paying to build them and those people in turn pay taxes and spend that money on other things and so on? Or is it simply the view that weapons don’t tend to be produced domestically?
With multiple layers of tax I have long believed that in the very long game everything turns into tax, but obviously some games are a lot shorter and others are a lot longer.
The defence multiplier depends solely on the onward spending of those paid to create defence. Much is outside the UK
There is no productive added value in the community – hence the limit to the possible multiplier at maybe 1
Excellent article, but I don’t think its complete enough to parry the “Ahh, but what about…” comments from detractors. Here are my questions…
1) I tell people all the time that there is no money except that which is created by government (either directly or on licence through commercial banks). So, how can it be possible that £1 spent by the Government can generate £3 in the economy if the Gov aren’t creating the other £2? (I suspect the answer is that the spending stimulates the use of savings from elsewhere (either personal or corporate) and so breathes life back into the ‘dead money’ of savings).
2) I also tell people that the Gov doesn’t need taxes to spend – in fact the tax take (which I accept has a clear and unbreakable relationship to Gov spending) is a separate factor from Gov spending altogether. Does describing the tax take in a “3 multiplier” situation as being “costless” and referring to tax as “revenues” not suggest that taxes DO pay for spending (even if ultimately they generate the taxes needed to pay for themselves)?
I regularly present a talk at work to new people which I call “How much does a nurse cost?” and I really do enjoy watching the jaws drop when I tell them the cost is nothing over the longer term, except that which said nurse can save. But I’ve always been troubled by the concept of the multiplier as it fits into the “Tax Doesn’t Pay For Anything” narrative.
Sorry if I’m fishing for free economics training, but it’s a stick the neo-libs and money-creation deniers will use to beat me with.
Ta!
The answer is that money is created by commercial, banks under government licence – and so £1 of government spend can entirely plausibly create more than £1 of economic activity because it stimulates spending through commercial banks, and also lending by them. So your point (1) needs adaptation for that. And the point you make about savings is, I think, wrong. Savings usually have nothing to do with this in a positive way.
And whilst it is true that in principle government does not need tax to spend the relationship is really not that simple. All spend is initially funded by government money creation but without tax we would very soon end up with inflation, so it is not true that the two are not related. They definitely are, but just not directly. So I am not wholly sure I agree with your point (2) either.
It’s true tax does not pay for anything. But it does make payment possible by controlling inflation and making the fiscal space for the state to spend. It’s not as neutral or useless as it sounds in your description. Sorry.
Mr Murphy
Your use of just £1 to describe the effects of The Multiplier is inspired as it makes it SOOO much easier to understand… everyone can relate to and understand a solitary quid, but not necessarily millions/billions.
Also once the concept is understood at the lowly level of a quid it becomes that much easier to apply to huge numbers. If anything it makes it inevitable, along the lines of “If that’s what can happen with a measily extra £1 government spend, just imagine what could happen with a few million/billion extra spent in the right place!!!!” I personally think this is one of the best articles you’ve written simply because it’s such an eye-opener without being a head banger too. That’s some achievement.
However, I don’t understand this bit: “Another agreement is that multipliers are higher when unemployment is higher. Again, that’s unsurprising. And, as @d_blanchlower argues, we have so much spare capacity due to underemployment in our economy multipliers should usually be high.”
That wasn’t my experience of unemployment. When I had an ok-paying full-time job I never had to worry about feeding my pre-payment leccy meter, or myself for that matter. I could afford to treat myself to the occassional cafe meal, night out, new shoes or clothes if I needed them, etc, etc, etc. But as soon I was made redundant I really had to watch the pennies and a half.
That pre-payment leccy meter was put on starvation rations so that I didn’t have to, although both the quality and quantity of food I could afford to buy reduced. Cafe meals, etc, became a thing of the past to be dreamed off. I was reduced to doing things like buying full fat milk and diluting it to make it go further. I even ended up loading my pockets with toilet roll from the local libary on a daily basis for God’s sake! In fact no public building with toilets was safe from me regarding toilet roll.
So surely high unemployment would decrease the multiplier effect as unemployed people put far less into local businesses, and even international businesses (Scottish Power got far, far less from me when unemployed)? That’s my take on it, and I don’t think I’m an exception to the rule. Although I doubt many unemployed people ended up as obsessive bog roll nickers…. or maybe they did!
Thanks for the kind comments
The reason why this works in the way I suggest is that in times of unemployment it is much easier to find people to recruit to increase the scale of economic activity
This conflicts with your micro experience – but this is a macro phenomenon and as is so often the way, they appear contrary to each other
The evidence strongly supports my suggestion
There’s another expression you could have added, Marginal Propensity to Consume (MPC) It’s the expression of how much of the £1 you’ve been given you spend. Rich people have a low MPC, because they can already afford everything they want, so don’t spend the money. Poor people spend that money because they need to. Interestingly, often at greater than 100%, because they use it to access and fund borrowing.
So the answer to rebooting the economy is simple, give money to poor people!!! Start by increasing benefits.
Agreed