I wrote recently about why I felt fiscal rules were bad news for any government seeking to offer a Keynesian / left of centre approach to economic management using a narrative logic to support my arguments. When doing so I suggested I might also offer a mathematical justification for the position I have taken. Let me do that now.
The logic that follows is based upon what is usually called national income accounting: what is offered are accounting identities that have to be right as a matter of fact; that is why they are called identities.
The analysis starts by looking at the composition of GDP from a number of perspectives. One is these is:
GDP = C + I + G + (X — M)
What this identity says is that total national income (GDP) is the sum of total final consumption spending (C) plus total private investment (I) plus total government spending (G) and the sum of net exports (X — M) where (X) is exports and (M) is imports.
Why the identities are useful is because GDP can be looked at it in more than one way, As a result this identity is also true:
GDP = C + S + T
This identity says that GDP is made up of household consumption (C) (as before) plus total saving (S) and total taxes paid (T).
It's pretty straightforward to then say that:
C + S + T = GDP = C + I + G + (X — M)
And without too much effort it's also possible to reorganise the expressions from this point to eliminate GDP from the equation and consumption (C) as well as it appears on both sides of the identity and so cancels out. The result is the following expression which is generally said to represent the sectoral balances within the national accounts, about which I have written quite often on this blog:
(I — S) + (G — T) + (X — M) = 0
In effect the the three sums have to balance to zero because they represent financial flows that are either net saving or borrowing and as a matter of fact every saver needs a borrower in double entry book-keeping terms.
It's important to note what expressions mean though. (I — S) is in effect the private sector balance: if investment exceeds saving the sector is borrowing and vice versa.
(G — T) is the government sector balance: either government spending is more or less than taxation.
And again (X — M) is the current account balance with the rest of the world and represents whether we are saving abroad, which happens when outflows including for goods, services and finance (X) exceed (M) for inflows, or vice versa.
To this point I should add that there is nothing exceptional about what I have noted: this is pretty basic economics. I have no idea whether what follows is normal or not: I haven't bothered to search the literature to find out. It's my analysis using the arguments to this point as a starting point.
Remember what I am trying to analyse is the impact of a fiscal rule. That rule can be fairly easily inserted into these identities. The fiscal rule being considered is that which says government spending in current items, which I will call (Gc) should balance tax income (T) but that government spending on investment, which I will call (Gn) may be funded by borrowing.
To put this another way:
G = (Gc + Gn)
and
(Gc - T) = 0
In that case this identity (for that is what fiscal rule has made it) can be cancelled out of this expression:
(I — S) + (G — T) + (X — M) = 0
which now becomes:
(I — S) + (Gn) + (X — M) = 0
This has now to hold true. Rearrange it and the following can also be said to hold true:
Gn = (S - I) + (M - X)
This is, I suggest, when things become interesting. What this says are are a number of things.
The first, and most obvious, is that government investment does not impact consumption. (C) is not a component in this identity.
Second, the identity says that government investment is funded by one of:
- An increase in private sector saving: this would be represented by new bank deposits if cash was created to pay for the investment or by new holdings of bonds sold to fund that expenditure; or
- By a reduction in private sector investment: this could be said to be the classic government spending crowding out private sector activity argument; or
- By a net inflow of funds from importers or overseas investors taking advantage of the bonds made available to fund the government investment; or
- By a reduction in exports, which would imply that productive resources otherwise diverted to this activity was instead allocated to the creation of the newly funded government infrastructure.
You can play around with these explanations, but I suggest that these are sufficient for the purpose of this analysis.
There is one important thing to add at this juncture though, based on the economic climate that has prevailed since 2008, which cannot be ignored if we are going to suggest which, if any of these factors is more important than the others. This additional factor is that there is now and has been for some time spare capacity in the UK economy.
There is argument about how much spare capacity there is: I tend (like Danny Blanchflower) to think that actual spare capacity is much bigger than that which the Bank of England estimate. It does not greatly matter for this argument; what I think is clear is that quite considerable sums could be expended by any government on investment activity right now without in any way reducing the capacity of the private sector to invest (which it is showing some reluctance to do) or impairing the UK economy's capacity to export.
I could go into differential calculus here but I won't because at that point most who have reached this point will depart the scene and that is not my wish. Suffice to say that whilst it may be rash to say that in the last noted identity that there is no possibility of a relationship between a change in (Gn) and either (I) or (X) that it is also not unreasonable, given the point just noted on excess capacity in the economy, to suggest that the relationships in question are very weak indeed. In fact, I'll suggest that they are likely to be so insignificant that for all practical purposes it can be said that in the identity in question that for all practical purposes:
I = 0
and
X = 0
implying that a change in government investment will have no impact on these two variables, which then leaves the identity as:
Gn = S + M
There should, perhaps, be no great surprise in this conclusion. Modern monetary theorists have long said that government debt is the same as private wealth. And that is exactly what this expression confirms: the net effect of borrowing to fund government investment (Gn) has been to either increase UK private savings (S), which can be owned by either individuals or companies, or to deliver an increase in the sums saved in the UK from those overseas (M) which will be reflected in the current account deficit. What will not change is (C), which is the apparent well being of households.
The proof of this might be considered to be the evidence from the economy. For more than a decade there has been government borrowing funded by deficits. Now admittedly there has been current government spending (Gc) funded by deficits as well (for a summary of the data to 2015 see here), which confuses the picture (I admit) but what is clear from the most cursory of headline data is that there has been GDP growth but that this has not fed through into wages which have been declining whilst private sector wealth and the trade deficit have grown.
In other words what the identity suggests has happened: what has been considered to be desirable investment has not lead to a growth in net consumption, which is what maters to most people. That;s not to say that there has been no growth, but most people have not benefitted.
The issues arising from that suggestion and most especially what can be done about (because I think there are actions to be taken that can have beneficial impact) need further theoretical exploration, which should follow (other work and holiday permitting) but for now I return to my opening thesis, which is that a fiscal rule of the type that has been proposed by John McDonnell and which was previously imposed by Gordon Brown on the apparent advice of Ed Balls will not assist those that Labour intend it to help. That fiscal rule will increase private sector wealth which, as we know, is poorly dispersed across the economy resulting in a cost of increasing inequality in the UK. And such a rule may make the UK feel as though it is increasingly dependent for anything to happen on an inward flow of foreign funds (sound familiar, anyone? Just think Hinckley C) but by itself such a rule will not deliver for the ordinary people of the UK.
This is not to say there is no growth: if the result of the investment is put back into this identity:
GDP = C + S + T
then S has grown, and so therefore has GDP, but the reward is going to savers with a low marginal propensity to consume and low effective tax rates (because that is UK policy on saving).
I should warn: of course these identities ignore spillover effects, of which there will be some, but the point I am making remains valid as far a I can see: a fiscal rule of the sort that John McDonnell has adopted permit investment but is likely to deliver many of its benefits to those with the greatest existing wealth. Unless linked to further policy measures (and this is the issue I am working on) that means the question has to be asked as to why Labour, irrespective of leadership, has been so keen to use such rules?
Comments are of course welcome. This has not been peer reviewed so I may have missed something glaringly obvious. But if so please tell me what.
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:
I’m not sure if you read FB comments – so reposting here: you’ve stated that every borrower has to be balanced by a saver – but that has been debunked (Steve Keen, Richard Werner, Ann Pettifor and others) and instead money understood to be created by debt (no savers needed to create new money) – how does that fit in with this analysis? I think that is almost always the elephant in the room.
They do not disagree with me I am sure
If all money is debt – and it is – there has to be a lender and a borrower or there is no debt
So the logic I present is right
Is new money created by the central bank debt ? E.g. if the BOE were to credit my bank account with a £100 today, who would incur any additional debt ?
Theoretically yes
But I have not got time to explain why and get away on holiday
Thanks Richard, it’s a vivid explanation. Does this imply that without a fiscal rule government investment (Gn) should be balanced by increased taxes (T) in order to avoid impacting the private and foreign sectors?
I.e. assuming:
(I – S) + (X – M) = K
K = (Gc + Gn – T)
where “K” is constant.
If (Gn) increases then (T) must also increase in order to avoid pro-cyclical impact on fiscal spending (Gc).
I’d love to read more about this, please could you point me at a text book?
Simon
No, the absence of a fiscal rule does not necessarily mean that
Give me time and I will explore what it might mean: without a rule more variables float free
And the point of Gn is that we we may want to really promote a spend
The issue is how to prevent is simply increasing S or M
Ample stuff on this on the web under national income accounting
Richard
To a dunce who has long wrestled (largely unproductively) with the learned opacity of much MMT exposition, this of yours comes as a shaft of light. Thank you.
“…the question has to be asked as to why Labour, irrespective of leadership, has been so keen to use such rules?”
My suggested answer would be:- because Labour has become sensitised to, and extremely wary of providing ammunition for, accusations of fiscal irresponsibility/incompetence – due to having languished in the wilderness for considerable periods in the recent past as a result largely of having had that albatross hung around its neck by its opponents. The blogosphere is still full of it, like a folk-myth.
So, in other words, I’m suggesting it’s no more than a tactic, camouflage if you will.
It may be
But people don’t want lies now
They need to be told what can be done and how
I have more respect than such misrepresentation might imply
I guess you’ve already seen this: https://medium.com/@james.meadway/richard-murphy-and-fiscal-credibility-rule-a-reply-156e701fb850#.cwf77r6ur.
I have
I will comment later when I’ve stopped laughing
None of this is really controversial of course, Richard. It is pure MMT:
(S — I) = (G — T) + CAD
which is interpreted as meaning that government sector deficits (G — T > 0) and current account surpluses (CAD > 0) generate national income and net financial assets for the private domestic sector.
Conversely, government surpluses (G — T < 0) and current account deficits (CAD < 0) reduce national income and undermine the capacity of the private domestic sector to add financial assets.
Expression (5) can also be written as:
(6) [(S — I) — CAD] = (G — T)
where the term on the left-hand side [(S — I) — CAD] is the non-government sector financial balance and is of equal and opposite sign to the government financial balance.
This is the familiar MMT statement that a government sector deficit (surplus) is equal dollar-for-dollar to the non-government sector surplus (deficit).
The sectoral balances equation says that total private savings (S) minus private investment (I) has to equal the public deficit (spending, G minus taxes, T) plus net exports (exports (X) minus imports (M)) plus net income transfers.
All these relationships (equations) hold as a matter of accounting and not matters of opinion.
I said all that
Then I pointed out something that as far as I know MMT has not said
But MMT is controversial, isn’t it. See http://www.thomaspalley.com/docs/articles/macro_theory/mmt.pdf.
Yes
But this is not MMT
This is bog standard national income accounting
Richard, I have a methodological point to make. Economics, like all the humanities, has attempted to clothe itself in the respectability of physics, and nowhere is this more apparent than when a humanities discipline uses algebra. Unfortunately what is missed in this effort is the realisation that terms in physics are defined uniquely, precisely and for ever. Here you use two different definitions of GDP and then write: C + S + T = GDP = C + I + G + (X — M). You couldn’t make that mistake in physics because a quantity termed “GDP” would have a single definition and would always have the same numerical value when measuring the same physical phenomenon, within the precisely known percentage errors of measurement. In your example the two GDPs would have different values if measured for any given period for any given economy, and so your identity is false.
Hence when you land up with (I — S) + (G — T) + (X — M) = 0 you embark on a compounding of errors. One of these is a classic MMT mistake, that negative (I – S) equates to government borrowing. In reality, even just taking the private banking sector (Monetary Financial Institutions in BoE terminology), savings are vastly greater than investment, yet the holding of government bonds by this sector is less than 10% of the balance sheet. The bulk of savings are not borrowed by government but are borrowed by the private sector for the purchase of assets.
You say: “Modern monetary theorists have long said that government debt is the same as private wealth.” Certainly, governments borrow from the private sector. But private wealth is vastly greater than government debt.
This does not necessarily invalidate your ideas on fiscal rules. MMT is just not a helpful way to get there.
But economics is not a physical science
And as a matte of fact GDP can be measured in different ways
So the suggestions made are not only made, but appropriately made and there is no compounding of errors
Nor is there any claim that I – S = G – T. You miss X – M in the process and cannot do that
In other words, I think you may be trying too hard
In 1998 UK imports and exports last balanced, so X-M = 0 at some point in that year. Are you seriously saying that S-I = G-T when X=M? That government borrowed all private-sector savings bar that which was used for private sector investment? (Note that you have to write either -(I-S) = G-T or S-I = G-T when X=M, following from your equations.)
X – M includes financial flows
‘Certainly, governments borrow from the private sector. But private wealth is vastly greater than government debt.’
most of that ‘wealth’ is created by bank debt. MMT says ONLY G can create net assets. None of the spending creates real net assets. This is fundamental.
see:http://neweconomicperspectives.org/2010/11/yes-deficit-spending-adds-to-private.html
This ‘fiscal rule’ business strikes me as bullshit and bizarre obfuscatory accounting.
We know from the example of Japan that the Central bank can keep buying (monetising) debt until the cows come home with huge debt/GDP ratios.
The real question is always: Inflation and resources available-nothing else matters-so why people like Meadway are doing a dance on a pin head I just don’t get.
Nope
Me neither
Richard, I’m afraid you cannot draw these kinds of conclusions by rearranging accounting identities.
You write Gn = (S — I) + (M — X), and interpret the absence of C as meaning government investment cannot increase consumption. But, since GDP = C+S+T you could equally have written
Gn = (GDP – C – T — I) + (M — X)
and say that government investment cannot affect savings. Both statements are wrong.
There is plenty of literature around the relationship between government spending/investment and consumption: some is empirical, some theoretical, but none relies upon accounting identities!
I could and will play with the identities
And I have not for a moment denied that there is a relationship between government spending and growth – indeed, I acknowledge there is one.
But there are two reasons for looking at things as I did
First they make sense in terms of the rule I was discussing: it is offered an identity in the form I used
Secondly, the conclusion I reached appears to reflect known outcomes which makes it worthy of consideration as an explanation (no more) of what is happening and how at present the net benefit from a particular type of spending might flow
I think that playing with the identities might show how the outcomes may be changed – I have just not had time to do that as yet and may not before I go on holiday
But it does not mean that looking at the identities and what they suggest is in any way wrong unless you think that to explore ideas in new ways is an error in itself
This seems to prove no more than if you assume consumption is a fixed parameter, then it is. Your model seems to allow no room for ‘multiplier’ effects resulting from investment that would reduce unemployment, nor for the positive effect on wages that would result from that. You rightly point to excess capacity but do not seem to have built this into your model. Depending on the type of investment, there could also be a positive effect on consumption through cost reduction, e.g. housing.
I accept the points, especially on the multiplier, which I am aware that this does not show
But what the model does show is that in a very particular set of circumstances (those we have) the multiplier may be as much outside as in the jurisdiction with the remainder of the overall benefit going to savers
It is worth pointing out that the literature does not look at multipliers with a fiscal rule of the type suggested by the way. No one has delivered one. I could, of course, have noted that they can’t, but that’s a minor sideline point!
One thing I’m confused about is you say;
“The first, and most obvious, is that government investment does not impact consumption. (C) is not a component in this identity.”
If this is the case then why use PQE? Why do governments spend in a downturn if it doesn’t encourage consumption? In fact what benefit is there to the economy of opposing austerity (there are obvious social benefits)? My understanding was the opposite, that spending does impact consumption.
If I get time before I go on holiday I will explain
C can be included – but not with that fiscal rule is my point (when I get to it)
And the whole argument is about that rule
I hopefully have followed the argument, but not sure I completely agree:
Hopefully this is the argument: Given the national accounting identity +Gn will = more saving, which will benefit savers but not more important parts of the economy.
Why I don’t completely agree:
You’ve argued that given the fact that S increases, because GDP = C+S+T, therefore GDP would increase. But given C+S+T=GDP=C+I+G, there would then surely be an increase in C+I+G.
Seeing as we wouldn’t expect there to be an increase in G in this instance, then this means either C or I would increase given an increase in S.
Which is why I don’t completely agree and why I would argue that it is not the case that I (or C) would = 0.
Also I’m not really sure why the argument is about the fiscal rule (ie that Gc should = 0). As it seems to me more about an argument against the other side of the policy, that the government should increase investment. As far as I can tell, what you are arguing is that essentially this wouldn’t affect the real economy ie C & I. (which seems intuitively hard to believe and as I’ve suggested might not be the case)
Gn would be the increase in G
That’s because Gn = Gc = G
So the increase in Gn would be the same as the increase in G
There is then no spillover as you assume
I accept that the system is not as rigid as the parameters imply, but I do not think you have found the flaw
My argument would be that it is pretty arbitrary where you put Gn in the scheme.
The reason being it can be seen as a one off injection into the economy. Eg. lets say to build (unfortunaley HS2, or more hopefully HS3).
So lets say the injection is £70bn, then that money which is Gn also gets spent by private sector contractors such as engineering firms/train makers for instance, in which case you can also class it as I.
So there is no reason especially to class it as an increase in G as opposed to an increase in I.
Sorry, but that’s nonsense
If you change the definitions you can say anything you like, but it’s no longer meaningful
Fair do’s I was assuming that the govt would simply print the money. Rather than create it out of savings. Which is I think why I was getting a bit confused.
Sorry busy day. Looks Ok but will go through the maths with a fine tooth comb and simultaneously write it up in LaTeX with proper math typesetting – I find it much easier to read. Will try to get it done tomorrow – will link to a PDF and will send you the LaTeX source code if you PM me.
Sean
Thanks – but I am quite sure this is not the final version. Might I suggest waiting?
Best
Richard
No worries; I normally go through dozens of drafts in any event. LaTeX lends itself to cut and paste. Will certainly hold fire on anything like a final version. I wish I could write 8000 words on a good day like you! I often find that the process of writing equations down carefully helps my thought process. I’m interested in reconnecting with economics; I lost all interest when Neolibearalism came into fashion.
8,000 is not common
6,000 is though
Thanks for your interest
For what it’s worth I have crudely typeset the analysis. The equations are OK as far as it goes but as you have indicated this is far from a final version; definitely needs more work
http://www.stbarnabaschapel.net/Murphy.pdf
Have a great holiday and relax for a few weeks
Sean
Many thanks
Appreciated
I fear that may travel with me….
Richard