From the TUC

Fiscal fallacies (2): accounting identities and the case for government loan-expenditures

10 Nov 2015, by in Economics

Related to the fallacy of deficit spending are arguments around accounting identities and the ‘principle of sectoral balancing’.

The economics is extended from the government’s balance (its deficit or surplus), to the balances for all ‘sectors’, i.e. households, companies and the rest of the world. On this basis there are four balances, though they originate in the National Accounts where they are further subdivided (e.g. into private / public, financial / non-financial corporations).

In all cases each ‘sectoral balance’ corresponds to the difference between totals payments and total revenues; e.g. for households, the revenues include salaries, interest earnings from savings, payments include taxes and consumption expenditure. For the rest of the world, the balance corresponds basically to the difference between exports and imports, adjusted for other (net) flows such as on dividends and foreign direct investment.

The logic of the argument follows from accounting considerations that dictate the four balances must add to zero. This is simply illustrated; below I have used OBR figures and added the two private sector balances together. (Strictly, the figures are National Accounts estimates of net lending / borrowing as a share of GDP.)

Sectoral balances, % GDP


(Whether the identity is simple to understand is another matter, but taken as given for present purposes.)

So over the past five years, while the public deficit has narrowed (slowly), the private sector has moved from surplus into negative territory and the rest of the world surplus has risen (i.e. the UK current account deficit has increased).

According to the ‘accounting identity’ case for expansionary policy, government spending cuts are synonymous with deficit reduction. The logic then continues: if the government reduces its deficit then either the private sector deficits or rest of world surplus must increase to maintain the balance. Given the unprecedented scale of both household indebtedness and the current account deficit, either outcome is regarded as unsustainable.

So to prevent these unsustainable outcomes, the conclusion is that the government must not reduce its deficit.

The argument is a variant or extension of deficit spending arguments, and this approach was contested in the first of these two posts.

To me the argument seems an extension of the household sector budget fallacy, with the claim that one of any number of households spending less must mean others spending more. Even simplified in this way the balanced outcome doesn’t seem obviously to follow. Then in the macro case, how can accounting identities dictate changes to the behaviour of households corresponding to the behaviour of government?

Outcomes across the whole of the macroeconomy do not follow so directly.

As in the previous post, the problem is that the impact of any change in spending on aggregate income is ignored.

The whole purpose of increasing government spending is to increase incomes, and the logic of four balances is based on actions not affecting income (‘given income’ is the jargon).

In fact, it is the change in income that that ensures the identities are maintained. This was a key part of Keynes’s original justification for public spending. Under the rival ‘crowding out’ argument, public spending would draw on private saving, so that they were no longer available to private activity. Keynes saw from the macroeconomic perspective saving was the result of public or private investment.  As he would later say,

“For the proposition that supply creates its own demand, I shall substitute the proposition that expenditure creates its own income” (Collected Writings, Volume XXIX,p. 81).

I am not sure how well the translation of this logic to the wider economy has been articulated. But plainly the starting point is very different when macroeconomic reactions on aggregate income are brought into play. The final sectoral positions do not follow from imposing a zero sum, but from understanding how the zero sum comes about through the changes to flows of incomes and expenditures for all sectors.  In the case of Keynes’s policy, the object of the exercise (both monetary and fiscal) was to revive private investment expenditures. This would suggest in the short-term a widening of the private sector deficit, in opposition to the narrowing of the public sector one.

None of this is to oppose increased spending or deny the threats to the economy from the current account or household debt on the present trajectory. But the outcomes of spending or otherwise on the part of government cannot be so simply translated to the rest of the economy. It seems odd to want to argue the case for spending in such a complicated way, but I am troubled as well by the logic – though stand very ready to hear alternative explanations.

10 Responses to Fiscal fallacies (2): accounting identities and the case for government loan-expenditures

  1. Brian Stobie
    Nov 11th 2015, 11:26 am

    Hi Geoff, I’m not sure what the ‘fallacy’ is you refer to in this article, or who makes this fallacy?

    A recent Guardian article from Ha Joon Chang and a number of noteworthy economists (OK this is an Appeal to Authority logical fallacy, but….)
    refers to the ‘principle of sectoral balancing’ in its argument against Osborne’s insane cuts, so I presume it isn’t that per se, it’s how the balances are used in argument?

    My understanding from my reading of MMT literature is that the sectoral balances describe the complete monetary economy including all sources,
    similar to an equation such as x + y + z = 0 where the terms x,y,z are defined such that the sum is always zero.
    So no-one ever argues that you have to do something to ‘make’ the the sum equal to zero, because it can never by definition be anything other than zero.

    So I’m puzzled when you say things like “Even simplified in this way the balanced outcome doesn’t seem obviously to follow. Then in the macro case,
    how can accounting identities dictate changes to the behaviour of households corresponding to the behaviour of government?”
    I’ve never seen anybody argue this, as its obviously wrong. It is correct however to flip it and say “changes to the behaviour of households corresponding to
    the behaviour of government dictate the (magnitude of) the accounting identities”.

    Also, when you say “According to the ‘accounting identity’ case for expansionary policy, government spending cuts are synonymous with deficit reduction.”,
    I’ve never seen anyone argue this either. George Osborne (for example) never refers to the ‘sectoral balances’ in his ridiculous claims, and MMT luminaries
    such as Bill Mitchell know very well that government cuts are NOT synonymous with defecit reduction and also that aggregate demand in the economy is of
    supreme importance.

  2. petermartin2001
    Nov 12th 2015, 2:16 am


    I can’t see why you have any problem with this.

    The sectoral balance analysis shows that any attempt by government to ‘balance its books’ by cutting spending and raising taxes is going to end in disaster. Its simply going to send the economy spiralling down into recession or even depression. We all knew that anyway, right?

    Sure, it might work, but that’s because we’ll all end up too poor to buy imports and too poor to save any money. That’s what’s happened in Greece as the Troika has forced them to cut their deficit!

    That means 25% plus unemployment!

    Not really what anyone wants, is it ?

  3. Postkey
    Nov 12th 2015, 10:47 am

    (S – I) ≡ (G – T) + (X – M) is an accounting identity. Say 2014.

    The equation
    (S – I) = (G – T) + (X – M) is ‘derived’ from the accounting identity, but the equality only holds when the economy is in equilibrium. That is, when planned aggregate demand {AD} is equal to planned income/output {Y}.

    If G is cut when the economy is in equilibrium, then the equality no longer holds. AD will fall {ceteris paribus} leading to a fall in Yo {and C}.
    Yo will fall until a new equilibrium is reached at, say, Y1.
    { Y1 < Yo}.
    The above equation will then hold, but with different {lower} values of S, G, T and Z {assuming I and X are exogenous variables}.

  4. Alex
    Nov 12th 2015, 7:00 pm

    Have to agree with other posters, that I haven’t seen anyone argue with sectoral balances in this way Geoff.

    Rather, the argument I’ve seen runs thus. Since the accounting identity is necessarily true, then IF the government achieves it’s aim of a lower public deficit, then either the private deficit must rise to compensate, OR our net export potion must improve. THEN the argument goes that well, we’re not exactly likely to see more exports given that out big trading partners like the Euro zone and China are in a mess, THEREFORE if the deficit is to come down in the way planned, the private sector must get into deeper debt.

    From this, the argument can run two ways. FIRST that if this happens it’s dangerous since private debt is still quite high after the end of the last crisis and having a government policy such that you hope the private sector borrows more is foolhardy. SECOND instead that whatever the dangers, that this is unlikely to happen because people would pay down debt and not borrow more.

    If you take the FIRST line of argument you have a clear argument against government policy – more private debt and, as usually happens eventually, another banking crisis from that, purely because the government decided to raise taxes/cut public spending. And if you take the SECOND line of argument, you’ve then hypothesised that net exports won’t improve and the private sector won’t start borrowing again, so EITHER the public debt won’t come down as the government claims it will (austerity is self-defeating), or if it does then it does so in a way that reduces GDP (austerity makes us poorer).

    That’s usually the rough contours of sectoral arguments I’ve seen anyway.

  5. petermartin2001
    Nov 13th 2015, 1:03 am

    PS Can anyone provide a data source for the graph in this article?


  6. Geoff Tily

    Nov 13th 2015, 10:26 am

    I took the figures from the Office for Budgetary Responsibility. They publish an excel spreadsheet with an immense amount of background detail on their forecast. From this page in the section ‘supporting documents’ go to supplementary economy tables. Within the spreadsheet it’s table 10.

    The alternative is the National Accounts; the most useful is the quarterly UK Economic Accounts publication:

    You need the ‘income and capital account’ tables for each sector, and the relevant aggregate is net lending / borrowing. Though these balancing items are collected together on table 1.6.B9.

    The definitions underpinning these measures are different (in that they cover all flows of revenue and expenditure, or resources/uses including transfers) to the equations on the various comments I have received here (and on Richard Murphy’s blog), a point of interest I think.

  7. Fiscal fallacies and sectoral balances: response to the critics
    Nov 19th 2015, 6:30 pm

    […] (‘Keynes wanted government loan expenditures NOT deficit spending‘ and ‘accounting identities and the case for government loan-expenditures‘) have generated a number of comments, including on Richard Murphy’s blog (here).  […]

  8. Postkey
    Nov 24th 2015, 9:20 am

    “I was also referred to a recent blog written by the senior economist at the British TUC – Fiscal fallacies (2): accounting identities and the case for government loan-expenditures – which appears to entertain the view that the sectoral balances framework provides a “case for expansionary policy”.

    Both these inputs are unhelpful.”

  9. Geoff Tily

    Nov 24th 2015, 9:28 am

    I am grateful for this detailed response, but will not get the chance to consider properly until after the spending review. In the meantime, ‘Postkey’ might like to take a look at my previous response to comments so far: