Short & Simple 9 – Spending Determines Income

We understand that, as a rule, total spending must equal total income (this was explained in part 4 of the series). But this raises a question. Is it spending that determines income or, instead, income that determines spending?

At first glance, it might seem plausible that income determines spending. After all, when an individual household receives income, it will invariably spend some of it. If a member of the household receives a pay increase at work, the household will have more income. It is quite likely the household will respond by increasing its spending, perhaps after a delay. It can now afford to spend more than before.

All this is true. But notice that the effect of income on spending is not automatic. A household might decide to spend all of its after-tax income. Or it might decide to spend some and save some. And the fraction it chooses to spend could be very high or not so high. Or, for that matter, an individual household might decide to spend more than its income. It could do this either by drawing down savings accumulated in previous periods or by borrowing.

It is actually spending that is most subject to choice. A household or business can decide to spend more, but its ability to receive more income depends to a considerable degree on the decisions of others.

A worker might like to earn more wage income, but the employer will have to agree to pay the higher wages or offer extra hours of work before this can occur.

Likewise, a business might like to make more sales, and step up its sales efforts accordingly, but whether it succeeds in its aim will ultimately depend on the spending decisions of its customers.

When somebody decides to spend, this causes income to be received by somebody else. The spending creates income.

This idea is sometimes expressed as “spending determines income”.

This is only possible because some spending can occur independently of current income. For instance, a household might spend in excess of income by drawing down savings or borrowing.

Share

11 thoughts on “Short & Simple 9 – Spending Determines Income

  1. It has to be spending before earning. We are users of currency. If the government fails to spend there will be a shortage of currency and so spending will automatically be constrained. This is what budget surpluses do to the economy.

  2. Do not forget government spending, which is a huge chunk of the total spending of a country. Government spending generates income to house holds and business too.

    The central (federal) government has much more discriminatory power over spending than a household or business. It doesn’t need necessarily to “draw down savings or borrow”. It can print money out of thin air. All it needs is an approved budget to do so.

  3. Pete,

    By now, my praises should sound repetitive. I’ll say it, anyway: very good and instructive.

    Let’s focus on this bit, which I found very telling:

    It is actually spending that is most subject to choice. A household or business can decide to spend more, but its ability to receive more income depends to a considerable degree on the decisions of others.

    Now that we are speaking of households, let’s think of the utility maximisation problem of microeconomics. I’m sure this all is very familiar to you, but I’ll beg your patience. This is the budget constraint in the simplest such model, corresponding to household 1 (thus, the subindex)

    Px*X[1] + Py*Y[1] + Pz*Z[1] = M[1]

    X, Y, and Z are the quantities of three goods. Household 1 decides how much of each good it buys. They are what some call “instrumental variables”. The P’s are their respective unit prices and are common to all households, and M is household 1’s money endowment in $. Both the P’s and M are parameters, they are exogenously given: constants.

    The LHS gives the total spending possible to the household, the RHS is its income: in this most basic model spending equal income. This is compatible with your description, yes?

    A second household could have the following budget constraint, which like the previous one, seems compatible with your premises:

    Px*X[2] + Py*Y[2] + Pz*Z[2] = M[2]

    ———-

    Then, you write (and this is where I have difficulty):

    When somebody decides to spend, this causes income to be received by somebody else. The spending creates income.

    I might be mistaken, but I interpret that what you say in the quote above is that in fact, M[1] and M[2] are not constant. Household 2’s X[2], Y[2], Z[2] affect M[1]; likewise, X[1], Y[1], Z[1] affect M[2].

    Sure, after a while that may be true: one household’s consumption decisions lead to higher sales, which leads to higher production, which leads to more work and income for the other household. But that income variation only occurs after they have made their decisions. They do not have an immediate effect on each other’s income and even if they did, I can’t see how the households could assess it.

    Moreover, that goes on: X[1], Y[1], Z[1], affecting M[2], then also affect M[1]; similarly for M[2].

    In practice, to me (and correct me if I’m missing something) this sounds pretty much like denying that there is any budget constraint.

    If that conclusion is more or less granted, then the question is why is there any budget constraint at all?

  4. The “magic money tree” has come in for adverse comment recently. Nobody wants to admit that it exists and that it is the sole property of currency ISSUERS, more than likely a sovereign nation’s Treasury. As a picture speaks a thousand words, have a look at https://www.quora.com/Why-are-there-terms-Horizontal-and-Vertical-money The Treasury including its own central bank – they are one and the same – spends (issues) its fiat money (units of account), brand new every day. Taxation is used to drain the units of account from the currency users, to prevent the resources available in the economy being overbought (inflation). Taxes are just the brake pedal, they don’t fund anything, they are not recycled for further government spending.

    Now, have a look at the latest “Whole of Government Accounts” (WGA) https://www.gov.uk/government/publications/whole-of-government-accounts-2015-to-2016 See if you can spot where the “magic money tree” appears.

    According to the UK’s IFRS prepared WGA accounts – same system as any multinational corporation – the nation is insolvent; liabilities much greater than assets; huge negative working capital; has been for years. This is more easily seen in the key facts and figures from 2013/14 https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/445748/PU1786_WGA_summary_report.pdf How come it is still trading? That’s the beauty of being a currency issuer with your own magic money tree, with an infinite number of money leaves available every day.

    BTW. The government’s Treasury always spends twice. When it keyboards your pension into your bank account (a liability for your bank); it effectively keyboards the same amount into your bank’s “reserve” account at the BoE, (an equivalent asset for your bank’s balance sheet). All horizontal money assets and liabilities have to balance and hence sum to zero.

  5. Hi Magpie. This is a good question for further teasing out the explanation. I haven’t yet got to most of the following in the “short & simple” posts. All this and more will be covered at some point.

    The initiating acts of spending are those that are autonomous of the circular flow of income. The expenditure of a currency-issuing government (as others in the thread have noted, and I will get to later in the series) is fundamental, because this spending is never subject to a budget constraint (despite what neoclassical theory pretends). Members of the non-government can also spend autonomously through borrowing. Unlike a currency issuer, they are (at the micro level) ultimately constrained in their capacity to borrow by their prospective future income streams. Nonetheless, inside this capacity to borrow, there is a considerable degree of autonomy in their spending decisions, since they could choose to borrow less than possible or as much as possible, or not at all.

    But the point made in the post is not only about spending through currency issuance or private credit creation. The point also applies to the decision of how much to spend out of income received. In other words, the receipt of income does not automatically (and may not at all) result in spending. The income, once created, can be saved. In contrast, an act of spending always creates income of the same amount, directly and immediately.

    Strictly speaking, decisions to spend occur at every chain of a multiplier process, even though (for good reason) we often treat spending after the initial autonomous act of spending as being induced out of income. For example, suppose the government spends $100. This immediately creates $100 of income for the recipient. (It’s not the focus of the present post, since we are talking here about income, but this act of spending also involves the crediting of a reserve account — the creation of new ‘government money’ — and the crediting of a commercial bank deposit, which is an expansion of broader measures of the ‘money supply’.)

    If the process stopped at this point, it is true to say that spending created income, rather than the other way round. But it remains true for all acts of spending. Let’s say the recipient of the government spending wishes to spend some of the new income received. Although, for analytical purposes, it often makes sense to assume that a particular fraction is consumed out of income (a particular marginal propensity to consume), in actuality the spending recipient — as with any spending recipient — gets to decide how much to spend out of the new income. Perhaps this particular spending recipient immediately uses $50 to spend. If so, the act of spending creates another $50 of new income. If instead the spending recipient had decided to spend $30, only $30 of additional new income would be created. And, of course, recipients of this subsequent spending have similar decisions to make.

    If spending recipients, on average, decide to spend a small (large) proportion of any income they receive, less (more) income will ultimately result from an initial act of autonomous spending. Although we often theorize this as a multiple of the autonomous spending (as if it is the initial autonomous act that automatically generates a multiple of income), strictly speaking it is the entire sequence of decisions to spend (or not) that determines how much income is created.

  6. Magpie, just to address the household budget constraint aspect explicitly:

    In micro, it can make sense to consider a household’s choices taking its income as given, ignoring interdependence with the rest of the economy for simplicity. This can work because a single household is small relative to the economy as a whole, and its decision is economically insignificant from the macro perspective.

    If we also constrain the household to spend precisely its given income, then it is true that the household’s income will equal its spending. But, in general, there is no necessity that any one household, firm or sector of the economy spends an amount equal to its income. The rule that spending must equal income is an aggregate one — total spending equals total income.

    Since in the neoclassical household budget analysis the household’s income is simply taken as given, the question of where its income came from, or how it was created, does not arise. But what we can observe is that the household is taken to have this income. It can now choose how to spend it. When it does spend, it will create new income equal to the amount the household chooses to spend, and a firm will now be faced with a similar question of what to do with this new income.

  7. John, André, acorn: You have all zeroed in on a point that I agree is key when it comes to thinking about causation. It is a point I am getting to in upcoming posts. Basically, spending can only be said to be the cause of income if at least some spending is autonomous (and so subject to choice). That is, it must be possible for some spending to occur independently of income. In a modern money system, government spending certainly can occur independently of income. Private spending through borrowing is another source of autonomous spending. For a nation, exports are another significant source of autonomous demand.

  8. It is worth examining a couple of fundamentals. The first being the spending multiplier. A fiscal stimulus, an injection of government spending into the economy, will suffer some leakages as it bounces around the economy facilitating transactions all over the economy; being taxed every time it moves, until the Treasury eventually gets all its spending back via taxation and charges. The leakages will be taxation; savings and imports. Prof Bill Mitchell explains at http://bilbo.economicoutlook.net/blog/?p=6949

    The second is the sectoral flow balances that show which sectors are spending more than their income; and, which are saving more than they are spending. Neil Wilson’s 5 sector chart at https://docs.google.com/spreadsheets/d/1ktX29jafu9iV3oFZ8GVLtAHaFmUkikdfYrw_BRPFUTU/edit#gid=3 is better than the ONS equivalent.

    You will see that the Treasury was running a 10.68% budget deficit in 2010 Q1. That deficit was paying for the imports and the panic savings of both the private sector non-financial firms and households. The banks and financial firms put on their tin hats and went into lockdown.

    You will see that in 2017 Q1 and the three Qs previous, that households have started “deficit spending”, just like the Treasury. Hence, the Osborne July 2010 austerity budget plan for transferring government sector debt into private sector debt, is working out quite well.

    Alas, the government’s Treasury has an infinite supply of Pounds Sterling – it being the currency issuer, it will never run out of them – everyone else on Neil’s chart is a currency, user, they don’t have a magic money tree.

  9. Thanks for your thoughtful replies, Pete.

    To avoid misunderstandings, here I will attempt to clarify two things. I’ll think about your second reply and comment on it later.

    One: I’m not talking about federal government budget constraint in a country with fiat currency. I don’t dispute MMT’s conclusion that they are unconstrained.

    So, I have no issue with the multiplier mechanism, with fiat money. It’s true, like you say, that For example, suppose the government spends $100. This immediately creates $100 of income for the recipient. If the [multiplier] process stopped at this point, it is true to say that [fiscal deficit] spending created income, rather than the other way round.

    I repeat: I’m cool with that.

    Although perhaps my doubt could extend to levels of government other than the federal, I am not talking about them either. My doubt is with households and firms only: members of the non-government, users, not issuers, of money, whether fiat or not.

    So, in the multiplier mechanism you mentioned, my doubts start here: But it remains true for all acts of spending. Let’s say the recipient of the government spending wishes to spend some of the new income received.

    My contention there and in general is that income (the recipient’s) is logically previous to spending (which I believe you were heading to in your second reply).

    Two: It’s okay if you prefer to leave this for the future. Don’t feel pressed to reply.

    ———-

    I also understand that

    Members of the non-government can also spend autonomously through borrowing. Unlike a currency issuer, they are (at the micro level) ultimately constrained in their capacity to borrow by their prospective future income streams. Nonetheless, inside this capacity to borrow, there is a considerable degree of autonomy in their spending decisions, since they could choose to borrow less than possible or as much as possible, or not at all.

    This, in a simple two-period model where the interest rate available to a household is R and its future income is N means that, inclusive of borrowing, the new money endowment (M’) could at most be

    M’ = M + N/(1 + R)

    In other words, the current endowment may vary between M and M’. The household chooses one and plugs it in its budget constrain as presented before. Once chosen, however, the endowment is still a constant. Income (current and future, in this case) still constrains its spending level. Income remains logically prior. No?

  10. Thanks for your clarifications, Magpie.

    Okay, if I’m understanding correctly, you have no issue with the idea that spending is autonomous when: (i) carried out by a currency-issuing government; or (ii) achieved through private borrowing (up to some limit).

    I know you realize this, but I’ll state it anyway in case it is not obvious to all readers. In the case of autonomous spending financed by private borrowing, the initial deposit created by the loan is not yet income. It is the act of spending on goods or services, by drawing down the deposit, that creates income. This is why spending through private borrowing can be regarded as autonomous (up to some capacity to borrow).

    Magpie, I think we agree on that, but you are pointing out that induced consumption comes out of income that is received prior to the spending:

    My contention there and in general is that income (the recipient’s) is logically previous to spending (which I believe you were heading to in your second reply).

    Yes. I can agree with that, except I would stress that: (i) the income of that household, which already exists and is yet to be spent, is the result of earlier spending; and (ii) the household spending about to take place, which is induced out of the income already received, will create new income for the recipient.

    Do we both agree on this?

    I had written a long response, but upon re-reading your latest comment, I sensed that most of it might just be repeating stuff we already both agree on.

    I’ll just summarize by saying I agree that some spending can appropriately be considered induced and that this comes out of income (though, as just noted, I would insist that the income it comes out of is the result of prior spending). Two conditions need to be satisfied if income is to be considered demand-determined. In the behavioral equation:

    Total Income = Autonomous Spending / Marginal Propensity to Leak

    total income can only be considered demand determined if:

    1. Autonomous Spending > 0; and
    2. Marginal Propensity to Leak > 0.

  11. Thanks Pete.

    You replied to all my doubts. A sensation, however, lingers that both of these things are true: (1) the Government is unlike households, but (2) households are not really entirely unlike the Government.

    Oh, well 🙂

Leave a Reply

Your email address will not be published. Required fields are marked *