The double-sided excuse

At Mike Norman Economics, Tom Hickey links to Erroneous Use Of The Sectoral Balances Identity [Updated] by Ramanan.

Ramanan reviews Andrew Lilico's Can public sector austerity coincide with private sector austerity?

Ramanan's post is impressive. Andrew Lilico's is not.


Lilico writes:

When we talk about "private sector deleveraging" what do we mean? We mean things like households paying off loans to the bank, or corporates paying off bonds or other loans. The vast, vast majority of such loans are loans private sector agents make to each other. So for every dollar reduction in borrowing made by one household or company, there is one dollar fall in savings by other households and companies.

(Forgive me my Americanisms. Andrew Lilico thinks in British pounds, but I think in dollars. Two places I had to change his word "pound" to "dollar" to help me think.)

Ramanan astutely observes:

Lilico confuses the terms borrowing and saving

He certainly does: For every dollar reduction in borrowing made by one household or company, there is one dollar fall in savings by other households and companies.

If only. If the reduction in demand (caused by the reduction in borrowing by some) was counterbalanced by an increase in demand caused by the reduced saving of others, there need be no recession, no output gap, no increase of unemployment.

Andrew Lilico's error is comparable to Eugene Fama and John Cochrane's as described by Paul Krugman:

Fama and Cochrane are asserting that desired savings are automatically converted into investment spending, and that any government borrowing must come at the expense of investment — period.

But in Lilico's case, the error is created by his use of the word "savings" when the correct word would have been "lending": For every dollar reduction in borrowing made by one household or company, there is one dollar fall in lending by other households and companies.

That much is true. But a reduction in lending is not the same as a fall in savings. A fall in savings implies money moving out of savings and into circulation without an act of lending. It implies a dollar-for-dollar substitution of demand by savers in the place of demand lost when spenders borrow less.

Lilico's simple word-substitution error explicitly describes this dollar-for-dollar shift in demand. Since it is dollar-for-dollar (or pound-for-pound) as Lilico describes it, GDP does not fall as a result of the reduction in borrowing. Output does not fall.

Ramanan writes:

The most fundamental error of Lilico of course is that he holds output constant in his entire argument. When discussing a scenario with sectoral balances, it is also important to keep in mind the behaviour of output.

Amen to that.



Ramanan also points out that

Lilico’s argument seems to think of the budget deficit as exogenous – i.e., under the control of the government but a careful study reveals that this ain’t so.

Yeah... But this depends on one's point of view. The people who call for austerity in government spending obviously think government spending can be cut and cut again. And obviously it can be done: It is being done. So I think Ramanan's argument here is weak. Remember, the object is always to convince people who disagree. Never only to preach to the choir.


My criticism of Andrew Lilico's article is unlike Ramanan's.

Lilico describes the sectoral balances identity:

...whatever the government doesn't borrow from the private sector in its own country it must be borrowing from foreigners.

He calls this "trivial". Then he allows a simplification: "we assume the position relative to foreigners doesn't change." Now the trivial has been reduced to pristine simplicity:

That implies that any reduction in government borrowing must precisely be matched by a rise in household borrowing. Conversely, any fall in household borrowing must be precisely matched by a rise in government borrowing.

Couldn't get much clearer than that. Even I understand it: A given economy, with its particular monetary balances, requires a particular level of borrowing for markets to clear. Notice in the above excerpt only borrowing is considered; not lending.

But Lilico says "the entire argument is utterly confused" from the start. From the off, he says.

Lilico says that private-sector borrowers essentially borrow only from the private sector. Not from the government, he implies out of the blue. Allowing the same simplification Lilico allowed above, let us say *all* private sector borrowing originates in the private sector. So when we speak of private sector deleveraging, he says,

The net change in the indebtedness of the private sector as a whole, relative to ... the government ... is zero.

Now he brings in lending, thoroughly confusing the issue:

Within the private sector, households could pay off all of their debts to each other, and that would (in an accounting sense) make no difference whatever to the net lending of the private sector as a whole to the government.

The goal of the sectoral balances approach is not simply to separate debt by sector. The goal is to see and understand what happens in the economy. In order to see and understand what happens, it is helpful to keep the sectors separate. Adam Smith did something similar when he identified the categories we call land, labor, and capital.

Andrew Lilico's logic is painfully out of focus. Reaching for a conclusion, he writes:

That clever-sounding national accounting identity at the start, once we ignore the external sector, says nothing more than that what the government borrows from the private sector is equal to what the private sector lends to the government

In Smith's day, Lilico would have been calling it foolish to separate labor from capital because what capitalists pay to labor is equal to what labor receives in payment.