Phantoms in the Dark Part 3

 

One of the assumptions that is sometimes adopted in discussing the impact of current government borrowing is that the economy is at full employment, including full employment of it’s financial capital (total accumulated savings), so that when the government consumes more of the country’s product then someone else must consume less.  If the government taxes the population to pay for its increased consumption then after-tax incomes decline and the population consumes less, which frees up product for the government to consume.  If the government doesn’t tax the people, and borrows from the stock of existing savings, that leaves less available savings for businesses to borrow to invest in productive capacity, or for households to borrow to invest in durable goods and housing.

A lot about that paragraph makes me squirm, and not just because half the words it contains are boring.  The assumption of full employment of labor is absolutely at odds with what we see around us right now.  The idea of full employment of the existing stock of savings also fails as a description of the current situation in the United States and much of the rest of the world too.  For example, the U.S. federal budget deficit for 2012 will be about $1.1 trillion, but the excess reserves (the extra money that banks have sitting idle at the Federal Reserve) is nearly $1.5 trillion.   In other words, the federal government of the United States could finance its entire deficit for 2012 from money the banks have just sitting idle in their accounts at the Fed.  We could just substitute Treasury bonds for their account balances; I’m not sure why they would object, either, since it would raise the interest rate they get from their deposits, and they’re not doing anything else with that money right now.

I want to be clear.  Crowding out does happen.  It happens all the time, and in a lot of different forms, not just because government consumes more but because anyone consumes more when markets are tight.  When you buy the last can of tomatoes from your local supermarket, the customers who come to the store after you are crowded out of the canned tomato market, at least at that store.  When markets become too tight, and crowding out happens too much for too long, prices rise for the things that are experiencing high demand.  In the case of accumulated savings available for borrowing, that would mean that when the market is tight and demand is high the interest rate on loans would rise.  

But I promise yesterday to discuss one specific variation of the “crowding out” assertion, the one that claims that crowding out must occur because Savings=Investment is an accounting identity. 

So here goes. 

When I hear that phrase my first reaction is to ask which of those two words, accounting and identity, the speaker does not understand.

Accounting, in the first place, is just a way of writing down what has happened in the world in some fixed period of time.  It’s not easy, because the world is immensely complex, and because people’s wealth depends on the accountants getting this right so people are fussy about details.  But it’s just an image of one slice of time, and it has no magical powers to change what it records.  To say that the accounting equations change what happens in the world is like saying that the weather map creates the thunderstorm, or the seismograph creates the earthquake.  That doesn’t mean that people can’t change their behavior after they see what the accounts record, but just because a picture of a sunset in Nepal may create a desire to visit Nepal doesn’t mean that the photograph caused the sunset. 

And in the second place, when the accounts are written down about a particular time period, that period is past.  Whatever happened is done, and can’t be changed by writing it down.

People who use this argument, that crowding out is a result of an accounting identity, might say that they don’t mean that it’s the act of accounting itself that has force, but that the accounting equation is a recognition of a force that’s already there in the economy.  So we have to move to the second word: the equation they have in mind is an identity!  That means it’s true by the construction of the equation, not because something in the world changed to make it true.  It’s always true, no matter what happens in the real world, and no matter how small or large the accounting period is.  If the government deficit doubles, it’s still true.  If investment expenditures double, it’s still true.  And if both of those things happen at once, it’s still true.

It’s true because in the accounting equation, savings is whatever is required to make it true. 

Part of the confusion may arise because the savings in the accounting equation is the savings that took place in a given time, and it has to be distinguished from all the savings from all of time that are available for borrowing.  It’s not the total of all money in bank accounts everywhere---the accounts only record the savings that have occurred in a slice of time; within the equation it’s a measure of the change in the amount of savings available within that accounting period, not the total accumulation across time.  Here’s how the Concepts and Methods of the U.S. National Income and Product Accounts describe how they derive savings from personal income:

“This account shows the sources and uses of income received by persons…The right side of the account shows the sources of personal income—such as employee compensation and interest and dividend income. The left side shows personal taxes and outlays and personal saving, which is derived as personal income minus personal taxes and outlays.” (emphasis mine)

Business savings and government savings are derived in a similar way: it’s whatever is left over from income when all direct uses of income have been subtracted.

Let’s go back to basics.  The first equation I saw when I first walked in to my first macroeconomics class was this: C+I+G+(X-M)=C+S+T.   The X-M is exports minus imports; let’s get rid of that right up front and pretend that the rest of the world doesn’t exist---not because it’s unimportant, but because this point is easier to see without it. So we’re left with C+I+G=C+S+T: Consumption+Investment+Government expenditures equals Consumption+Savings+Taxes paid.

C is in this equation twice, and holds the same value in each place, but conceptually it is not the same C.  On the left side, it represents the consumption goods sold: it’s a source of income to the businesses or people who are selling these things.  The revenue from that sale goes into someone’s bank account---in most cases, bits of it go into a lot of people’s bank accounts, since a lot of people worked to make, transport and sell the products.  On the right side, it’s consumption goods purchased: it’s a use of income.  Revenues from that C come out of someone’s bank account.  But since purchase and sale are two sides of the same transaction, these have to be the same number.   Transactions like this have no impact on savings, actually: the funds come out of one bank account and go into another---usually out of a household bank account, where it had been personal savings, and into a business bank account where until it is spent on something or paid out in dividends it is current retained earnings.  In either case, it’s part of the whole nation’s savings.

The same is true of all the variables on the left and right sides: everything shown on the left is something that goes into someone’s bank account, and everything on the right is something that comes out.  This is important to understand: because when investment goods sold in this equation increases, that new revenue all goes into someone’s bank account.  As with consumption goods, the money spent on investment goods comes out of one account and into another. 

But when goods and services sold to Government increases, that sale also goes into the seller’s bank account---but it doesn’t come out of any other private account; it comes out of the Treasury’s account at the Federal Reserve.  Contrary to the usual view, government deficits don’t crowd out private investment within this accounting framework: they are income to the people who have sold goods or services to the government, so they increase savings by the amount of the deficit.  This new savings is then available for the government to borrow, if it wants to do that to cover its deficit.  Government deficits create the savings that government borrowing might absorb: nothing is crowded out, and private investment in the accounting equation is not affected in any way.  This somewhat surprising result is a preview of the even more surprising and contrary functional finance vision, but in this case it’s also just a description of the way the National Income and Product Accounts really work.

The fact that we can’t use the accounting identity to “prove” crowding out doesn’t mean that it never happens.  It means we have to work a bit harder, and put some conditions on the circumstances, to show how it happens. For example, crowding out may occur because there actually is a shortage of some resource the government is trying to buy.

In sum, it’s not impossible, or even very hard, to construct an economic model where crowding out occurs.  But within the National Income and Product Accounts, government deficits simply increase the recorded savings, and government borrowing absorbs them: nothing else has to change to make this work.  This is an unavoidable result of the fact that the accounting equation in question is an identity, an equation that can’t be false no matter what the real world does.