In Part 1 of this two-part series, I explained that there were (at least) two distinct frameworks in play, when discussing the concepts of Saving and Investment, and their possible (or necessary!) equality. Specifically, I contrasted the arguments that economists had over the (in)stability of the market economy based on households’ portfolio decisions, with an accounting framework where saving by definition always equals investment.
The first article primarily laid out the accounting tautology (which went hand-in-hand with InFi episode #94, where my guest Charles DuBois made the case in a few different ways. If you haven’t already read that first article, you should definitely review it before tackling the present one. Here, I am going to reconcile the two camps, which at first seem to contradict each other.
It is common to see those laying out the accounting tautologies go on to criticize the economists with their own, allegedly confused statements on the topic of Saving and Investment. For those who listed to the InFi episode, you heard that DuBois himself got into that a bit, for example when he said that saving had nothing to do with “living below your means” but rather was due to investment spending.
To give two more specifics, I can quote from the two articles I linked in the prior post—the first being an intuitive introduction to the S=I identity, while the second was a very rigorous exposition of the accounting. From the first, intuitive article, after the author has explained the accounting approach, we see the following swipe at the typical economics treatment of the subject:
Probably the most egregious example of an economist who gets the S=I identity wrong is Gregory Mankiw. Mankiw is commonly regarded as one of the most important macroeconomists, and his macro and micro textbooks are among the most commonly used. In the third edition of his microeconomics textbook he explains the S=I entirely wrong.
Mankiw writes
Consider an example. Suppose that Larry earns more than he spends and deposits his unspent income in a bank or uses it to buy a bond or some stock from a corporation. Because Larry’s income exceeds his consumption, he adds to the nation’s saving. Larry might think of himself as “investing” his money, but a macroeconomist would call Larry’s act saving rather than investment.
This is entirely wrong. In order for Larry to save a portion of his income in the form of a financial asset the rest of the economy must be running a deficit. Larry’s saving does not add to the national wealth.
Mankiw then embarks on an elaborate rationalization where he discusses how the additional saving in Larry’s bank account causes additional investment. This entire discussion is wrong. An identity must be true in all possible economies, so you cannot appeal to some particular fact of the banking system when justifying it. In particular, the identity must hold in a world where, for example, people’s decisions about saving and investment have nothing to do with the interest rate. It is beyond embarrassing for the economics profession that someone who is so prominent within the field can make such a basic error, and it can remain in macroeconomics textbooks unchallenged for three editions.
(To avoid confusion, in the block quotation above, the plain italicized words are those of the article author, while the bold are from Mankiw’s textbook discussion.)
And to show that I’m not overreacting to one source, here is an excerpt from the exhaustive, scholarly article. Again, after meticulously laying out the accounting approach to these matters, the author then proceeds to (in his mind) dismantle the confusion reigning among the economists. Here is just a taste:
In the following, we present a prominent example of a problem entanglement: the loanable funds model (see Robertson (1934) for an early systematic exposition). As will be explained in more detail, the model confuses production, financial saving, non-financial saving and credit. This has disastrous consequences for policy because it concludes that an increase of non-firm financial saving through a reduction in consumption would necessarily lead to higher investment. However, in a monetary economy, this is not likely to be the case, as will be argued in the critique that follows the model’s presentation.
I trust that these two excerpts (as well as my reference to the remarks from Charles DuBois in our InFi discuss) show that I am not inventing strawmen. It is common in this literature to (a) establish the accounting approach in which S=I by definition and then (b) use this framework as a springboard from which to criticize the typical economist who praises saving as the source from which investment spending flows.
In this blog post, I can’t hope to definitively settle the dispute. Rather, I want to sketch out for you, the reader, what the typical economist has in mind with such statements, and why the accounting tautologies do not destroy the “loanable funds model” and related concepts.
Being familiar with both frameworks, I’ve been pondering the essence of the divide. I believe it fundamentally derives from thinking of the economy either in terms of “real,” physical resources and finished goods, versus thinking in terms of financial magnitudes measured in monetary terms.
In the previous post, I focused on the accounting approach, which uses financial magnitudes denominated in the currency unit. As we saw, here S=I by definition, for any specified time period.
But the typical economist is concerned with more than mere accounting. He wants to explain the social benefits of saving and investment. To put it bluntly: Why do we care about such things? Why are we devoting so much of our time to analyzing these terms?
In the typical economist’s framework, society has a scarce endowment of real resources, including natural resources (what used to be called simply “land”) such as coal deposits and forests, but also human labor of various skill types and also a collection of various types of tools and equipment (“capital goods”).
At any given time, the people in society can channel their scarce resources into the production of consumption goods (giant TVs, sushi dinners, diapers, etc.) and investment/capital goods (hammers, MRI scanners, communication satellites, etc.). There is a tradeoff between these two broad categories: At least if we assume the economy is operating at capacity, then increasing the production of (say) giant TVs necessarily requires a reduction in the flow of newly-produced (say) communication satellites in that period.
In this approach, it is obviously the case that saving precedes investment in a logical sense: We can’t physically produce more hammers unless we first have the necessary resources, and we choose to not channel those resources into making more coffee tables. There is no sense in which the physical production of hammers could somehow call into existence the real resources needed to produce those very same hammers, in the way that some proponents of the accounting approach argue that investment spending makes possible the aggregate (net) saving that finances it.
In closing, let me showcase a start illustration of the different ways economists and accountants think about such matters. From the intuitive article laying out the S=I identity, we can quote the author’s initial example:
In this simple example, the author drives home the point that without spending on investment goods, there’s no way for both individuals to “get ahead” in terms of financial assets. One person’s saving can only come at the expense of the other person’s dissaving, of the same dollar amount.
The author then shows what would happen if both individuals tried to “live below their means” by spending less than their income:
Now before I proceed to the fundamental issue, let me first deal with a quibble: Strictly speaking, I think it is ambiguous whether to classify money as a financial asset, rather than as a commodity. This is obvious if we are dealing with a metallic money, such as literal gold coins. Clearly, if more gold is mined, stamped into coins, and added to various households’ portfolios, that increase in “cash balances” doesn’t correspond to someone else’s higher liabilities. This is an important difference between money in the form of coins versus claims on money (such as bonds).
I would argue that even a non-tangible item such as a bitcoin—if it ever became used as money—would likewise be a commodity (and this indeed is how certain regulations treat cryptocurrencies). When someone “possesses” 300 satoshis, his asset doesn’t correspond to someone else’s liability, in the same way that someone’s $300 in a checking account corresponds to the bank’s $300 liability.
Once we consider the possibility of commodity money, we see that it is possible for everybody in society to accumulate higher cash balances. However, I understand that the accounting proponents of the S=I identity might object that these owners would be invested in the acquisition of more physical goods, namely yellow discs that we call “gold coins.”
So, to be as fair as possible to their perspective, let’s work with the exact example of the author quoted above. Suppose there are only two people in the economy, and each starts out with $100 in currency. Each person desires to add 50% to his cash balances, i.e. each person wants to spend only $50 of his usual $100 income (and thus add $50 to his original $100 cash balance).
As the author points out, this is impossible (at least if we assume the monetary authorities refrain from printing more money). When one party restricts his consumption spending by $50, if he doesn’t in the same period replace it with $50 in investment spending, then the other party’s income must fall by $50 for that period. Thus, each person’s attempt to “live below his means” thwarts the other guy’s attempt, leaving them with the same cash balance but a lower income and consumption spending for that period.
Yet although the attempt at universal saving seems elusive from the accounting perspective, from the economicperspective we can observe that by refraining from spending $50 on pizza (or going to the movies, etc.), each household “frees up” the real resources that otherwise would have gone into the production of those consumer goods, at least in principle. When prices fall for these inputs, it makes it possible for businesses to eventually hire them to divert into goods intended for sale in the future.
And so we can see what Mankiw was getting at in the quotation shown above. Admittedly, economists may sometimes be sloppy in flipping back and forth between the “real” and the nominal framework, but that’s because in our actual world, financial decisions govern the flow of real, physical resources. This is a very complex relationship, and economists of different schools of thought obviously have different theories about it. But it’s not that they are wrong to wrestle with such ideas; they occur on a different plane from the mere accounting. Or rather, the economists are trying to connect the accounting tautologies with physical goods.
Before closing, let me make one final observation showing how price adjustments help to reconcile the two approaches: If everybody in society wants to increase his cash balances by 50%, this actually is possible, even with a fixed stock of money. The answer is that economists think people are ultimately motivated by real cash balances. In other words, when someone thinks, “I want to increase my cash balance from $100 to $150,” it’s not because there is something special about those numbers per se. Rather, the person surveys the prevailing prices in the economy, and wants the ability to buy “$150 worth of stuff” at a moment’s notice, rather than a mere “$100 worth of stuff.”
Consequently, if everyone attempts to “live below his means” by restricting spending, then this fall in demand will presumably lead to a fall in prices. The process continues until everyone’s $100 in nominal cash balances can now buy 50% more “stuff.” Then the households resume spending all of their income, because they have achieved their desired level of “real” cash balances.
I have admittedly dealt very cursorily with a complicated subject. But I hope I have demonstrated in this two-part series that there is something deep and important at play, when economists discuss S=I as an empirical tendency, while others think of S=I as an accounting identity. It’s not that one camp is right and the other is wrong. Rather, they are both correct in their respective frameworks, and the more sophisticated members of each camp will try to understand how the two frameworks interact with each other. After all, the reason we care about financial assets, household wealth, etc., is that this corresponds to physical realities such as the amount of food eaten each month or the number of new sports cars produced.
Twitter: @infineogroup, @BobMurphyEcon
Linkedin: infineo group, Robert Murphy
Youtube: infineo group
To learn more about infineo, please visit the infineo website