Saturday, February 18, 2012

Savings and Debt

Much like spending and income, savings and debt are opposites from an individual perspective, but are the same thing when considering the economy as a whole.  The only way to earn interest on savings is to loan it to someone else and charge interest.  So there is no trope about our economy that bugs me more than the idea that Americans have too much debt and not enough savings.  This is a problem we can't have -- savings and debt go hand in hand, we can't have one without the other.  Perhaps we have too much debt and too much savings.  Or you could say that the middle class has too much debt and not enough savings, but the corollary of this is that someone else -- perhaps the government or the wealthy -- will need to have less savings or more debt to solve the problem.  "Too much debt" is not a problem that can be solved by encouraging more saving -- more saving is an anti-solution that puts pressure on Wall Street to search the mountains and valleys for anyone who could be convinced to go into more debt.

If for any reason the phrase "fractional reserve banking" ran through your head as you were reading the last paragraph, you're off base.  Sure, the statement applies to fractional reserve banking, but the point is it applies to any sort of banking system.  In fact, the debt that fueled the housing bubble was not created primarily through fractional reserve banking but rather through financial securities such as CDOs.  The only important difference between fractional reserve banking and other forms of finance for the purpose of this discussion is that, in fractional reserve banking, up to 90% of savings become someone else's debt, whereas with other forms of finance 100% of savings are someone else's debt.  As a result of this, the most debt the fractional reserve banking system can possibly create is nine times the money supply, whereas financial instruments like bonds and CDOs allow money to be loaned, spent, loaned, spent, over and over again until one of three things happens: 1) no one is interested in loaning money, 2) no one is interested in borrowing, or 3) the financial system becomes unstable.

Let's continue the thought experiment from the last post and examine another way that money can constrain an economy. 

In our last post we kept the concentration of wealth fairly even to keep the problem simple.  Now let's consider what happens when the economy becomes less equal over time.  In the real world, a handful will become far richer than average, and the richest will be much further above average than the poorest are below average.  In the last post there were about as many above average as below average and those with extra money did about as much extra spending as those short of money were cutting back.  Will someone with a lot of extra money spend even more?  Generally, yes, but not in linear proportion to how much money they have -- a person with $100,000 in cash may only spend at 5 times the rate as someone with  $10,000.  So given what we discussed last post, we might expect spending -- and therefore incomes -- to fall over time as inequality grows.  But this changes when we consider savings and debt.

Finance offers a compelling alternative to spending for those who are not living paycheck-to-paycheck -- if you don't need to spend your money right away, why not earn a little interest on it and have more when you need it?  And when you do this, that money is loaned to someone who is planning to spend it right away -- so spending does not slow down as a result of your saving; the economy continues humming right along!  Given a functioning financial sector and a competent Fed, it would seem that we can have our unequal economy and eat it too -- the gains can go to the top, and living standards for the middle class can improve too thanks to cheap credit!  This is the perverse cycle of trickle-down economics -- a dollar is spent by the middle class, it goes to Wall Street as profit, is put into the financial system where it is loaned back to the middle class, where it can be spent again.

Of course, this can't go on forever, but what exactly brings it to a stop?  Several things gradually change as the distribution of wealth changes, but it's helpful to start by dividing money into two categories based on how holders intend to use it:  1) Money available for capital -- this is money available for investment, money that owners hope to turn into more money, perhaps by buying stock, starting a business, or loan it for interest.  The money for these activities disproportionately comes from the very wealthy.  2)  Money available to spend on the product of capital -- spending money, to meet basic needs or wants.  This money comes from everyone, and is most of the money of the middle class.  At first glance the first might seem more important for the creation of business, but only the second can make it financially worthwhile to create businesses in the first place.  Without the second, investment can only be a ponzi scheme.

For investors looking to invest in business, the average return on investment is determined by the ratio between the two.  So if gains from economic growth go disproportionately to the top (to capital), then this ratio changes over time to reduce the return on investing in business.  This has two consequences of importance:  interest rates go down, as finance becomes relatively more attractive as a use for investment dollars, and the fraction of capital allocated to business will decrease while more capital is used to extend credit to the middle class in the form of mortgages, better credit card deals, college loans, and so on.  As the middle class gets easier credit, the money becomes available to spend on the output of capital again, thus solving the short term economic problem and masking the long term economic problem slowly developing.

Lower interest rates encourage more people to take on debt.  You might think it also discourages people from saving, but this isn't necessarily true.  I've certainly never changed my 401k contribution based on interest rates and don't know anyone who has.  And it's not clear that I rationally should -- lower interest rates mean I get less out of saving, but they also mean I need to save more to have the retirement I want.  As for the very wealthy who hold most of the savings, they don't necessarily view their savings as savings in the sense that they plan to spend them in the future; they are their primary source of income and they will prefer to keep that income around by only spending out of interest.  So a drop in interest rates is like a pay cut, and withdrawing from savings means cutting income further.

Role of the Fed

A common misperception is that the Fed has the ability to set interest rates, and also the ability to control inflation.  What they can actually do is trade one for the other.  So the Fed made a choice to trade high inflation for high interest rates in the early 80s, and they have acted to bring about lower interest rates in recessions, but the fact that both interest rates and inflation have come down since the 80s and are now near zero is not the Fed's doing and it is not in their power to undo it.  The normal Fed policy for a recession with inflation below target would be to print money and use it to buy bonds, thus increasing inflation and savings levels and causing former bond holders to look elsewhere to invest their money.  But this strategy can only work if there is still debt left to be created that financiers expect they can make interest on -- otherwise the money just ends up sitting in an account somewhere doing as much good as if it were never printed.

Where are we now

Some say the subprime mortgage bubble was a misallocation of capital, but I don't think this gets it quite right.  If capital was misallocated, where are the people of sound finances who were unable to get mortgages because banks were lending their money to people who wouldn't be able to pay them back?  Where are the nations with sound long-term finances which were not able to take on debt because everyone was buying Greece's bonds instead?  The problem was and is not a misallocation of capital; it is a misallocation of money to capital.  For the foreseeable future, we need to find ways to make the economy more labor friendly and less capital friendly.  A return to stability requires a return to the rising middle class wages of the post-war era and away from the rising credit lines of the last three decades.


  1. How do you reconcile these two seemingly contradictory statements?

    "And when you do this, that money is loaned to someone who is planning to spend it right away -- so spending does not slow down as a result of your saving; the economy continues humming right along!"

    "But this strategy can only work if there is still debt left to be created that financiers expect they can make interest on -- otherwise the money just ends up sitting in an account somewhere doing as much good as if it were never printed."

    BTW, liked your comment over at Noah's site about prices being emergent phenomena.

  2. The second statement is the exception to the first. The first statement basically applied from 1980-2007, and the second is the wall we've run up against since. Rising inequality, increasing debt (and savings), these things hardly affect the economy at all, until they affect it quite severely. The Fed's strategy of pushing down interest rates during recessions works just fine for dealing with them, until it doesn't work anymore. Our institutions are going to be resistant to the changes necessary to get us out of this mess, because they have learned certain things work and other things don't matter, and for a while they were right.

    Thanks! It's been interesting watching the discussion on microfoundations, and seeing how ingrained this idea is among economists on both sides of the debate that you need to come up with a model that can be solved and reduced to an equation explaining the behavior of the economy throughout all of time. This assumes away so much of what can go on, what almost certainly does go on in any real economy. Maybe economics just hasn't been taken over by people who can program like all the real sciences, but I tend to think that the most you should ask of a theoretical model is that it can be played and its behavior can be observed; insisting they also can be solved is to throw out most possibilities without any evaluation.