Commentary: About That Crowd Out

A coherent view of financial inter-mediation has yet to be integrated into macroeconomic theory, a shortcoming which was laid bare in spectacular fashion when the Global Financial Crisis hit in 2008. There has been a flurry of new economic thinking outside the halls of the academy, but many of field's elite have stuck largely to traditional orthodoxy. That's a shame. As facts change, I change my mind. I think a great economist said that.

It is understandable that the top macro theorists who were tasked with explaining the financial crisis didn't measure up to the challenge. The current generation of macro-economists grew up on New Classical Macroeconomics, a movement founded in the 1970s in response to the apparent "failure" of Keynesian policies to jump start the economy during the stagflation crisis. This new school of thought sought to derive macroeconomics from classical micro economics. In terms of describing individual choice and real exchange, classical micro is a highly successful theory. However, it is wholly unequipped to address issues which affect money-production economies, and clings to the quaint notion that money is "just a veil" which facilitates barter. However, eight years after trouble in the financial sector put the world on a lower growth trajectory, the idea that in the long run money and finance don't matter for growth, and therefore can be abstracted away when formulating theory, is the intellectual equivalent of sticking one's head in the sand.

Unfortunately, the refusal of the economics profession to develop an accurate account of finance has led policy makers astray. The recently release 2016 Long Term Budget Outlook from the CBO makes the audacious claim that the US economy will be five percent smaller in 2046 than it would be if the national debt were reduced as a percentage of GDP. The mechanism by which this supposedly occurs is that households and firms would spend more on government bonds and less on productive investments. The government and private sector would also "compete" for funds, and a large public debt load would thus crowd out the funds available for investment.

Prima Facie, it's a convincing story which quickly unravels upon further examination. First, the so-called "loanable funds" approach where a fixed quantity of savings is allocated by the banking sector to productive investments has been all but abandoned by monetary theorists. Rather, than "lending out" reserves, banks simply credit the deposit accounts of borrowers. The bank thus takes on a deposit liability and acquires a loan as an asset. When the bank deposits are used by the borrower to make a purchase, settlement occurs between banks with reserves at the central bank. If the lending bank falls below its regulatory reserve requirement, it borrows reserves, which the central bank supplies on demand and at a specified price. (The Fed Funds rate in the US, the cash rate in Australia, ect.) No banker checks to see if he has money in the vault before he lends. Rather, as put by the great Canadian economist Basil Moore, "Banks make loans, and then look for reserves elsewhere."

If the Fed or other central banks did not fix interest rates but rather the quantity of liquidity in the banking system, the Neo-classical account of bank lending would hold. Banks would be reserve constrained in their lending activities, and borrowers would ultimately be competing for the funds of savers. In order for the market to reach equilibrium, the rate of interest would rise and fall with the demand for funds. However, modern banks operate in exactly the opposite manner. The rate of interest is fixed, and quantity of reserves is allowed to vary.

The significance of this arcane feature of money market structure has been over-looked precisely because it occurs in no other domain of the economy. In the short run at least, the quantity of any specific commodity is fixed and its price varies so that markets clear. However, suppose that the government had the magical power to create and destroy cell phones at will, and used this ability fix the price of a cell phone at one hundred dollars, regardless of market demand. If demand for phones rose, the government would create new phones to stop the price from rising, and if demand fell, it would destroy phones to stop the price from falling. In this fantastical world, the use or consumption of phones by one agent, even the government itself, could never crowd out the use of cell phones by another agent. But on the money market, this dynamic is not fantasy but reality. The naked truth, well understood by virtually all central bankers, is that nobody has to "save" so that others can borrow. An army of savers did not abstain from eating 500,000 ice cream cones off the dollar menu at McDonald's so that Mr. X could by his house.

But the US government does not borrow from banks, rather it sells securities. So could it somehow "use up" money held outside the traditional banking system? No. Traditional commercial banks acquire short term Treasury securities precisely in the same way they make loans. But most long term bonds are held outside the banking sector. However, a vast network of 'shadow banks' exists which ultimately borrows from the same elastic supply of reserves (again, ultimately supplied b the Fed) that banks do to finance security positions.

So the dreaded crowd out can never occur on the liabilities side of the private sector's balance sheet. What about on the asset side? Specifically, if the US Treasury issued much more long term debt than the private sector wished hold, longer term yields would have to rise so that the bond market reached equilibrium. If investors could earn high returns in risk-free government bonds, they would probably rebalance their portfolios towards government debt and away from productivity enhancing investments in R&D, plant, and equipment. In the long run, this would slow growth. However, here it is crucial to remember that the US government, as the sole issuer of the US dollar, has a monopoly on risk-free dollar denominated financial assets and thus controls their quantity and availability in the market place.

This was demonstrated during the various rounds of quantitative easing implemented by the Federal Reserve. By replacing interest bearing long term bonds with the ultimate short term asset, reserves at the Fed which at the time yielded a paltry 25 basis points, the Fed created a scarcity of long term risk-free investment outlets. This caused bond yields to fall, which in turn made holding bonds less attractive to investors, who then rebalanced their portfolios towards riskier assets.

If juicy yields on government debt ever threatened to cause investors to allocate too much of their portfolios towards Treasurys and not enough towards investments in the real economy, nothing prevents the government, either through QE or debt buybacks, from bringing down longer term yields by issuing more short term assets. Conversely, in an over-heating economy "reverse-QE", where the government retires short term debt by issuing long term debt, could be used when increases in the Federal Funds rate fail to result in higher long term rates. (As was the case from 2004-2006).

In sum, the size and composition of the Federal Government's liabilities is a policy tool, not a policy objective. Larry Summers recently criticized the CBO for its report on infrastructure, but prefaced his remarks by calling the institution a national treasury. I think the CBO does an adequate job of scoring specific bills, but a rather poor job on macro-oriented analysis. But who can blame them? When you are navigating with an outdated map, its easy to end up in the ditch.