When the Federal Reserve was created in 1913, the United States had been through a series of banking panics that had started soon after the Civil War and continued with a startling frequency. The Fed's mission was to act as a lender of last resort to troubled banks; by a process of regress, the theory went, banking panics would be less frequent as the public was aware that the Federal Reserve was there to provide liquidity.
Many of us, viewing things from our modern perspective, will be surprised to learn that banks at one time issued their own paper currencies, with the currencies being convertible into gold and/or silver via precious metal reserves held by the banks. It seems like such a system would allow for wild printing press activities, but it contained its own very sharp internal policing mechanism: notes printed by an issuing bank would often end up deposited at a second bank, and the second bank---vigilant and suspicious---could go to the issuer and demand convertibility. A bank that was seen as being guilty of mischief could be exposed quickly.
The abilities of private banks to print money against their reserves was gradually eroded by the government, first by predatory taxation and then by a requirement that US charter banks hold government debt as a reserve requirement (this was enacted during the Civil War and was a method used to finance the war effort rather than to make the banking system more effective). After the Civil War, restrictions on currency issuance, combined with the single-unit banking rules already discussed, let to a number of banking panics. It was after the 1907 troubles that the Federal Reserve was created, with twelve strategic member banks coordinating efforts and able to print a national currency. There was nothing magical about this---another way to solve the liquidity problems that the banking system faced after the Civil War would have been to just relax the rules on private money printing in the existing system. Congress solved a problem that it had been at least partially responsible for creating in the first place.
The Fed's initial mission was primarily to act as the "lender of last resort" to troubled banks. Over the years, this mission has changed as Keynesian notions of activist monetary policy have taken hold. The modern incarnation of the Federal Reserve is bound by legal mandate (the Humphrey-Hawkins Act) to closely monitor two macroeconomic metrics---inflation and unemployment---and to attempt to steer between them, and to "connect" monetary policy to the President's fiscal policy objectives (dangerous). The Federal Reserve Open Market Committee meets every six weeks and makes decisions regarding monetary policy based on a variety of aggregated macroeconomic data points. Decisions are notionally meant to increase or decrease the money supply, but accurate measurement of the money supply has become increasingly difficult as the range of highly liquid instruments and credit facilities in the economy has become vast.
So instead of explicitly targeting the money supply, the Fed targets bank reserves. If the Fed decides that the economy is slowing and requires more liquidity, it will "buy" Treasury securities held by the banks ("buy" is in quotes because it implies that the banks have a choice---they don't, the accounting entry happens automatically in the reserve account deposits that the banks must hold with the Fed) and give the banks cash. The cash reserves would expand credit through the economy vis-a-vis the bank multiplier effect of a fractional-reserve system (as discussed earlier). If, on the other hand, the Fed decides that inflation is becoming a risk, the decision can be made to "sell" Treasury securities to the banks, which gives them Treasury securities on reserve at the Fed and takes away cash, thus contracting the money supply in at least this sense.
The primary monitoring device used by the Fed to execute all of this is a single, very important interest rate---the Fed Funds Rate. The FFR is the rate at which banks lend to each other overnight. The Fed sets a target rate for the FFR and then tries to achieve that target through open market operations---the buying and selling of Treasury securities. After the Fed makes a rate decision, open market operations are conducted on a near-daily basis to try to hit and maintain the desired level.
Raising and lowering the FFR has several important effects on the economy, but for our purposes the one to keep in mind is that it changes the time-value of money. Higher cash reserves in the banks encourage them to lend more, as if they had more customer deposits. More customer deposits would indicate greater savings, which would mean that more people are delaying consumption for the future. When rates are cut by the Fed, the pricing signal, at least temporarily, indicates that the time-value of money has decreased. If, on the other hand, everyone wanted to consume today, you would have to give them a higher interest rate to get them to delay consumption, and thus the time-value of money would increase.
In a free market, the FFR would be the result of an unplanned, emergent discovery process between supply and demand forces. A low rate would mean that banks were flush with cash because a lot of people were saving, so there was also a lot available for credit expansion. Projects could be begun today---investments made, loans taken out, people hired---with the knowledge that depositors were delaying consumption for the future (thereby making said projects less risky). When the Fed lowers the FFR by increasing the cash reserves in the banking system, it is in a sense "tricking" the economy into believing that the savings rate is higher than it really is.
The central pillar of modern finance is the concept of net present value, or the way that future cash flows from an investment or project are discounted to take into account the time value of money and risk. All other things being equal, a project that starts generating cash flows tomorrow is better than one that starts generating them next year, since you could take the cash flows that started tomorrow and invest them, earning a return while the cash flows from the other project did not even start rolling in for many months. However, if the discount rate being applied to the cash flows of both projects was zero percent, they would look equally attractive.
When the time-value of money is manipulated downwards by means of central planning apparatus, projects that normally might not look economically attractive can look a lot more appealing. This can help to reflate asset values in an economic downturn. As we will soon see, the projects that are most sensitive to interest rate moves tend to be the ones that are expected to generate cash flows out into the longer-term future, like you find in capital-intensive industries such as housing, infrastructure construction, mining, and commercial real-estate development. This has had serious ramifications for us recently.
One may ask why flooding the market with new money by increasing bank cash reserves does not simply lead to everyone just raising prices to acknowledge that there is new money in the system. Imagine that two people were on a desert island and had two machetes and two dollar bills as their total economy. Each machete is worth a dollar. Now two more dollar bills wash up. The economy is not really richer; each machete is now worth two dollars instead of one. Creating money does not mean that the level of real goods and services has increased.
Well, the theoretical reasoning behind the role of monetary stimulus (and perhaps the keystone of the whole Keynesian argument) is that wages and prices are "sticky." Rather than responding immediately to money printing, wages and prices will remain the same for awhile and thus the new spending money will be able to push up aggregate demand. Imagine a drag race between a car marked "full employment" and a car marked "inflation". The starter gun is called "Keynesian monetary and fiscal stimulus packages." The stickiness of wages and prices argument states that the full employment car will respond more nimbly to the stimulus signal and will race ahead faster than the inflation car will.
There are reasons to believe that wages and prices are becoming less sticky as businesses become more agile and union labor takes up a smaller percentage of the economy. If this proposition is true, we would expect Keynesian monetary stimulus to be less and less effective, and possibly for the economy, like a junkie, to require greater and greater stimulus effects to get the same results . We'll get into the core Keynesian models, such as the famous "IS-LM" general equilibrium model, in a future discussion.
All competent central banks, including the Fed, keep a close eye on inflation. If too much money is printed, inflation may break free, the friction of sticky prices and wages may be overcome in a violent way, and the Fed may lose control as other, nominal interest rates rise because people understand that the printing presses are at work and thus they demand greater compensation for lending it now and getting less valuable money back in return in the future.
Alas, there are several complicating factors involved in inflation-targeting that may result in monetary policy mistakes occurring even in the most competent hands. I will choose just three of them and touch on them very briefly.
The first issue is that the measurement of inflation may be flawed. The Consumer Price Index is an aggregate metric that is meant to show changes in the prices of a basket of goods and services. The CPI has historically been modified in ways that may give it an underestimation bias, perhaps because government expenditures pegged to inflation-based cost-of-living adjustments and heavy borrowing from foreign financing sources have created a desire to lower the official inflation numbers. Three fairly recent modifications have included the use of geometric instead of arithmetic calculations for period-to-period CPI change; a modification for "substitution effects" (if the price of beef rises against the price of chicken, to use one example, the new CPI calculation assumes that consumers will substitute more chicken for beef and thus the chicken numbers will be used for price change purposes); and, most controversially, a "hedonic adjustment" that makes a downwards adjustment to an increase in price if it is judged that a corresponding increase in product quality has also taken place.
The second issue is that there *may* be a natural deflationary drift in a successful capitalist system, as competition eventually causes goods and services to be "commoditized". Commodities reward the lowest-cost producer because that firm will be able to win the inevitable price wars that erupt between producers of identical or nearly-identical products, and other firms will have to adopt smaller-volume, differentiated niche strategies in order to avoid having to compete on price.
The globalization of manufacturing and some services to lower-cost/comparative advantage production in China and India may have a deflationary economic impact, as would technological improvements that lower the costs of, say, VCRs in the early 1980s or computing power today. This could cause problems because it would mean that, in order to target a rate of inflation of, say, 3%, the Fed might not be starting with a default rate of 0%---it might be facing a continuous 1-2% deflationary headwind ( a slow, benevolent "supply side" deflation that is not the result of crisis "demand shock" deflation, which may run at 5-10% and be extremely destructive, but rather a positive result of capitalism, access to more production capacity as markets globalize, exploitation of technological means to search for best values, etc.). Thus, the Fed might systematically run a dangerously loose monetary policy because the focus on price stability (i.e., mild inflation) may itself be counterproductive.
The last issue that I'll cover that may confound inflation-targeting efforts is political in nature, and is based on the perceived need for "coordinated" fiscal and monetary policies. In the coordination dilemma, the Fed is not able to truly act as an independent inflation watchdog, but instead is leaned on to keep rates low in order to help finance heavy borrowing on the fiscal side.
A historical footnote providing us with a clear example of this political pressure in action took place when William Martin, a salty-dog veteran of the Federal Reserve chairmanship who was usually known for his insistence on insulating the Fed from political populism, faced Lyndon Johnson's need to fund both escalating US involvement in Vietnam and his Great Society programs. Johnson had to run fiscal deficits that, at the time, were considered very large. Large government borrowing efforts can create a "crowding-out" effect in the bond markets, as competition for funding causes interest rates to rise. Martin ended up, probably against his better judgment, financing Johnson's debt-fueled fiscal program efforts by keeping interest rates low (and changes in M1---a measure of the money supply that was useful at that time---more than doubled), although he did raise rates in 1965 despite Johnson's sharp protestations.
Richard Nixon, for his part, made his low-rate interests very clear to his Fed chairman selection, Arthur Burns, and Burns did as he was asked/told. The result of this "stimulus-accomodative" policy is now well known: runaway inflation during the 1970s. The 70s also saw the arrival of a macroeconomic-monster that the Keynesian conceptual framework had never described: stagflation. What made stagflation so difficult for policymakers is that it revealed a problem with the Phillips Curve, the analytical device that depicted a strong inverse relationship between unemployment and inflation. An increase in inflation would correspond with a decrease in unemployment, and vice versa: the task for policymakers was to manage the "inflation-unemployment" tradeoff, which was possible using the standard Keynesian toolkit of monetary and fiscal stimulus efforts when unemployment was unacceptably high. During the 1970s, both inflation and unemployment rose at the same time---this was an inconceivable result for the Phillips Curve. A modified version has since been developed as part of the New Keynesian Synthesis school of macroeconomics, and it suggests that there is a short-term inflation/unemployment tradeoff (due to the "sticky" wages and prices effect previously described), but in the long-term the relationship breaks down.
In the next installment, we will look at how the Federal Reserve lowered interest rates to very low levels in the wake of the dot.com crash and 9/11, and how doing this sent a signal to the market that encouraged borrowing and fueled the housing bubble.
FURTHER READING: Allan Metzler's "A History of The Federal Reserve" (volume 1 available now; volume 2 is meant to be out in Feb 2010) is a painstakingly researched chronicle of Fed policy since its inception in 1913. Of particular interest in the political pressure placed on the Fed to pursue expansionary monetary policies. Chris Farrell's "Deflation" contains a nice argument for the positive aspects of "supply side" deflation, as does George Selgin's concise, fascinating "Less Than Zero". My heroes Hayek, Rothbard, Hazlett, and von Mises of the Austrian school of economics provide some penetrating discussions of distortions and resource misallocations resulting from centrally planned interest rates. Selgin, an authority on "Free Banking" (banking treated as any other industry), also has a number of works available that discuss issues with central banks and unlimited reserve generation capacity when external monetary disciplines (such as the gold standard) are absent.