Unit 7 · Money, Banking and the Federal Reserve

Imagine a town with no money. A dentist wants a haircut. The barber does not need dental work; he needs shoes. So the dentist must find a shoemaker with a toothache, trade a filling for shoes, then trade the shoes for a haircut. Economists call this problem the double coincidence of wants: in barter, each side must want exactly what the other has, at the same moment. Trade happens, but slowly and rarely. Money solves this. The dentist fills a tooth for $150, hands the barber $30, and keeps $120. Nobody has to want anything but money.

That is the first job of money: it is a medium of exchange, something everyone accepts in payment. The second job is to be a unit of account, the common measuring stick for prices. Without it, a store would need to post the price of a sandwich in eggs, in haircuts, in bus rides and in every other good. With dollars, one price tag does the work: $9. The third job is to be a store of value. If you mow a lawn today, you can hold the $40 and buy something next month. Fresh fish cannot do that. Dollars can, at least while prices are stable.

Anything that does all three jobs is money, no matter what it is made of. In prisoner-of-war camps during World War II, cigarettes became money: soldiers priced chocolate and soap in cigarettes, accepted them in payment, and saved them for later. Notice what money is not. A credit card is not money; it is a loan you must repay with money. A stock is a store of value but not a medium of exchange, since you cannot buy groceries with a share of a company. Money is the set of things that do all three jobs well.

For most of history, money was a thing with its own value. Gold and silver coins, salt in parts of Africa, tobacco in colonial Virginia and cowrie shells across the Indian Ocean all served as money. Economists call this commodity money: the money is worth something even if nobody uses it as money. A gold coin can be melted into jewelry. That is its strength and also its weakness. Commodity money is heavy to carry, easy to shave or clip, and its supply depends on luck at the mine rather than on the needs of trade.

The next step was paper that promised a commodity. A bank or government printed a note that said, in effect, hand this in and we will give you one ounce of silver. The paper was light and easy to carry, but its value came from the metal it could be traded for. This is representative money. The United States issued silver certificates for decades; a $1 silver certificate was a claim on a dollar’s worth of silver held by the Treasury. The note itself was just paper, but the promise behind it was real metal.

Today’s dollar is neither. Pull a bill from your pocket and read it: it promises you nothing but itself. It is fiat money, from a Latin word meaning let it be done. The government declares it legal tender, which means a creditor must accept it in payment of a debt, and the Federal Reserve controls how much of it exists. Its value rests on two things: people’s confidence that others will accept it tomorrow, and the central bank’s job of keeping its supply in line with the economy so that prices stay roughly stable. Most countries now use fiat money, because its supply can be managed rather than dug out of the ground.

Each form has a trade-off. Commodity money limits how much money can exist, which protects against runaway printing but can starve a growing economy of cash. Fiat money can expand with the economy, but if a government prints too much, prices rise and the money loses value. Germany in 1923 and Zimbabwe in the late 2000s both saw prices double in days as their governments printed without limit. Fiat money works only as long as the people who manage it are trusted to use restraint.

Why cookies fail as money and dollars succeed comes down to a short checklist. Good money is durable: it survives being handled thousands of times. It is portable: you can carry enough to buy a car. It is divisible: it breaks into small units so you can pay $3.47 exactly. It is uniform: one $20 bill is worth exactly the same as any other, so nobody has to inspect each one. It is scarce: the supply is limited, so it holds value. And it is accepted: people take it without argument. Cookies fail on durability, uniformity and scarcity. Gold failed on portability and divisibility. Paper dollars pass every test but one, and that one takes a central bank.

The test paper cannot pass on its own is scarcity. Paper is cheap, so someone must decide how much money to create and hold to that decision. In the United States that someone is the Federal Reserve, the country’s central bank. It does not decide by whim. Its job, set by Congress, is to keep prices roughly stable and employment high, and the amount of money in the economy is one of the levers it uses. When the Fed lets money grow much faster than the goods and services people produce, each dollar buys less. That is inflation. When money grows in step with output, prices hold steady and the dollar keeps its value.

Here is the arithmetic in a tiny economy. Suppose a village produces 1,000 loaves of bread a year and has $2,000 in circulation that changes hands once. The price of a loaf settles near $2. Now the village doubles its money to $4,000 while still baking 1,000 loaves. The same bread is chased by twice the dollars, and the price drifts toward $4. Nothing about the bread changed; only the count of dollars did. This is why the dollar’s value ultimately rests on the discipline of the people who control its supply and the confidence that they will keep it.

You can measure how much money exists. The Fed reports a narrow measure, M1, that counts cash in circulation plus checking and similar accounts you can spend right away, and a wider measure, M2, that adds savings accounts and other near-money. Notice that most money in the United States is not paper at all. It is numbers in bank accounts, moved by debit cards and transfers. The paper in your wallet is a small slice of the money supply. Most dollars have never been printed.

Picture a new bank on opening day. Its first customer deposits $10,000. The bank now owes that customer $10,000, payable on demand. But the customer will not ask for all of it tomorrow, and neither will the next hundred depositors. On any given day only a few people withdraw. So the bank keeps part of the money as reserves, cash in its vault or in its own account at the Federal Reserve, and lends the rest. If it holds 10% in reserve, it keeps $1,000 and can lend $9,000. This is called fractional reserve banking, because the bank holds only a fraction of what it owes.

The lending is how the bank earns a living. Suppose it lends the $9,000 to a landscaper buying a used truck at 8% interest for one year. At year’s end the landscaper repays $9,000 plus $720 in interest. Meanwhile the bank pays its depositor 2% on the $10,000, which is $200. The bank’s gross earnings are $720 minus $200, or $520. Out of that come the teller’s wages, the building, the software and losses on loans that are never repaid. What is left is profit. Economists call the gap between the lending rate and the deposit rate the spread, and it is the core of the banking business.

The bank is doing something valuable for both sides. The depositor gets safety, convenience and a little interest without having to find a trustworthy borrower herself. The landscaper gets a truck this year instead of saving for three years. The bank sits in the middle and specializes in the hard part: judging who will pay back a loan. Economists call banks financial intermediaries, go-betweens that move money from people who have more than they need right now to people who can use it productively.

The arrangement has one built-in danger. The bank owes $10,000 on demand but holds only $1,000 in cash. If every depositor showed up at once, it could not pay. As long as people trust the bank, they do not all show up, and the system works. If trust breaks, everyone runs for the door at the same time. That is why banks are regulated, examined, and required to hold capital of their own, and why the government insures deposits. The next lessons show how the lending multiplies money and what stops a run.

Follow a single $1,000 deposit through a banking system where every bank keeps 10% in reserve. Bank A receives the $1,000, holds $100, and lends $900 to a student for textbooks. The bookstore deposits the $900 in Bank B. Bank B holds $90 and lends $810 to a mechanic. The parts supplier deposits $810 in Bank C, which holds $81 and lends $729. Add up the deposits so far: $1,000 + $900 + $810 + $729 = $3,439. The original $1,000 has become $3,439 in checking accounts, and the chain is not finished.

Each round the new deposit is 90% of the last one, so the amounts shrink toward zero. If you carry the chain all the way, the total is $1,000 divided by 0.10, which is $10,000. The general rule is the money multiplier: the maximum increase in deposits equals the first deposit divided by the reserve ratio. With a 10% ratio the multiplier is 1 ÷ 0.10 = 10. With a 20% ratio it is 1 ÷ 0.20 = 5, and the same $1,000 could become at most $5,000. A higher reserve ratio means less lending and less money created.

Notice what did and did not happen. No bank printed anything. Each bank only lent out money it had. Yet the total of everyone’s account balances is now far larger than the original cash. That is because a bank deposit is money, and each loan created a new deposit somewhere. The student’s textbooks, the mechanic’s parts and the bookstore’s balance are all real. This is the ordinary way most money comes into being in a modern economy: banks make loans, and the loans become deposits.

The textbook number is a ceiling, not a forecast. In real life some borrowers keep cash instead of depositing it, and banks often hold more reserves than the minimum, especially in nervous times. Both leaks shrink the multiplier. Since 2020 the Federal Reserve has set the required reserve ratio at zero, but banks still hold large reserves for safety and because the Fed pays interest on them. The lesson to keep is the direction: more lending means more money, and banks that pull back on loans shrink the money supply just as surely.

In the early 1930s a rumor could kill a bank. Suppose word spread that First Street Bank had made bad farm loans. Even a depositor who did not believe it had a reason to withdraw, because if enough others withdrew first, the vault would be empty when she arrived. So everyone hurried, the line stretched down the block, and a bank that might have been perfectly sound ran out of cash by noon. This is a bank run, and it is the flaw built into fractional reserve banking. Between 1930 and 1933, thousands of American banks closed, and depositors lost savings they had thought were safe.

In 1933 Congress created the Federal Deposit Insurance Corporation, the FDIC. It insures deposits at member banks, so if a bank fails, the FDIC pays depositors back, currently up to $250,000 per depositor, per bank, for each ownership category. The FDIC is funded by fees that banks pay, not by taxes. The genius of the design is that it changes the depositor’s reasoning. If your money is guaranteed whether or not you rush to the window, there is no reason to rush. Runs on insured deposits almost disappeared after 1934.

Insurance solves one problem and creates another. If depositors are protected no matter what, they stop caring how risky their bank is, and a bank owner might take bigger gambles with other people’s money. Economists call this moral hazard: protection against a loss can encourage the very behavior that causes it. That is why insurance comes with rules. Bank examiners inspect loan books, banks must hold capital of their own so that owners lose first when loans go bad, and regulators can shut a failing bank on a Friday and reopen it under a new owner on Monday.

Runs have not vanished; they have changed shape. In March 2023, a large California bank failed after depositors, many holding balances far above the insured limit, moved tens of billions of dollars out in about a day using phones and laptops. No line formed on the sidewalk. The lesson of 1933 still held: a bank whose depositors lose confidence cannot survive on its own, however sound its loans may be, and the speed of a modern run leaves regulators very little time.

Interest is the price of using someone else’s money for a while. Suppose you borrow $500 from a cousin to fix your bike, and she asks for 6% a year in simple interest. The $500 is the principal, the amount borrowed. The interest rate is 6% a year, written as 0.06. After one year you owe $500 × 0.06 = $30 in interest, so $530 in all. After two years, simple interest just doubles: $500 × 0.06 × 2 = $60, for a total of $560. The formula is Interest = Principal × Rate × Time, with time in years.

Work one more. A student lends $1,200 to a friend at 5% simple interest for three years. Interest is $1,200 × 0.05 × 3 = $180. The friend repays $1,380. Change one number at a time and watch what happens. Double the rate to 10% and the interest doubles to $360. Double the time to six years at 5% and the interest doubles to $360 as well. Simple interest grows in a straight line: the same $60 every year, because it is always figured on the original $1,200 and never on the interest already earned.

Time can be less than a year. If a payday lender charges $15 to borrow $100 for two weeks, that is 15% for two weeks. To compare it with a bank loan you convert to a yearly rate. There are about 26 two-week periods in a year, and 15% × 26 = 390% per year in simple terms. The dollar amount looked small; the rate is enormous. Always convert to a yearly rate before you compare two ways to borrow. That is exactly what the law requires lenders to print, as the next lessons show.

Two common mistakes. First, forgetting to convert the percent: 6% is 0.06, not 6, so $500 × 6 would be absurd. Second, treating the rate as a total rather than a yearly figure. A car loan at 7% for five years does not cost 7% once; simple interest would be 7% each year, or 35% of the principal over five years before the loan is paid down. Interest is a rate per year unless the lender clearly says otherwise.

Now change one rule. Instead of paying interest only on the principal, the bank pays interest on the whole balance, including interest already earned. This is compound interest. Put $1,000 in an account paying 5% compounded once a year. After year one: $1,000 × 1.05 = $1,050. After year two the 5% applies to $1,050, not $1,000: $1,050 × 1.05 = $1,102.50. After year three: $1,102.50 × 1.05 = $1,157.63. Simple interest would have given $1,150 after three years. Compounding gave $7.63 more. Small so far, but watch what time does.

Keep going. At 5% compounded yearly, $1,000 becomes about $1,629 after 10 years, about $2,653 after 20 years, and about $4,322 after 30 years. Simple interest at 5% for 30 years gives only $2,500. The gap opened slowly and then widened fast, because every year’s interest becomes next year’s principal. The formula is Balance = Principal × (1 + rate) raised to the number of years. You do not need to memorize it to see the shape: a curve that bends upward, not a straight line.

A handy shortcut is the rule of 72. Divide 72 by the yearly rate and you get roughly the number of years for money to double. At 6%, 72 ÷ 6 = 12 years. At 8%, about 9 years. At 3%, about 24 years. Use it to picture two students. Aisha puts $2,000 in an account earning 6% at age 18 and never adds another dollar. By 30 it is about $4,000; by 42 about $8,000; by 54 about $16,000; by 66 about $32,000. Her friend waits until 30 to put in the same $2,000 and reaches about $16,000 at 66. The twelve-year head start was worth one more doubling.

Compounding works against you exactly as hard when you are the borrower. A credit card balance of $1,000 at 24% a year, compounded monthly at 2% per month, grows to about $1,268 in a year if you pay nothing. Each month’s interest is added to the balance, and the next month’s 2% is charged on the larger amount. The same curve that builds a saver’s account digs a borrower’s hole. Which side of the curve you stand on is one of the most important money choices you will ever make.

A car dealer offers a loan at 1.2% a month. A credit union offers 12% a year. A payday storefront offers $20 per $100 for a month. Which is cheapest? To make loans comparable, federal law since 1968 requires lenders to state the APR, the annual percentage rate: the yearly cost of the loan, including interest and most required fees, on one scale. The dealer’s 1.2% a month is about 14.4% APR. The payday loan is 20% a month, roughly 240% APR. The credit union’s 12% wins. Never compare a monthly rate with a yearly one; find the APR on each offer first.

Why do two people get different APRs on the same car? Because the lender is guessing how likely each is to pay back, and a credit score is its main guess. In the United States the most common scores run from 300 to 850. Higher is better. A score is built from your borrowing history: whether you have paid bills on time, how much of your available credit you are using, how long you have had accounts, how many new accounts you have opened lately, and whether you have handled different types of credit. Payment history and how much you owe count the most.

The score has a price attached. Suppose two graduates each finance a $15,000 used car for five years. Dana’s score is high and she gets 6% APR; her interest over the loan comes to about $2,400. Luis’s score is low after some missed phone bills and he is offered 14% APR; his interest comes to about $5,900. Same car, same five years, and Luis pays roughly $3,500 more. A poor score also raises apartment deposits and insurance rates and can even cost a job offer. Building a good score is one of the cheapest ways to save money over a lifetime.

How you build one is not mysterious. Pay every bill by its due date, even the minimum, because one late payment reported to the credit bureaus can drop a score sharply. Keep card balances low compared with the limit; using less than about a third of your available credit is a common guideline. Keep old accounts open, since a long history helps. Do not apply for many cards at once. And check your credit report, which you are entitled to see free from each of the three national credit bureaus, for errors. The score rewards patience and reliability, not income.

Jordan has $3,000 from a summer job and two goals: a $1,500 deposit on an apartment next spring and a retirement fund he will not touch for forty years. Putting all of it in one place would be a mistake, because the two goals call for different tools. Money needed soon should be saved: kept in an insured account where the balance cannot fall, even if it earns only 1% to 4%. Money not needed for decades can be invested: put into assets such as stocks or bonds whose value moves up and down but that have historically grown faster than savings accounts over long stretches.

The key idea is the trade-off between risk and return. A savings account is nearly riskless, so it pays little. A bond is a loan to a company or a government that pays fixed interest; it carries some risk that the borrower fails to pay, so it usually pays more than a savings account. A stock is a share of ownership in a company; its value can double or fall by half, so investors demand a higher expected return to hold it. Over the past century, broad U.S. stock indexes have averaged returns well above bonds and savings, but with years of steep losses along the way. Higher expected return comes bundled with the chance of loss. There is no honest offer of high return with no risk.

Why does the time horizon matter so much? Because losses that would ruin a one-year plan usually wash out over forty years. If Jordan invested his apartment money and the market fell 30% that winter, he would sign the lease $450 short. If the same drop hits his retirement account at 19, he has decades for it to recover and keep compounding. So the rule is: match the tool to the timeline. Short goal, safe and boring. Long goal, accept the ups and downs for the growth.

One more habit protects investors: do not put everything in one company. Spreading money across many stocks and bonds is called diversification, and it lowers risk without lowering expected return much, because when one company stumbles another may thrive. A fund that holds hundreds of companies at once does this for you at low cost. The person who put a whole paycheck into one hot stock and the person who spread it across a broad fund both took risk, but only one of them could lose nearly everything on a single piece of bad news.

Return to Marisol’s jar. Every idea in this chapter is in that story. The coins were money: a medium of exchange, a unit of account and a store of value, though a store that inflation slowly drained. The bank was a financial intermediary that kept a fraction in reserve and lent the rest. Her deposit helped create new deposits elsewhere through the lending chain. The FDIC label meant she had no reason to run. The interest she earned was her share of the spread. And the choice between the jar, the account and a stock fund was the choice among no return, low return with no risk, and higher expected return with risk.

Interest rates tie the pieces together, because a rate is a price like any other. Savers supply funds; borrowers demand them. When many people want to borrow and few want to save, lenders can charge more, and rates rise. Higher rates then coax more saving and discourage some borrowing until the two balance. When savings are plentiful and borrowers are few, rates fall. The one big difference from the market for lemonade is that a central bank, the Federal Reserve, deliberately pushes the short-term rate up or down to steer the whole economy. The next chapter is about how it does that.

A few numbers to carry forward. Simple interest is principal × rate × time. Compound growth is principal × (1 + rate) to the number of years, and 72 divided by the rate roughly gives the doubling time. The money multiplier ceiling is 1 divided by the reserve ratio. The FDIC limit is $250,000 per depositor per bank. Convert every loan offer to an APR before comparing. Pay on time and keep balances low to build a credit score. Match the tool to the timeline: save for soon, invest for far away, and diversify.

Finally, notice how much of this system runs on trust. Fiat money works because people expect others to accept it. Fractional reserve banking works because depositors expect the bank to pay on demand. Lending works because lenders expect to be repaid, and they price that expectation as an interest rate. When trust holds, money moves from those who have it to those who can use it, and both sides gain. When trust breaks, as in 1933, the system seizes. Much of what governments and central banks do is aimed at keeping that trust intact.

In October 1907, a failed attempt to corner the stock of a copper company set off runs on several New York trust companies, which were bank-like firms. Depositors lined up to pull out cash, and one large trust company closed its doors. The United States had no central bank to lend cash to sound firms caught in a run. Instead, the banker J. P. Morgan gathered other bankers in his library and pressed them to pool money to stop the panic. It worked, barely. Congress drew a lesson: a country should not depend on one private banker to rescue its banking system.

After years of study and debate, Congress passed the Federal Reserve Act, and President Woodrow Wilson signed it in December 1913. The law created the Federal Reserve System, the central bank of the United States, usually just called the Fed. Its first job was to act as a lender of last resort: when a sound bank faces a sudden rush of withdrawals, it can borrow cash from the Fed, pledging its loans as security, instead of shutting its doors. Suppose a bank owes depositors $100 million but has $10 million in cash. If it can borrow $30 million from the Fed for a few weeks, it can pay everyone who asks and the panic fades.

Over time the Fed’s jobs grew to four. First, it conducts monetary policy: it manages short-term interest rates and the amount of money and credit in the economy to keep prices stable and employment high. Second, it supervises and regulates many banks. Third, it runs much of the payment system that clears checks and moves electronic payments between banks. Fourth, it is the bank for banks and for the U.S. Treasury, holding their accounts and handling their payments.

Two common mistakes. The Fed does not physically print dollar bills; the Treasury’s Bureau of Engraving and Printing does, and the Fed orders and distributes them. And the Fed does not set taxes or decide how the government spends money. That is fiscal policy, made by Congress and the President. The Fed’s lever is the price and quantity of money and credit, not the budget.

The Fed is built on purpose to spread power out. At the top is the Board of Governors in Washington: seven members nominated by the President and confirmed by the Senate. Each governor serves a single 14-year term, and the terms are staggered so that one begins every two years. Do the arithmetic: 7 seats times 2 years is 14 years. A President who serves one four-year term can normally fill only two seats, unless governors leave early. One governor is named Chair and serves four years in that role, also with Senate approval.

Below the Board are twelve regional Federal Reserve Banks, each serving a district: Boston, New York, Philadelphia, Cleveland, Richmond, Atlanta, Chicago, St. Louis, Minneapolis, Kansas City, Dallas and San Francisco. The Federal Reserve Bank of Chicago, on LaSalle Street, serves the Seventh District: all of Iowa and most of Illinois, Indiana, Michigan and Wisconsin. Reserve Banks examine banks, supply cash to banks in their district, and study the local economy. Eight times a year the Fed publishes the Beige Book, a report that gathers what each district’s businesses and bankers are seeing.

Monetary policy is made by the Federal Open Market Committee, the FOMC. It has twelve voting members: the seven governors, the president of the New York Fed, and four of the other eleven Reserve Bank presidents, who take turns voting for one-year terms. That is 7 + 1 + 4 = 12. The Chicago Fed president votes every other year, alternating with Cleveland. All twelve presidents sit at the table and join the discussion every time, even in years they do not vote. The committee holds eight scheduled meetings a year and can meet in between if something urgent happens.

A common mistake is to picture the Fed as one bank in one building, or as an ordinary cabinet department that takes orders from the President. It is neither. It is a system of public and regional parts, created by Congress and answerable to Congress, but built so that no single official, region or election can easily control it. The next lessons show why that design matters.

In 1977 Congress wrote the Fed’s goals into the Federal Reserve Act: maximum employment, stable prices and moderate long-term interest rates. Because steady prices and strong employment tend to keep long-term rates moderate on their own, people usually speak of a dual mandate, meaning two goals. Maximum employment means the highest level of employment the economy can sustain without pushing inflation up. Price stability means prices rise slowly and predictably enough that families and businesses can plan without worrying about them.

Stable prices does not mean zero inflation. In 2012 the Fed announced an inflation target of 2% a year, measured by a price index for personal consumption spending. At 2%, a basket of goods that costs $100 this year costs $102 next year and about $121.90 after ten years ($100 × 1.02 raised to the 10th power). Why not aim for 0%? First, price indexes tend to overstate inflation a little, because they miss some quality improvements. Second, a little inflation keeps interest rates a bit higher on average, which leaves the Fed room to cut them in a slump. Third, falling prices, called deflation, can be dangerous: people delay purchases waiting for lower prices, and debts become harder to repay.

Maximum employment has no fixed number. It depends on things the Fed cannot control, such as workers’ skills and how quickly people find new jobs. So the Fed estimates it and revises its estimate as the economy changes. When unemployment is high and inflation is low, both goals call for the same move: lower rates to encourage spending. The hard case is when inflation is high and unemployment is also rising, as in the oil shocks of the 1970s. Then fighting inflation costs jobs, and protecting jobs lets inflation run. The Fed has to weigh the two.

Every bank has an account at the Fed, the way you have a checking account at a bank. The money in those accounts is called reserves. At the end of a day, one bank may have more reserves than it wants and another may be short. So they lend to each other overnight. Suppose Bank A lends Bank B $50 million for one night at a yearly rate of 4.35% (an example). For one night that earns about $6,000: $50,000,000 × 0.0435 ÷ 360 days, the convention used in this market. The interest rate on these overnight loans between banks is the federal funds rate.

The FOMC does not order banks to charge a particular rate. It announces a target range, such as 4.25% to 4.50% (an example), and then uses its other tools to keep the actual rate inside that range. The range is usually one quarter of a percentage point wide. Traders measure rate changes in basis points: one basis point is one hundredth of a percentage point, so a quarter point is 25 basis points. When the Fed raises the target by 25 basis points, the range in our example moves to 4.50% to 4.75%.

Why should a car buyer care about an overnight rate between banks? Because it sets the floor under many other short-term rates. When banks’ own cost of money rises, they charge more on loans. Many banks set their prime rate, the rate for their most creditworthy business borrowers, about 3 percentage points above the top of the Fed’s range. Credit card rates are often tied to the prime rate, so they move within weeks of an FOMC decision.

Watch one common mistake: the difference between percentage points and percent. If a rate rises from 4% to 4.5%, it rose by half a percentage point, or 50 basis points. But measured as a percent change, it rose 12.5%, because 0.5 is 12.5% of 4. News reports usually mean percentage points. When you hear the Fed raised rates by half a point, it added 0.5 to the rate, not 0.5% of the rate.

Before 2008, banks held only small amounts of reserves, and the Fed steered the federal funds rate by adding or removing a few billion dollars of reserves, making them a little scarcer or more plentiful. After the 2008 crisis, the Fed’s large bond purchases filled banks’ accounts with trillions of dollars in reserves. With so many reserves around, small changes in quantity no longer moved the rate. The Fed needed a different handle. Congress had authorized it to pay interest on reserves, and it began doing so in October 2008.

Here is why that handle works. Suppose the Fed pays interest on reserve balances of 4.40% a year (an example). A bank with $100 million in reserves earns $4.4 million a year just by leaving the money at the Fed, with no risk at all. Would that bank lend reserves overnight to another bank at 3%? No. It would keep them at the Fed and earn 4.40%. So the rate the Fed pays sets a floor that holds up the federal funds rate. If the Fed raises the rate it pays to 4.65%, banks demand more from any borrower, and the federal funds rate climbs with it.

Some lenders in the overnight market, such as money market funds, cannot hold reserves at the Fed and so cannot earn that rate. For them the Fed offers a second facility, called overnight reverse repurchase agreements, which pays a rate at the bottom of the target range. Together these two administered rates, rates the Fed simply announces, keep the federal funds rate inside the range. The Fed calls this way of operating an ample reserves system, because it keeps reserves plentiful and controls the rate by setting what it pays rather than by rationing quantity.

The common mistake is to think the Fed still works mainly by adding and draining small amounts of reserves each day, as older textbooks describe. It can still do that, but since 2008 the everyday lever is the rate it pays. When you read that the FOMC raised rates, what changed that afternoon was the rate the Fed pays on reserves and on its overnight facility.

Suppose the Fed buys $1 billion of U.S. Treasury bonds from a securities dealer. It pays by adding $1 billion to the reserve account of the dealer’s bank. No paper changes hands; the Fed creates the reserves with a keystroke. The banking system now has $1 billion more in reserves than before. If the Fed sells $1 billion of bonds instead, the buyer’s bank pays with its reserves, and $1 billion of reserves disappears. These purchases and sales are open market operations, named because the Fed trades with dealers in the open market rather than buying straight from the Treasury. The trading desk at the New York Fed carries them out.

Buying bonds also changes interest rates, because bond prices and interest rates move in opposite directions. A Treasury security is a loan to the federal government. Suppose a bond pays $40 a year and you buy it for $1,000: your yield, the yearly return, is 4%. If heavy buying by the Fed pushes that bond’s price up to $1,050, the same $40 a year is now only about 3.8% of what a new buyer pays ($40 ÷ $1,050). Higher bond prices mean lower yields, and lower yields on government bonds pull down rates on mortgages and business loans that follow them.

When short-term rates were already near zero in 2008 and again in 2020, the Fed used this tool on a giant scale. It bought trillions of dollars of long-term Treasury bonds and mortgage-backed securities to push long-term rates down. This is called quantitative easing. The reverse, letting bonds mature without buying new ones so that the Fed’s holdings shrink, is called quantitative tightening. The Fed’s total holdings went from under $1 trillion before 2008 to nearly $9 trillion in 2022.

A bank that runs short of cash can borrow straight from its Reserve Bank at what is called the discount window. It must pledge collateral, assets such as loans or bonds that the Fed can keep if the loan is not repaid. The rate charged is the discount rate. For the soundest banks, it is usually set at the top of the FOMC’s target range. Suppose a bank in Rockford faces heavy withdrawals and borrows $50 million overnight at 4.50% (an example), pledging $70 million of business loans. It pays the Fed about $6,250 for the night and meets every depositor’s request. This is the lender of last resort at work.

For most of the twentieth century, the Fed also set reserve requirements: a minimum share of checking deposits that each bank had to keep as reserves. For large banks the requirement on checking accounts was 10% for many years. Raising the requirement meant banks could lend less. Remember the money multiplier from the last chapter: at a 10% requirement, the ceiling is 1 ÷ 0.10 = 10; at 12.5%, it is 1 ÷ 0.125 = 8. A small change in the rule could shift lending across the whole country, which made this tool blunt and rarely used for fine steering.

In March 2020 the Fed cut reserve requirements to zero, where they remain. Banks did not empty their accounts. They still hold large reserves, because they need cash to pay depositors, because other banking rules require them to hold safe, easy-to-sell assets, and because the Fed pays interest on reserves. So a modern textbook question about the multiplier shows how lending creates deposits, but the Fed no longer uses the requirement to steer the economy. Its everyday tools are the rates it pays, backed by open market operations and the discount window.

Follow one quarter-point increase. On Wednesday afternoon, the Fed raises the rates it pays on reserves and its overnight facility, and by Thursday the federal funds rate is a quarter point higher. Within days, banks raise their prime rate by the same quarter point, and rates on credit cards and many business lines of credit follow. This first link is quick and nearly one-for-one. The path from the Fed to spending is called monetary transmission, and the later links are slower and looser.

Car loans come next, but less tightly. Take Denise’s $25,000 used SUV, financed for 60 months. At 5% her monthly payment is about $471.78; at 6%, about $483.32; at 7%, about $495.03. Each extra percentage point adds roughly $12 a month, or about $700 in total interest over five years. Total interest at 6% is about $3,999; at 7% it is about $4,702. For one buyer, $12 a month may not change the decision. For the buyer who was already stretching, it can be the difference between buying now and waiting.

A mortgage, a loan to buy a home, works differently. A 30-year mortgage rate tends to follow the yield on ten-year Treasury bonds, and that yield depends on what investors expect the Fed to do over many years, not just at the next meeting. So mortgage rates sometimes rise before the Fed acts, when markets expect it to, and sometimes barely move after a decision everyone saw coming. The common mistake is to say the Fed sets mortgage rates. It does not; it moves the short-term rate and influences expectations, and markets set the rest.

Now add up the effects. Higher rates discourage borrowing for cars, homes and business equipment. They reward saving, so some households spend less. They tend to raise the value of the dollar against other currencies, which makes American exports more expensive abroad. Together these reduce total spending in the economy. With less spending, businesses hire more slowly and raise prices more slowly. That last step, slower inflation, is the goal, and it can take a year or more to arrive.

When unemployment is rising and inflation is below 2%, the Fed uses expansionary monetary policy, also called easing: it cuts its target so that borrowing is cheaper and spending picks up. When inflation is running above 2% and the economy is hot, it uses contractionary monetary policy, or tightening: it raises its target so that borrowing costs more and spending cools. In the language of aggregate demand, the total spending in the economy at each price level, easing pushes demand up and tightening pulls it down.

What borrowers and savers really care about is the real interest rate: the stated, or nominal, rate minus inflation. Suppose you save at 5% while prices rise 3%. Your money grows 5%, but it buys only about 2% more, so the real rate is about 2%. Now suppose a savings account pays 0.5% while inflation is 6%. The real rate is about 0.5% minus 6%, or negative 5.5%: your savings lose buying power every year. A central bank that holds its rate far below inflation is easing, even if its rate is rising.

The United States has seen both kinds of hard tightening. In 1979, with inflation above 11%, the Fed under Chair Paul Volcker pushed short-term rates close to 20%. Inflation fell to under 4% by 1983, but the cost was a deep recession, with unemployment near 11% late in 1982. Four decades later, inflation climbed to about 9% in 2022. The Fed raised its target from near zero to above 5% in about sixteen months, its fastest increases since the early 1980s. Inflation fell to around 3% within about two years, while unemployment stayed under 4% for most of that time.

Economists still debate how much of the 2022 to 2024 decline in inflation came from the Fed’s increases and how much from supply problems easing as the pandemic faded. What most agree on is that tightening works mainly through slower spending, and that its cost, in lost jobs or slower growth, depends on how far inflation must fall and how much people trust the Fed to follow through.

Picture a shower with a slow valve. You turn it hotter, nothing happens, so you turn it more, and a minute later you are scalded. Monetary policy works the same way. A rate change reaches bank rates within days, but its full effect on spending, hiring and prices can take a year or more. The economist Milton Friedman called these policy lags long and variable. The danger is that the Fed keeps tightening because inflation is still high, not seeing that its earlier moves have not yet worked, and then overshoots into a recession.

Rates also have a floor. The Fed cannot push its target much below zero, because people and banks would rather hold cash, which pays 0%, than lend at a negative rate. Economists call this the zero lower bound. In December 2008 the Fed cut its target to a range of 0% to 0.25% and kept it there for seven years. With no room to cut, it turned to other tools: buying long-term bonds, and forward guidance, which means telling the public where rates are likely to go. If families and businesses believe rates will stay low for years, long-term rates fall today.

The hardest case is a supply shock, a sudden change in the cost or availability of key inputs. Suppose oil prices double. Gasoline and shipping cost more, so inflation jumps, and at the same time businesses cut output because their costs rose. Should the Fed raise rates to fight inflation, deepening the slump, or cut rates to protect jobs, letting inflation climb? There is no answer that helps both goals. Most central banks look past a shock they expect to fade, but act if people start to expect high inflation to continue.

One more limit: the Fed sets one rate for the whole country. If factories in Peoria are laying off workers while a fast-growing city elsewhere is overheating, the same rate applies to both. Monetary policy cannot aim at one region, one industry or one group of workers. Those targeted jobs belong to fiscal policy, made by Congress and the President.

By 2007, falling home prices meant that many mortgages would never be repaid, and banks and investment firms that held them began to take large losses. In September 2008, the investment bank Lehman Brothers failed, and lending between financial firms nearly froze. The Fed acted as lender of last resort on a scale it had never tried. It cut its target to a range of 0% to 0.25% by December 2008, lent to banks and to other kinds of financial firms, and began buying mortgage-backed securities and Treasury bonds. Its balance sheet, the list of everything it owns and owes, grew from under $1 trillion to about $4.5 trillion by late 2014.

In March 2020, as the pandemic shut down much of the economy, the Fed moved faster. In two emergency meetings that month it cut its target back to 0% to 0.25%. It bought Treasury bonds and mortgage-backed securities at a record pace and set up emergency lending programs for businesses and for state and local governments, several of them backed by money Congress provided through the Treasury. It also cut reserve requirements to zero. Its balance sheet climbed from about $4.2 trillion to nearly $9 trillion by 2022.

Supporters argue that these actions kept a severe financial crisis from turning into a second Great Depression and helped the job market recover faster after 2020 than after 2008. Critics raise several concerns. They argue that years of very low rates and bond buying pushed up the prices of stocks and houses, helping people who already owned them. Rescuing firms that took big risks may encourage more risk next time, the moral hazard problem from the last chapter. And some argue the Fed kept policy too easy for too long in 2021, adding to the inflation that followed, alongside government spending and supply problems. Economists weigh these arguments differently.

Imagine a mayor facing reelection in November who could set interest rates. Cutting rates in the spring would bring cheaper car loans and more hiring by fall. The higher inflation would show up the next year, after the votes were counted. That temptation is why most countries give monetary policy to a central bank with central bank independence, meaning its decisions do not need approval from the President or Congress. Studies across many countries have found that more independent central banks have tended to deliver lower inflation.

The Fed’s independence is built into its design. Governors serve 14-year terms, longer than any President, and the law allows them to be removed only for cause, not over a policy disagreement. The Fed pays its own costs from the interest it earns on the bonds it holds, instead of asking Congress for money each year, and it sends its leftover earnings to the Treasury. No official outside the FOMC can overturn its rate decisions.

Independence is not the same as being unaccountable. Congress created the Fed and can change its law at any time. The Chair testifies to Congress twice a year on monetary policy. After every FOMC meeting the Fed releases a statement and the Chair holds a press conference. Detailed minutes come out three weeks later, and full transcripts are released after five years. This openness is a form of accountability: the public can judge whether the Fed is meeting the goals Congress set, and Congress can respond.

People disagree about the balance. Some argue that unelected officials hold too much power over jobs and savings, and want more oversight, broader audits or narrower goals. Others argue that closer political control would bring the election-year temptation back and, over time, higher inflation. Both sides accept that the Fed’s power comes from Congress, and that its credibility depends on people believing it will do what it says.