Central Banks: Functions, Monetary Policy Tools, Independence and Credibility (Complete Guide)

Lender of Last Resort — in a financial panic, the central bank should lend freely, at a penalty rate, against good collateral. It exists because a bank run is a coordination failure: a solvent bank fails simply because everyone expects it to fail.

A commercial bank takes deposits and makes loans, seeking profit. A central bank is the monopoly supplier of the reserves in which commercial banks settle with one another, and it pursues public objectives, not profit. It cannot run out of the currency it issues. This is not a matter of degree — it is a difference in kind, and it is why the central bank can always act as lender of last resort in its own currency, and never in a foreign one.

The central bank buys $1,000 of bonds from a bank. Reserve ratio is 10%.

Geometric series: 1,000 × (1 + 0.9 + 0.9² + …) = 1,000 × 1/(1−0.9) = $10,000

Money multiplier m = 1/0.10 = 10. An injection of $1,000 of base money supports $10,000 of deposits.

McLeay, M., Radia, A., & Thomas, R. (2014), “Money Creation in the Modern Economy,” Bank of England Quarterly Bulletin, Q1

The claim: Three economists at the Bank of England published an article stating plainly that the money multiplier account taught in introductory textbooks describes the causation backwards.

What actually happens: A commercial bank does not wait to receive a deposit before lending. When it grants a loan, it simultaneously creates a matching deposit in the borrower’s account by keystroke. Loans create deposits — not the reverse. The bank then acquires whatever reserves it needs, after the fact, in the interbank market or from the central bank.

The binding constraint is therefore not reserves. It is (i) the profitability of lending at the prevailing interest rate, (ii) the bank’s capital requirement, and (iii) borrower demand for credit. The central bank influences money creation by setting the price of reserves, not by rationing their quantity.

Why it matters: This resolves the great puzzle of the QE era. Central banks expanded the monetary base enormously and broad money barely responded, because reserves were never the constraint. The multiplier is not a mechanism. It is an ex post ratio between two numbers, and it moves around freely.

AP and Cambridge both still require the money multiplier calculation. Do it. Then, in an evaluation paragraph, note that the Bank of England’s own research holds that loans create deposits and reserves adjust endogenously — and that most advanced economies now impose no reserve requirement at all, which makes 1/r undefined. Demonstrating you know both the model and its limitations is exactly what top bands reward.

The total coefficient on inflation is 1.5, not 1. If inflation rises one point, the nominal rate must rise by more than one point so that the real rate rises and policy actually tightens. A bank that matches nominal rates to inflation one-for-one has done nothing. Raising them less than one-for-one loosens policy into an inflation. Violating this principle is the standard explanation for the Great Inflation of the 1970s.

Kydland, F. E., & Prescott, E. C. (1977), “Rules Rather than Discretion: The Inconsistency of Optimal Plans,” Journal of Political Economy 85(3). Nobel Prize 2004.

The set-up: A government announces it will deliver low inflation. Workers believe it and settle for modest nominal wage increases. Now the government has an incentive to renege: with wages locked in, a burst of unexpected inflation cuts real wages, raises employment, and looks like competent management.

The unravelling: Workers are not fools. They anticipate the temptation, and demand higher nominal wages up front. The government, facing those higher wages, delivers the inflation anyway — because not doing so would now cause a recession.

The equilibrium: high inflation, and no gain in employment whatsoever. Society ends up strictly worse off than under a binding commitment to low inflation.

The lesson: The problem is not bad policymakers. It is the discretion that good policymakers cannot credibly renounce. This is the intellectual foundation of central bank independence — and it is the same structure as the non-credible threat in game theory. Barro & Gordon (1983) formalised it as a repeated game; Rogoff (1985) proposed the solution of delegating policy to a “conservative” central banker who weights inflation more heavily than society does.

Goal independence: the bank chooses its own objective. Rare — the ECB has it; most do not.

Instrument independence: the government sets the target (e.g. 2% inflation); the bank chooses how to hit it, free from political direction. This is the standard arrangement, adopted by the Bank of England in 1997. It preserves democratic accountability over ends while removing political interference in means.

The inheritance: US inflation in double digits. A decade of Federal Reserve chairs had raised rates, then retreated the moment unemployment rose. Nobody believed a disinflation would be seen through. Inflation expectations were unanchored.

The action: Volcker raised the federal funds rate to unprecedented levels — above 19% at its peak — and, critically, held them there through a severe recession. Unemployment exceeded 10%. Farmers drove tractors to Washington and blockaded the Fed building. Congress threatened impeachment.

The outcome: Inflation fell from over 13% to under 4% within three years, and stayed down for a generation.

The theoretical significance: The sacrifice ratio — output lost per point of disinflation — is not a fixed structural parameter of an economy. It depends on credibility. Volcker’s recession was expensive precisely because nobody initially believed him. Once the belief was established, it became a durable public asset that his successors inherited for free. Credibility is capital: costly to build, cheap to hold, and ruinous to lose.

The crisis: Yields on Italian and Spanish government debt were spiralling. The mechanism was a self-fulfilling panic: investors feared the euro would break up, so they demanded higher yields; higher yields made default more likely; which made break-up more likely. A textbook multiple-equilibrium problem.

The intervention: ECB President Mario Draghi told a conference the ECB was ready to do “whatever it takes” to preserve the euro. He added that it would be enough.

The result: Spreads collapsed almost immediately. The ECB subsequently announced the Outright Monetary Transactions programme — under which it never purchased a single bond.

Why it worked: The panic equilibrium existed only because investors believed no one would backstop the market. A credible commitment to unlimited intervention eliminates the bad equilibrium — and because the bad equilibrium no longer exists, the intervention never has to happen. This is Diamond–Dybvig (1983) logic applied to sovereigns: deposit insurance works precisely to the extent that it is never claimed. The central bank’s most powerful instrument turned out to be a sentence.

1. The democratic deficit. Unelected officials make decisions with vast distributional consequences. Instrument independence answers this in theory; in practice, QE and macroprudential regulation are hard to describe as purely technical.

2. Distributional effects. QE raises asset prices. Assets are held by the wealthy. The Bank of England’s own analysis conceded this, while arguing that employment gains offset it for the poor. Both may be true; the net effect requires weights economics cannot supply.

3. The zero lower bound. Conventional policy runs out of room. Everything below the ZLB is improvisation.

4. Long and variable lags. Monetary policy affects inflation with a delay of 12–24 months. A central bank is always steering by a rear-view mirror.

5. Fiscal entanglement. A bank holding vast quantities of government debt pays interest on the reserves it created while earning the old, lower coupon. When rates rise it makes losses requiring Treasury indemnification. QE effectively converted long-term fixed-rate public debt into overnight floating-rate debt — a real cost, almost wholly absent from the 2009 debate.

6. Mandate creep. Central banks are now asked to address climate risk, inequality and financial stability. Each additional objective dilutes the accountability that justified independence in the first place. Tinbergen’s rule is unforgiving: you need at least as many instruments as targets.

The reserve ratio is 20%. The public holds cash equal to 5% of deposits. The central bank injects $500 of base money. Calculate the change in the money supply.

Given r* = 2%, π* = 2%, current π = 7%, output gap = −1%. (a) What rate does the Taylor rule imply? (b) The bank instead sets 7%. Explain, with numbers, why this fails to tighten policy.

Bank A holds long-term government bonds worth $100m at maturity but currently trading at $85m because rates have risen. Depositors demand $90m. Bank B has made $100m of loans to borrowers who will certainly default. Depositors demand $90m. Should the central bank lend to either?

• Compute the money multiplier, including with a cash drain and excess reserves.
• Show open market operations on a money-market diagram: buying bonds shifts MS right, lowering the nominal rate.
• Trace the transmission: lower i → higher I and C → AD shifts right → P and Y rise.
• Know the three conventional tools: OMO, discount rate, reserve requirement.
• At the zero lower bound, MS is effectively horizontal — further injections do not lower i.

“Evaluate the case for central bank independence.”

1. Define independence, and distinguish goal from instrument independence. Most students miss this and cap their marks.
2. Explain time inconsistency (Kydland & Prescott): the government’s temptation to inflate is anticipated, so society gets the inflation without the employment. This is the analytical core.
3. Note the game-theoretic structure: it is a non-credible commitment, resolved by delegation (Rogoff’s conservative central banker).
4. Cite Alesina & Summers (1993): independence correlates with lower inflation and no output cost.
5. Use Volcker to show that credibility determines the sacrifice ratio, and Draghi to show that credibility can substitute entirely for action.
6. Evaluate: the democratic deficit; distributional effects of QE; mandate creep and Tinbergen’s rule; the reverse-causation objection to Alesina–Summers.
7. Conclude conditionally: independence solves a commitment problem for monetary policy narrowly defined. It is a weaker justification the further the bank strays from that mandate.

• “QE is printing money.” QE creates reserves, held by banks at the central bank. Households cannot spend reserves. This distinction explains why a decade of QE produced no inflation while eighteen months of fiscal transfers produced a great deal.
• Treating the money multiplier as causal. Loans create deposits; reserves adjust afterwards.
• Claiming a rate rise “fights inflation” without checking the real rate. See the Taylor principle.
• Saying a central bank can go bankrupt in its own currency. It cannot. It can make accounting losses; that is different.
• Confusing independence with unaccountability. Instrument independence requires a democratically set target.
• Recommending a bailout of an insolvent bank as “Bagehot.” Bagehot said the opposite.

Exercise 1 — Does Independence Cause Low Inflation, or Do Inflation-Averse Societies Build Independent Banks?

Alesina and Summers found a strong negative correlation between central bank independence and average inflation across the OECD, with no cost in output volatility. This is presented as the empirical case for independence.

But independence is not randomly assigned. Germany’s hostility to inflation, forged in the 1920s, produced both the Bundesbank and a low-inflation political culture. The correlation may reflect a common cause. Construct an identification strategy that would isolate the causal effect of independence — what instrument, natural experiment, or discontinuity could you exploit? If no credible strategy exists, on what basis do we believe the policy works at all?

📄 Read: Alesina, A., & Summers, L. H. (1993). “Central Bank Independence and Macroeconomic Performance: Some Comparative Evidence.” Journal of Money, Credit and Banking, 25(2), 151–162. Read alongside Kydland & Prescott (1977) and ask whether the theory would still justify independence if the empirical correlation vanished.

Exercise 2 — What Is the Cost of a Promise That Is Never Tested?

Draghi eliminated a self-fulfilling panic by promising unlimited intervention. The ECB never bought a bond under the programme. On the ledger, the intervention was free.

But a commitment that would be honoured has real costs in the states of the world where it is called upon — and those states did not occur precisely because of the commitment. Is the expected cost therefore zero? Consider moral hazard: if governments know the central bank will backstop their debt, does the promise itself create the fiscal indiscipline it was designed to insure against? Formalise this as a game, and identify whether OMT’s conditionality requirements solve it.

📄 Read: Diamond, D. W., & Dybvig, P. H. (1983). “Bank Runs, Deposit Insurance, and Liquidity.” Journal of Political Economy, 91(3), 401–419. Nobel Prize 2022. Map their two-equilibrium structure onto sovereign debt markets and identify exactly which assumption does the work.