Skip to content
Menu
Finabase Learn

Economics · Foundations

Recession vs. depression: how deep, how long, and who decides

What separates a recession from a depression, with the fall in real GDP in the Great Depression and in all twelve US recessions since 1948 set side by side.

On this page
  1. What a recession is, and who says so
  2. What a depression is
  3. The numbers side by side
  4. Worked example: what a 26% fall means
  5. Recession vs. depression at a glance
  6. Which word to use when
  7. Pitfalls
  8. How Finabase does it
  9. Frequently asked questions
  10. Summary
  11. Check your understanding

Every downturn brings the same question in the news: is this a recession, or something worse? The two words describe the same kind of event, a broad fall in economic activity, at very different sizes. A recession is a normal part of the business cycle; the United States has had twelve since 1948. A depression is a rare and catastrophic one. The country has had one in living memory, the Great Depression of the 1930s, and nothing since has come close.

What surprises most people is that only one of the two words has an official keeper. A committee of economists decides when US recessions begin and end. Nobody decides when a depression begins. This page explains both, then puts the numbers side by side: how far output fell, for how long, and what happened to jobs and prices.

What a recession is, and who says so

In the United States, recessions are dated by the National Bureau of Economic Research, a private non-profit research organization. Its Business Cycle Dating Committee defines a recession as a significant decline in economic activity that is spread across the economy and lasts more than a few months. It looks at employment, real income, spending, industrial production and other monthly measures, not just GDP, and it announces its decisions long after the fact, often a year later.

The popular rule of "two quarters of falling GDP" is not the committee's definition. It is a rough signal that usually agrees with the official dates and sometimes doesn't. What is a recession? How the NBER dates US recessions covers how the committee works and when the rule of thumb failed.

What a depression is

There is no committee, no formula and no official list. Economists use "depression" for a downturn so severe and long-lasting that "recession" doesn't describe it. Two informal yardsticks circulate: a fall in real GDP of more than 10%, or a contraction lasting more than two or three years. Neither is a rule anyone enforces. In practice, in the United States, "the Depression" means one event.

The NBER does date the Great Depression's two contractions as ordinary cycles: from August 1929 to March 1933 (43 months) and from May 1937 to June 1938 (13 months). What sets them apart is everything the dates leave out.

The numbers side by side

Fall in US real GDP, peak to trough: the Great Depression and every recession since 1948 1929–33 (annual data) −26.3% · 43 months 1937–38 (annual data) −3.3% · 13 months 1948-11 to 1949-10 −1.7% · 11 months 1953-07 to 1954-05 −2.5% · 10 months 1957-08 to 1958-04 −3.6% · 8 months 1960-04 to 1961-02 −1.3% · 10 months 1969-12 to 1970-11 −1.1% · 11 months 1973-11 to 1975-03 −3.1% · 16 months 1980-01 to 1980-07 −2.2% · 6 months 1981-07 to 1982-11 −2.6% · 16 months 1990-07 to 1991-03 −1.4% · 8 months 2001-03 to 2001-11 −0.4% · 8 months 2007-12 to 2009-06 −4.0% · 18 months 2020-02 to 2020-04 −9.1% · 2 months Real GDP: BEA via FRED (GDPCA annual from 1929; GDPC1 quarterly from 1947). Dates: NBER.
Fall in real GDP from the highest to the lowest quarter around each NBER recession since 1948, and from 1929 to 1933 and 1937 to 1938 in annual data. Months are the NBER peak-to-trough length. Data: BEA via FRED; dates: NBER.

The chart compares the fall in real GDP, output adjusted for inflation. The Depression figures come from annual data, which smooth out the worst quarters, so its true peak-to-trough fall was if anything larger than shown.

Downturn Months of contraction Fall in real GDP Peak unemployment
1929–33 (Depression) 43 −26.3% (annual) About 25% (1933)
1973–75 16 −3.1% 9.0% (May 1975)
1981–82 16 −2.6% 10.8% (November 1982)
2007–09 18 −4.0% 10.0% (October 2009)
2020 2 −9.1% 14.8% (April 2020)

Three things stand out:

  1. Depth. The worst post-war recession by output, 2020, cut real GDP by about a third as much as the Depression did, and the typical post-war recession by a tenth or less.
  2. Length. The longest post-war contraction lasted 18 months. The Depression's first leg lasted 43, and in annual data real GDP didn't regain its 1929 level until 1936, then fell again in 1938.
  3. Prices. Consumer prices fell 27.2% from August 1929 to March 1933. No post-war recession has seen more than a brief dip in the yearly inflation rate below zero. Falling prices raise the real burden of debt, which is one reason the Depression fed on itself (Deflation: falling prices, debt deflation and why the Fed aims for 2%).

Worked example: what a 26% fall means

Real GDP in 1929 was about $1,191 billion in today's chained 2017 dollars, and $877 billion in 1933. Imagine the same proportion applied to the US economy of 2026: a fall of 26.3% from the third quarter's real GDP of about $24.6 trillion would remove about $6.5 trillion of yearly output, more than the entire annual output of most countries.

Set that against 2007–09, the downturn that became known as the Great Recession. Real GDP fell 4.0% from peak to trough, and unemployment doubled to 10.0%. It was the worst downturn in seventy years, and it was still roughly one-sixth the size of the Depression's fall in output. The name "Great Recession" itself reflects the judgment that, terrible as it was, it was not a depression.

Policy is the other big difference between then and now. In the early 1930s the Federal Reserve let the money supply shrink as banks failed, there was no federal deposit insurance, and there were no unemployment benefits or other automatic stabilizers to cushion incomes. Since then the FDIC, unemployment insurance and a central bank that cuts rates and lends in a panic have changed how deep a downturn can go. Economists still argue over how much each mattered, but every post-war recession has met a far larger policy response (Fiscal policy in recessions: 2001, 2008–09 and 2020 compared).

Recession vs. depression at a glance

Feature Recession Depression
Official definition Yes, NBER committee (US) None
Typical fall in real GDP 0.4% to 4% since 1948 (9.1% in 2020) 10% or more; 26.3% in 1929–33
Typical length 2 to 18 months since 1948 Years
Unemployment Rises a few points, often to 6–10% About 25% in 1933
Prices Inflation usually slows Often falling (deflation)
Banks Some stress Waves of failures
US examples 12 since 1948 1929–33 (and the relapse of 1937–38)

Which word to use when

  • News of a downturn now: "recession", and only once the NBER has dated it if you want the official label. Until then, "a possible recession" is accurate.
  • Comparing with history: use the numbers, not the label: how far real GDP fell, how long it took, how high unemployment rose.
  • "Depression": reserve it for a downturn that approaches the 1930s in scale, or for national cases such as Greece after 2008, whose economy shrank for several years in a row.
  • Markets: stocks often fall before a recession is announced and recover before it ends. Dating matters less to an investor than the direction of earnings and interest rates (How the economy affects stocks: growth, inflation, jobs and rates).

Pitfalls

  • Treating two negative quarters as the definition. It isn't, in the US. 2022 had two negative GDP quarters in the first estimates and no recession was declared.
  • Comparing quarterly and annual numbers directly. Annual averages understate peak-to-trough falls.
  • Calling every sharp fall a depression. The 2020 fall was the deepest since the 1940s but lasted two months; "depression" implies years.
  • Forgetting revisions. GDP is revised for years. The figures here are the current vintage, frozen in October 2026.

How Finabase does it

Finabase does not carry a GDP history or recession dates. The economic calendar on the Economy page lists the GDP release, the jobs report and other indicators with actual, forecast and previous values, and its unemployment chart shows the latest monthly rate. The markets' reading of recession risk is on the Treasury page, where an inverted yield curve has preceded most recessions (Yield curve inversion: what it signals and its recession record). The chart and table on this page are ours, built from BEA and BLS series on FRED and the NBER's published dates.

Frequently asked questions

What is the difference between a recession and a depression?

Size and length. A recession is a broad decline in activity lasting more than a few months; since 1948 US recessions have cut real GDP by 0.4% to 9.1%. A depression is a far deeper and longer downturn: in the Great Depression real GDP fell 26.3% from 1929 to 1933.

Who decides if the US is in a depression?

Nobody officially. The NBER dates recessions; "depression" is an informal term with no agreed threshold.

Was 2008 a depression?

No. The 2007–09 recession was the longest since the 1930s, at 18 months, with real GDP down 4.0% and unemployment at 10.0%. Those are severe recession numbers, far from the Depression's.

How long did the Great Depression last?

The NBER dates the first contraction from August 1929 to March 1933 and a second from May 1937 to June 1938. In annual data real GDP regained its 1929 level in 1936, but unemployment stayed high until the early 1940s.

Summary

A recession is a significant, widespread decline in economic activity lasting more than a few months, and in the United States the NBER decides when one starts and ends. A depression is an informal word for a far worse event. The numbers make the difference concrete: twelve post-war recessions cut real GDP by between 0.4% and 9.1% and lasted 2 to 18 months, while from 1929 to 1933 real GDP fell 26.3%, unemployment reached about a quarter of the labor force and prices fell 27%. Even the Great Recession of 2007–09 was about a sixth the size. Use "recession" for the cycle and keep "depression" for the rare collapse.

Check your understanding

4 questions on this lesson. Each answer comes with the reason, and nothing is scored.

  1. Who officially decides that the United States is in a depression?

    1. The NBER Business Cycle Dating Committee
    2. Nobody: "depression" has no official definition
    3. The Federal Reserve
    Show the answer

    B. Nobody: "depression" has no official definition The NBER dates recessions. "Depression" is an informal term for a downturn far deeper and longer than an ordinary recession.

  2. Real GDP fell 4.0% in 2007–09 and 26.3% from 1929 to 1933. About how many times larger was the Depression's fall in output?

    1. About twice
    2. About six to seven times
    3. About twenty times
    Show the answer

    B. About six to seven times 26.3 ÷ 4.0 ≈ 6.6. The worst downturn in seventy years was still a fraction of the Depression.

  3. The 2020 recession cut real GDP by 9.1%, more than any other post-war recession. Why is it still not called a depression?

    1. Because GDP is not used to date recessions
    2. Because it lasted two months, and depressions last years
    3. Because unemployment did not rise
    Show the answer

    B. Because it lasted two months, and depressions last years Depth alone does not make a depression. The 2020 contraction was the shortest on record; output recovered within about a year and a half.

  4. Which feature of the Great Depression made it feed on itself in a way post-war recessions did not?

    1. Falling prices, which raised the real burden of debts fixed in dollars
    2. Rising interest rates on savings accounts
    3. A shortage of workers
    Show the answer

    A. Falling prices, which raised the real burden of debts fixed in dollars Consumer prices fell 27.2% from 1929 to 1933. Debts did not shrink with them, so borrowers defaulted, banks failed and spending fell further.

Sources and further reading

  1. NBER — Business Cycle Dating (definition of a recession and the committee’s criteria)
  2. NBER — US business cycle expansions and contractions (peak and trough dates from 1854)
  3. U.S. Bureau of Economic Analysis — Real gross domestic product, annual from 1929 (FRED GDPCA) and quarterly from 1947 (FRED GDPC1), chained 2017 dollars
  4. U.S. Bureau of Labor Statistics — Labor force, employment and unemployment, 1929–39: estimating methods (Monthly Labor Review, 1948)
  5. U.S. Bureau of Labor Statistics — Civilian unemployment rate (FRED UNRATE) and CPI-U, not seasonally adjusted (FRED CPIAUCNS)