Real vs nominal GDP: deflators, chained dollars and why both matter

Nominal GDP is the value of an economy's output at the prices of the time. Real GDP is the same output with the effect of changing prices taken out. The first answers "how many dollars changed hands for final goods and services?"; the second answers "how much more stuff did the economy actually produce?"

The difference matters whenever prices are moving. In the second quarter of 2022, US nominal GDP was 10.3% higher than a year earlier, a pace that would look like a boom. Real GDP was up only 2.3%. The other eight points were inflation. Knowing which number you are looking at, and how the conversion between them works, is the difference between reading the economy and misreading it.

This guide explains how the Bureau of Economic Analysis (BEA) turns nominal GDP into real GDP, what the GDP price deflator is, why real GDP is quoted in "chained 2017 dollars" and why those dollars don't add up, and when nominal GDP is the more useful number. For the basics of what GDP measures and how it is released, start with GDP explained: how gross domestic product is measured and read.

Two measures of the same output

Imagine an economy that produces only pizzas and laptops.

Real output grew roughly 18.5% while nominal output grew 8.7%, so the overall price level of this economy fell: the falling laptop price outweighed the rising pizza price. That is the first lesson: nominal and real growth can differ in either direction, and the gap is the change in prices.

The second lesson is that the answer depends slightly on which year's prices you use: 18.6% or 18.4%. The choice matters more the further apart the two periods are and the faster relative prices change.

The deflator: from nominal to real

The ratio of the two measures is a price index for everything GDP covers.

In 2017 the two measures are equal by construction: average nominal and average real GDP were both $19,612.1 billion. Before 2017 real GDP is above nominal (prices were lower then, so valuing old output at 2017 prices raises it); after 2017 it is below.

BEA publishes two closely related price measures for GDP. The implicit price deflator is the ratio above. The GDP price index is calculated directly with the same chain-type formula as real GDP. They are almost identical in practice.

What the deflator covers, and how it differs from CPI

The GDP deflator prices everything produced in the US: consumer goods, business equipment, new homes and government services. It excludes imports, because imports are not domestic production. The Consumer Price Index (CPI) prices a basket that urban consumers buy, imports included, and nothing businesses or governments buy. So an oil-price jump hits CPI quickly through gasoline but reaches the deflator only through domestic oil production. Over the year to Q3 2026 the GDP price index rose 3.7%; CPI was up 3.7% in the year to September 2026. Those numbers often differ by half a point or more.

Chained dollars: why real GDP isn't in "2017 prices"

The obvious way to compute real GDP is the one in the pizza example: pick a base year and value every year's output at that year's prices. The US did this until the mid-1990s. The problem is the one the example hinted at. Goods whose prices fall fastest, such as computers, tend to be the ones whose output grows fastest. Valuing today's huge volume of computers at the high prices of a distant base year exaggerates growth, and the exaggeration gets worse the further you move from the base year.

BEA's solution is to chain short comparisons together:

1. For each pair of adjacent periods, measure the growth in quantities using the prices of both periods, combined in a Fisher index: the geometric mean of the two answers.
2. Multiply ("chain") these period-to-period growth rates together to build a quantity index over any span of time.
3. To express the index in dollars, multiply it by nominal GDP in the reference year (currently 2017) and divide by 100. The result is GDP in chained 2017 dollars.

Because the weights are updated every period, chained real growth rates are not distorted by a stale base year. The price is a loss of simple arithmetic.

Chained dollars don't add up

BEA states it directly: chained-dollar values of components are not additive, because the relative prices of each period differ from those of the reference year. Real consumption plus real investment plus real government plus real net exports does not equal real GDP, and the gap grows the further you are from 2017.

Two practical rules follow. Don't compute a component's "share" of GDP from chained dollars; use nominal shares instead (in Q3 2026, consumption was 67.6% of nominal GDP). And don't add chained-dollar components to get contributions to growth; BEA publishes its own contributions to percent change tables that do it correctly.

Real and nominal growth over time

The two lines move together, separated by inflation. A few points from the data:

• 2022: the widest gap since the early 1980s. In Q2 2022 nominal GDP was up 10.3% on a year earlier, real GDP 2.3%, and the deflator 7.8%, its fastest four-quarter rise since 1981.
• Q3 2026: nominal +5.9%, real +2.1%, deflator +3.7%.
• 2009: in a deep recession with little inflation the lines nearly meet. In Q2 2009 nominal GDP was down 3.3% on the year and real GDP down 4.0%.
• 2020: in Q2 2020 nominal fell 6.7% and real 7.4%.

Before 2000 the gap was sometimes wider still. In Q4 1974 nominal GDP was 8.4% higher than a year earlier while real GDP was 1.9% lower; the deflator rose 10.5%. In Q2 1980, nominal GDP grew 8.0% while real GDP fell 0.8%. In both cases a reader watching only dollar totals would have missed a recession. The deflator's peak in the series was 10.9% in Q1 1975. Periods like these, with high inflation and weak output, are the subject of Stagflation: high inflation, weak growth and the 1970s lesson.

Over the long run the difference compounds. From Q1 2000 to Q3 2026 nominal GDP grew 3.3 times, real GDP 1.77 times; the rest is the rise in prices.

When to use which

Use real GDP when the question is about production: is the economy growing, is it in recession, how fast is productivity rising, how does this expansion compare with the last one? Recession dating, the business cycle and "GDP growth" in the news all mean real.

Use nominal GDP when the question is about dollars:

• Debt and deficits. Debt-to-GDP divides a nominal dollar debt by nominal GDP. Inflation raises nominal GDP and so can lower the ratio even with no real growth; see Government debt and deficits: how US federal borrowing works.
• Tax revenue and company sales. Both are collected and reported in current dollars. An S&P 500 company's revenue tracks nominal spending, which is one reason profits can rise strongly during inflation. The link is explored in How the economy affects stocks: growth, inflation, jobs and rates.
• Interest rates. Long-term bond yields are often compared with nominal GDP growth, the dollar growth rate of the economy's income. Real yields are compared with real growth; see Real interest rates: nominal rates minus inflation, explained.

Pitfalls

• Mixing measures. Comparing this year's nominal growth with last year's real growth, or a nominal series (retail sales, see Retail sales report: what it measures and how to read it) with real GDP, mixes inflation into the comparison.
• Adding chained dollars. Shares and sums built from chained-dollar components are wrong, sometimes by a lot for components with fast-changing prices.
• Treating the deflator as consumer inflation. It covers investment and government and excludes imports. For what households pay, use CPI or the PCE price index (PCE inflation: the Fed’s preferred gauge vs the CPI).
• Reference-year changes. When BEA moves the reference year in a comprehensive update, every chained-dollar level changes. Growth rates barely do. Compare growth rates, not levels, across vintages.
• Quality change. Real GDP is only as good as the price indexes behind it. Measuring the price of a better product (a faster laptop at the same price) is hard, and economists debate how well it is done.

Real, nominal and price measures compared

How Finabase does it

Finabase does not publish GDP or deflator data. The numbers, tables and the chart on this page are ours, built from BEA data distributed through the St. Louis Fed's FRED database and frozen on October 30, 2026.

The app's market data are all nominal: quotes, returns and charts are in current dollars and not adjusted for inflation. The Treasury page lists nominal Treasury yields from one month to 30 years; subtracting expected inflation from those gives the real rates discussed in Real interest rates: nominal rates minus inflation, explained. The Discover page's Economy topic collects web articles about the economy, including GDP releases.

Frequently asked questions

What is the difference between real and nominal GDP?

Nominal GDP values output at the prices of the period; real GDP removes the effect of price changes so periods can be compared. In Q3 2026 US nominal GDP was $33.0 trillion at an annual rate and real GDP $24.6 trillion in chained 2017 dollars.

Which is better, real or nominal GDP?

Neither; they answer different questions. Real GDP tells you whether the economy produced more; nominal GDP tells you how many dollars of income and spending there were, which is what debts, taxes and revenues are measured against.

What does "chained 2017 dollars" mean?

It means real GDP is measured with a chain-type index that updates price weights every period, then scaled so that it equals nominal GDP in the reference year, 2017. The values look like dollars but are an index in dollar units.

Can nominal GDP rise while real GDP falls?

Yes, when prices rise faster than output falls. In Q4 1974 nominal GDP was 8.4% higher than a year earlier while real GDP was 1.9% lower.

What is the GDP deflator?

It is nominal GDP divided by real GDP, times 100: a price index for everything produced in the US. It stood at 134.1 in Q3 2026, against 100 in 2017.

Summary

Nominal GDP counts output at current prices; real GDP strips out price changes, and the ratio between them, the GDP deflator, is a price index for all domestic output. BEA measures real growth with a chain-type Fisher index that updates its weights every period, which avoids the distortions of a fixed base year but means chained-dollar components no longer add up to the total. The gap between the two growth rates is inflation: 10.3% nominal against 2.3% real in Q2 2022, 5.6% against 3.4% annualized in Q3 2026. Use real GDP to judge the economy's output and nominal GDP for anything measured in dollars.