50 Years of GDP Growth from the IMF World Economic Outlook
Analyze This! is our series of live QuantumLayers demonstrations. We connect a public data source, let the platform chart it and test it, and report what comes back. Every chart on this page is fetched live from QuantumLayers. This time: real GDP growth for eight major economies from the IMF’s World Economic Outlook, 1980 to 2031, including the IMF’s own projections.
The World Economic Outlook is the IMF’s flagship forecast, published twice a year. Its headline series is real GDP growth: the annual change in an economy’s output after adjusting for inflation. The WEO version has a feature most sources lack: it includes the IMF’s projections for the next several years.
The data comes from QuantumLayers’ built-in IMF, Real GDP growth (%) WEO template, which needs no API key and requests indicator NGDP_RPCH for the United States, China, Japan, Germany, India, the United Kingdom, France and Brazil. The IMF serves it in SDMX, a format used by statistical agencies, and QuantumLayers decodes it into a table with a TIME_PERIOD column and one column per country. It recognised the plain years in TIME_PERIOD as dates. We marked the dataset public, which is what lets the charts below load for anyone.
One thing to keep in mind from the start: values from 2026 onward are IMF projections, not history. They are drawn as dashed lines in the charts below.
Fifty years in one chart
Figure 1: Annual real GDP growth for eight economies, fetched live from QuantumLayers. Dashed segments are IMF projections. Click a country in the legend to hide or show it. Source: IMF World Economic Outlook.
▶ Open the live ql-chart on QuantumLayers
Over the whole period, average annual growth was 8.3% for China, 6.1% for India, 2.6% for the US, 2.4% for Brazil, 1.9% for the UK, 1.7% for France, 1.6% for Japan and 1.5% for Germany. Most lines share two deep dips, in 2009 and 2020. In 2009, China and India slowed but kept growing.
China changed regime around 2011
The structural break test (CUSUM) found a break in Chinese growth around 2011. Before it, growth averaged 10.1% a year. After it, 5.4%, a drop of 46%.
Figure 2: China’s annual real GDP growth, fetched live, with the structural break QuantumLayers detected. Dashed: IMF projections.
▶ Open the live ql-chart on QuantumLayers
From 1992 to 2011, Chinese growth averaged 10.6% a year, with peaks of 14.3% in 1992 and 14.2% in 2007 (the highest year in the series was 15.2% in 1984). From 2012 it stepped down to a lower band, from 7.8% in 2012 to 6.1% in 2019, and the IMF projects a continued slide to 3.3% by 2031. A straight trend line through all 52 years would describe China as gently decelerating. The data says something different: two distinct growth regimes, with a clear shift between them.
Japan’s break came twenty years earlier
Japan has its own break, around 1991, the year its asset price bubble burst. Growth averaged 4.3% before it and 0.8% after, a fall of 82%.
Figure 3: Japan’s annual real GDP growth, fetched live, with the structural break QuantumLayers detected. Dashed: IMF projections.
▶ Open the live ql-chart on QuantumLayers
The decade after 1991 is known as Japan’s lost decade, and the data shows it never really ended. Japan has had eight years of negative growth since 1993, including -5.9% in 2009 and -4.3% in 2020, and the IMF projects growth of around 0.6% a year through 2031.
The Western economies move as a group
The correlation tests found strong links between the advanced economies: France and the UK at r = 0.82, the UK and the US at 0.73, Germany and France at 0.72, Germany and Japan at 0.66. Shared recessions drive much of it. 2009 and 2020 were negative for all five, and 2021 was a strong rebound for all of them.
Figure 4: Each point is one year from 1980 to 2031, fetched live, with the trend line QuantumLayers fitted (r = 0.82). Source: IMF World Economic Outlook.
▶ Open the live ql-chart on QuantumLayers
The two most extreme points are 2020 and 2021. The UK had the most extreme swing in the sample, from -10.0% in 2020 to +8.5% in 2021. China does not appear in any of the strong correlations: its growth has followed domestic policy more than the Western business cycle.
What the statistics flagged
A forecast of a forecast. QuantumLayers ran Holt-Winters and SARIMA six years ahead. Because the data already ends with the IMF’s projections to 2031, those forecasts cover 2032 to 2037 and are fitted partly on forecasts. Read them as a demonstration of how the models behave, not as predictions.
Two models, two assumptions. For the US and Germany, SARIMA found the series stationary and pulled growth back toward the long-run average: the US from 1.8% to 2.6%, Germany from 0.6% to 1.5%. For China and Japan, the series needed differencing, so SARIMA extended the recent direction instead: China to 2.6%, Japan to roughly 0%. Holt-Winters, which tracks the trend directly, took China furthest, to 1.8%. Mean reversion suits economies with a stable long-run rate; trend extension suits economies mid-shift.
Ignore the percentages near zero. The headlines described Japan’s projection as a 101% decrease, because 0.6% to roughly 0% is a large relative change of a small number. Read the levels.
A cleaner way to run it. Use the date filter on the insights page to keep years up to 2025, run the forecasts on outturns only, and compare them with the IMF’s projections for the same years. Where the two diverge, the IMF’s judgement departs from what the historical pattern alone would suggest.
How this was built
- Connect. We picked the IMF real GDP growth template in QuantumLayers’ REST API connection form and saved it without changes. QuantumLayers decoded the IMF’s SDMX response and typed the year column as a date.
- Analyze. We ran the automated insights (structural breaks, average levels, correlations, outliers and forecasts) over all eight countries.
- Publish. With the dataset public, each chart here is drawn with Chart.js from an unauthenticated request to QuantumLayers’
ql_get_chart_dataendpoint, with the IMF’s projection years dashed. The purple buttons open the same charts on QuantumLayers’ hostedql-chartpage.
The other IMF templates (inflation, unemployment, government debt, current account and GDP per capita) use the same URL pattern. To add a country, add its ISO3 code, such as KOR or MEX, to the +-separated list in the URL. When the IMF publishes a new edition, the connection picks it up on its next sync.
Data: IMF World Economic Outlook, the edition available when we connected in November 2026. Values from 2026 onward are IMF projections. Figures reflect the dataset at request time.
This post is part of Analyze This!, a series of live QuantumLayers demonstrations. For the economic sources you can connect the same way, see Economic Data in QuantumLayers: FRED, World Bank and IMF, and for how the forecasts work, How QuantumLayers Forecasts Time Series. Start building at quantumlayers.com.
