Earlier this year we greatly expanded the functionality of the US Treasuries & Economic Indicators Time Series page, allowing users to compare interest rates against other economic data such as GDP and inflation metrics. It is important to note that such economic data released by the US government is frequently revised. This means that the historical data you are looking at today could be very different from how it was originally reported. On the Time Series page, we provide you the option to view the values as their current official datasets or as the values were reported when the data releases were first made public.
Some metrics get revised more radically than others. GDP is notorious for undergoing numerous revisions that could result in drastic differences from the initial report over time. This is why recessions are often only realized by the government reporting agencies well after the fact. For example, in the figure above, real GDP growth for the Fourth Quarter of 2021 was 6.9% in the "Advance" estimate of the Bureau of Economic Analysis, reported January 27, 2022. On September 26, 2024, a "comprehensive revision" by the BEA changed this value to 7.9%.
Comprehensive revisions occur regularly, in which case the BEA may make changes to the last 5 years of reported GDP figures. Comprehensive revisions can even overwrite what that BEA calls "Final" GDP estimates. The Q4:2021 "Final" estimate, provided in a press release dated March 30, 2022, stated the real GDP growth to be 6.9%.
On our new page titled Economic Data Revision History provides a table of economic data revisions, allowing you to see the number of changes to a given data set over time. At the time of its release, you can view the history of revisions to CPI, PPI, real GDP growth, and the unemployment rate. More datasets will be available soon!
Why many GDP revisions are labeled on 6/17/2024
When viewing the GDP economic revisions, a large amount of the data has a "Revision Date" of 6/17/2024. That is the date which this website began aggregating the real GDP growth metrics from BEA data. We felt it appropriate to label those data points with that particular date because we know that's what the figures were reported by the BEA as of that date. We do not have information as to the originally reported values of those data points and their revision history. All we can verify is how the BEA reported them on 6/17/2024.
Most of the other GDP revision dates and originally reported values have been obtained from the BEA press release archive. Unfortunately, this archive only has press releases going back to 1994. We do not have a history of revisions for GDP growth prior to that.
CPI y/y revisions
The "seasonally adjusted" CPI, which is used for month-to-month comparisons, is subject to revision for up to 5 years. The "not seasonally adjusted" CPI figures are revised infrequently, and usually only occur when the Bureau of Labor and Statistics changes calculation methodology. Some significant revisions are noted around these years:
1919: The BLS began publishing separate consumer price indexes for 32 cities. This was the initial attempt to systematically measure changes in consumer prices across different urban areas.
1940: The first major revision of the CPI was implemented, which included a shift to a more comprehensive sample of goods and services and updated weights based on the 1934-36 Consumer Expenditure Survey.
1953: Another significant revision introduced a new base period (1947-49) and expanded the geographic coverage of the index.
1964: The CPI was revised to use the 1960-61 Consumer Expenditure Survey for weights, and the base period was updated to 1957-59.
1978: This revision introduced a new base period (1972-74) and incorporated changes in the sampling and pricing methods. It also marked the beginning of the CPI for All Urban Consumers (CPI-U) and the CPI for Urban Wage Earners and Clerical Workers (CPI-W).
1988: The base period was updated to 1982-84, and the methodology was further refined to improve accuracy.
2000: Starting in 1998, BLS began using a geometric mean formula for most basic indexes that mitigates lower level substitution bias and reflects shifts in consumer spending with item categories as relative price change
