WASHINGTON (Reuters) - U.S. government statisticians say they have addressed a methodology problem, which analysts had argued tended to understate economic growth in the first quarter.
The Bureau of Economic Analysis, the government agency that constructs gross domestic product data, said this week a three-phase plan to resolve residual seasonality was complete with Friday’s release of a comprehensive revision of GDP numbers.
“We have worked quite closely with our partners around this,” Erich Strassner, associate director for National Economic Accounts at the BEA, told reporters. “Through our tests we do not show any signs of residual seasonality in our estimates.”
The BEA announced in 2016 that it was undertaking a component-by-component review of data series included in GDP to look for possible residual seasonality. The agency applied seasonal adjustment improvements to the GDP series going back to 1929 for the annual data and 1947 for the quarterly figures.
The government adjusts economic data to remove fluctuations such as seasonal weather patterns and holidays that normally occur at roughly the same time and magnitude every year, to make the series easier to interpret and analyze.
To make the seasonal adjustment process more transparent, the BEA is now publishing not seasonally adjusted GDP data, with every release of the quarterly reports. The minimum period for seasonal adjustment was raised to five years from three years.
“We know we are not done in the sense that moving forward with every annual update we are going to be looking at the most recent 15-year period to test for residual seasonality as well as the whole time period of statistics,” said Strassner.
Economists believed residual seasonality had been most prevalent in first-quarter GDP data. The BEA sharply revised up the 2016 first-quarter GDP growth estimate to a 1.5 percent annual rate from the previously reported 0.6 percent pace. It also raised the 2017 first-quarter GDP growth estimate to a 1.8 percent pace from a 1.2 percent rate.
“Naturally improved seasonal adjustment plays a role in these revisions,” said Strassner. “However we cannot separately identify the portion that can be attributed to improved seasonal adjustment from the incorporation of newly available source data.”
The government also incorporated new data from the 2012 census. Over the 2012-2017 period, GDP growth for the second quarters was on average revised up by one-tenth of a percentage point, and 0.3 percentage points for the third quarters.
Growth for the fourth quarters was revised up by two-tenths of a percentage point over that period. Overall, the benchmark revisions showed little change in the economic picture from what was previously reported.
But the saving rate was far much higher than previously estimated between 2010 and 2017. The 2017 saving rate was raised to 6.7 percent from 3.4 percent. The BEA said the revisions accounted for misreporting of proprietor income. There was little change to previously published inflation estimates.
The government also made improvements to measures of price changes for software, medical and communication equipment. Other measures taken include changes in definitions to more accurately reflect the changing the U.S. economy and to provide consistent comparisons with data for other countries.
Editing by Chizu Nomiyama