About Foreign Trade |
Schedule B |
STATISTICS: Press Releases | Trade Highlights | Country/Product Data | State Export Data | Historical Series | Special Request
Questions Concerning Change to Chained-Dollar (Fisher) Methodology
Q1: Why is Census adopting the Fisher chained methodology for its constant dollar (real) series?
A1: Census is replacing its fixed-weighted constant dollar series with a new chained series to improve the quality of the data series and increase the consistency of the real import and export statistics produced by the Department of Commerce. The new data will be based on the same chained methodology used by the Bureau of Economic Analysis in the National Income and Product Accounts (NIPA). The chained methodology produces more accurate indexes during periods of large price changes by eliminating the substitution bias (which occurs when less expensive goods are substituted for more expensive ones, for example, chicken in lieu of beef or margarine instead of butter).
A2: No, the trends will be similar, but the data will not match for at least two reasons. First, there are coverage and definitional differences between the current dollar NIPA and Census based trade data. Second, for the monthly FT-900, Census will use monthly chaining. The NIPA are quarterly and use quarterly chaining.
Census basis data reflect the movement of goods between foreign countries
and customs territory of the United States, which includes the 50 states,
the District of Columbia, Puerto Rico, the U.S. Virgin Islands, and
all U.S. foreign trade zones. They do not include monetary gold or shipments
between the United States and Puerto Rico or the U.S. Virgin Islands.
The NIPA include the sale of monetary gold held at the Federal Reserve
Bank, and shipments between the United States and Puerto Rico or the
U.S. Virgin Islands. However, shipments between Puerto Rico or the U.S.
Virgin Islands and foreign countries are not included in the NIPA. These
are the most significant differences between Census basis goods data
and BEA's NIPA import and export data. More information about Census
basis trade data can be found at: http://www.census.gov/foreign-trade/www/press.html
A3: The current methodology compares the current period's trade with that in the base year (currently 1996). With the new methodology, the chained dollar data are constructed by comparing each month to the previous month, weighting those changes by the importance of each good, and chaining back to the base year. Thus, changes in the goods traded are incorporated very quickly into the deflators. In addition, when the chained dollar series are rebased, the period-to-period changes are maintained.
One obvious difference is that the chained dollar series are not additive. For example, the sum of the six component import series in Exhibit 10 will not sum to total imports. The difference between the sum of the categories and the total will vary from month-to-month and between the two real import and export exhibits, (Exhibits 10 and 11).
As stated in A1, the chained methodology produces more accurate indexes during periods of large price changes by eliminating the substitution bias. The fixed-weighted series tends to misstate growth in periods farther from the base period.
Q4: Does Census plan to continue producing the fixed-weighted constant dollar data?
A4: No, based on BEA's experience, the new methodology is more accurate. Census plans to discontinue the fixed-weighted constant dollar series as of April 2003 statistics (June 2003).
A5: Yes, by the time the new data series is released in June 2003, Census will have chained dollar data available back to January 1994.
A6: Census plans to make 3 years of chained dollar data available one
week after the release of January 2003 statistics. These data will have
a 1996 base year comparable to the published fixed-weighted data for
the same period. The data will be available on our Web site and users
will be able to connect to the data through the FTD Web News Box. Data
users are cautioned, however, that these data are for comparison purposes
only. The new chained dollar series will have a 2000 base year.
Q7: What product categories will be available with the new chained methodology? What will the press release tables look like? Will there be any significant differences in content or appearance?
A7: There will be no change in what product categories are available in the monthly FT900 (Exhibits 10 and 11). The 1-digit end-use categories, petroleum, and non-petroleum will be available for the new chained dollar series.
The formats of Exhibits 10 and 11 in the monthly FT900 will change very little. The major difference will be the addition of a residual column to both exhibits. The residual in Exhibits 10 and 11 will be the difference between the sum of the categories and total imports or exports.
A8: BEA is planning to change its base year to 2000 late in 2003. To avoid the need to rebase so soon after adopting the new methodology, as well as to minimize the disruption to data users, Census is electing to change base years at the same time we introduce the new chained dollar series.
A9: We will continue to revise the previous month's data every month and to revise the current and previous quarters at the end of each quarter. Also, we will continue to revise the most recent three years in our annual revision report that is released with the April FT900 each year.
Q10: Where can I get more information concerning the methods used to calculate the chained indexes?
A10: BEA's Web site www.bea.gov has many articles. One article of particular interest is in May 1997's Survey of Current Business, BEA's Chain Indexes, Time Series, and Measures of Long-Term Economic Growth by J. Steven Landefeld and Robert P. Parker www.bea.doc.gov/bea/articles/national/nipa/1977/0597od.pdf.
Source: FTDWebMaster, Foreign
Trade Division, U.S. Census Bureau, Washington, D.C. 20233
Location: MAIN: STATISTICS:PRESS RELEASES
Created: 30 July 2002
Last modified: 12 July 2011 at 04:01:20 PM