Rapid Estimates of U.S. GDP: Timeliness, Estimating Methods & Accuracy Supplemental Handout

advertisement
Rapid Estimates of U.S. GDP:
Timeliness, Estimating Methods & Accuracy
Supplemental Handout
Dave Wasshausen
International Seminar on Timeliness, Methodology and
Comparability of Rapid Estimates of Economic Trends
Ottawa, Canada
www.bea.gov
May 27-29, 2009
Estimating Methods: PCE
▪
Retail Control Method:
 Subset of monthly retail sales used as extrapolator for most PCE goods
▪
Price-Times-Quantity Method:
 Used when accurate direct information is available on quantities and prices, or
when consistent data on sales are not available
▪
Judgmental Trend:
 Primarily used to extrapolate services -- where monthly sales and quantity
data are not available:
 Heating and cooling degree days
 Monthly payroll data
 Population growth, changes in CPI for relevant category; and adjustment factor.
▪
Imputations:
 Used where the service cannot be directly observed:
 Rental value of owner-occupied housing & banking services.
 Important for capturing all economic activity and for maintaining
measurement that is consistent under different institutional arrangements.
 Accounts for about 1/8 of PCE.
www.bea.gov
2
Estimating Methods: Investment
▪
Commodity Flow Method:
Shipments
+ Transportation costs, trade margins, sales taxes, and imports
- Exports, changes in inventories, and sales to consumers and government
= Private fixed investment in equipment
▪
Price-Times-Quantity Method:


Used when accurate direct information is available on quantities and prices, or when
consistent data on sales are not available
Examples include petroleum and natural gas structures & motor vehicles
▪
Inventories: Adjusted to a replacement cost basis
▪
Structures: Value-Put-In Place data (plus adjustments for brokers commissions)
▪
Judgmental Trend:


www.bea.gov
Primarily used for missing third month (structures, foreign trade, most inventories)
Foreign trade assumptions in commodity flow can be a significant source of revision to
investment but not GDP because of offsetting revision in net exports.
3
Estimating Methods: Net Exports
▪ Monthly foreign trade reports from the Census Bureau
are a virtual monthly census
 Adjustments for commercial gold, petroleum that moves through
pipelines, etc.
 Trend extrapolation and partial extrapolators (aircraft and
petroleum) for missing third month.
 Large source of revision; usually offset by inventory revisions.
▪ Trends:
 Trend extrapolation and partial extrapolators (aircraft and
petroleum) for missing third month for goods;
 Trend extrapolation and partial extrapolators for trade in
services:
 Examples include monthly Treasury data on financial
transactions, monthly commercial data on international
passenger travel;
 Later replaced with quarterly international services survey data
www.bea.gov
4
Estimating Methods: Government
▪ Federal spending extrapolated using Monthly
Treasury Statement:
 Adjustments for transfers to derive final expenditures
(which account for only about 1/3 of Federal Spending).
 Timing adjustments
▪ State & Local spending extrapolated using a
variety of trend and indicator data, including:
 Monthly “Value-put-place” data by type of structure
 Monthly S&L payrolls.
▪ Quarterly depreciation estimated from trend
www.bea.gov
5
Download