The Robustness of the Split-Trend Stationarity Hypothesis for the U.S

Perron (1989) argued that real gross national product (GNP) has neither a unit root nor a simple trend but is instead stationary around a linear trend that changes slope in 1973. The robustness of Perron's split-trend stationarity (STS) findings is investigated for other measures of economic ac...

Full description

Saved in:
Bibliographic Details
Published in:Southern economic journal 1992-07, Vol.59 (1), p.9-9
Main Authors: Davis, George K, Kanago, Bryce
Format: Article
Language:English
Subjects:
Comparative studies
Economic activity
Economic models
Economic theory
Economics
GNP
Gross National Product
Hypotheses
Quarterly
Stationary
Statistical analysis
U.S.A
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Perron (1989) argued that real gross national product (GNP) has neither a unit root nor a simple trend but is instead stationary around a linear trend that changes slope in 1973. The robustness of Perron's split-trend stationarity (STS) findings is investigated for other measures of economic activity using different methods of determining lag lengths. The breakdown of real GNP is studied based on the type of good produced: manufactured goods, structures, and services. It is found that Perron's result is sensitive to the measure of real economic activity and to the number of lagged differenced residuals. There is no support for the STS hypothesis for employment or construction. Results for industrial production, manufacturing, and real GNP also fail to provide persuasive evidence for the STS hypothesis. However, it is found that services behaved differently than the other measures and may be stationary around a split trend.
ISSN:0038-4038
2325-8012