Book Chapter
Current Uses of the RAS Technique: A Critical Review

Many analysts are using nonsurvey techniques to estimate national and regional input-output tables, interregional trade flows, and other types of economic data. Analysts’ extensive use of these techniques reflects the delicate compromise they must make between the cost of collecting actual data and the accuracy of the models with which they work. For adjusting and/or updating input-output tables, analysts most widely use the RAS procedure. Based upon the structure of a national or a different, but supposedly similarly structured, regional table, they employ the technique both to update national (regional) tables and to estimate regional tables. They use a base national (regional) table, A, and separately estimate marginal row (r) and columnand column (s) controls for the predicted year. They then iteratively adjust the flows in A first to sum to the respective rs and then to the respective ss until the new row and column totals in the new matrix are as close as designated to marginal (actual) totals.

I sincerely thank Drs Malte Möhr and William H. Crown, with whom I discussed these views in the early 1980s, and Dr Xiannuan Lin, who gave me very helpful comments on this paper. Polenske, Crown, and Möhr (1986) and Möhr, Crown, and Polenske (1987) contain supporting information concerning the important issues I review here.

Title
Publication TypeBook Chapter
Year of Publication1997
AuthorsPolenske KR
Book TitlePrices, Growth and Cycles
Pagination58-88
PublisherPalgrave Macmillan
CityLondon, UK
ISBN978-1-349-25277-0
Abstract

Many analysts are using nonsurvey techniques to estimate national and regional input-output tables, interregional trade flows, and other types of economic data. Analysts’ extensive use of these techniques reflects the delicate compromise they must make between the cost of collecting actual data and the accuracy of the models with which they work. For adjusting and/or updating input-output tables, analysts most widely use the RAS procedure. Based upon the structure of a national or a different, but supposedly similarly structured, regional table, they employ the technique both to update national (regional) tables and to estimate regional tables. They use a base national (regional) table, A, and separately estimate marginal row (r) and columnand column (s) controls for the predicted year. They then iteratively adjust the flows in A first to sum to the respective rs and then to the respective ss until the new row and column totals in the new matrix are as close as designated to marginal (actual) totals.

I sincerely thank Drs Malte Möhr and William H. Crown, with whom I discussed these views in the early 1980s, and Dr Xiannuan Lin, who gave me very helpful comments on this paper. Polenske, Crown, and Möhr (1986) and Möhr, Crown, and Polenske (1987) contain supporting information concerning the important issues I review here.

URLhttps://link.springer.com/chapter/10.1007/978-1-349-25275-6_4
DOI10.1007/978-1-349-25275-6_4