Benchmark Reconciliations Revisited Bettina Aten and Alan Heston1 ...
20
pages
English
Documents
Cet ouvrage est temporairement indisponible
20
pages
English
Documents
Cet ouvrage est temporairement indisponible
1
Benchmark Reconciliations Revisited
1
Bettina Aten and Alan Heston
Summary ; Producers of benchmark comparisons are naturally interested in whether the
results from one benchmark are consistent with a later benchmark using growth rates
between the two to produce a comparison in the later year. Our interest is somewhat
different. We recognize that there are differences in price samples, weighting and other
factors that will always lead to some differences between extrapolated and new
benchmark estimates. In the spirit of some strands of research on the statistical
discrepancy in national accounts, we focus on the question of whether there is anything
to be learned about the world from examining such differences. The technique we
employ looks at several pairs of OECD benchmark comparisons for 1990s, first 1990 - 96
as officially reported and then less intensively at other benchmarks in the 1990s using a
common aggregation procedure. The approach is to examine whether there is strong
autocorrelation among the residuals of the countries. We then examine some possible
influences that might reduce this auto - correlation, including price similarity, distance and
a measure of real exchange rates between each pair of countries. The main conclusion
of the paper i s that our measure of real exchange rate movements has a strong
explanatory effect. While this is a suggestive result it would require more research to
establish the nature of any ...
Voir
Benchmark Reconciliations Revisited
1
Bettina Aten and Alan Heston
Summary ; Producers of benchmark comparisons are naturally interested in whether the
results from one benchmark are consistent with a later benchmark using growth rates
between the two to produce a comparison in the later year. Our interest is somewhat
different. We recognize that there are differences in price samples, weighting and other
factors that will always lead to some differences between extrapolated and new
benchmark estimates. In the spirit of some strands of research on the statistical
discrepancy in national accounts, we focus on the question of whether there is anything
to be learned about the world from examining such differences. The technique we
employ looks at several pairs of OECD benchmark comparisons for 1990s, first 1990 - 96
as officially reported and then less intensively at other benchmarks in the 1990s using a
common aggregation procedure. The approach is to examine whether there is strong
autocorrelation among the residuals of the countries. We then examine some possible
influences that might reduce this auto - correlation, including price similarity, distance and
a measure of real exchange rates between each pair of countries. The main conclusion
of the paper i s that our measure of real exchange rate movements has a strong
explanatory effect. While this is a suggestive result it would require more research to
establish the nature of any ...
Publié par
Langue
English
