Discussant: Paul Van den Bergh

Abstract

Efforts to Harmonise Statistical Methodologies and Best PracticesThe paper by Carson illustrates how the Fund is trying to push the statistical envelope to meet the challenges of globalisation. It focuses on the Fund's efforts to improve the statistical coverage of FDI, public finance and financial soundness indicators. With respect to the latter indicators, the BIS has been participated actively in the preparation of the Compilation Guide and the promotion of the indicators in IMF seminars around the world. Efforts were made to ensure that the methodology used for the indicators, particularly those for the banking sector, were consistent with definitions of the Basel Committee on Banking Supervision as well as with the BIS methodology for its international banking statistics (e.g. approaches to consolidation). The BIS was also very supportive of the proposal to include statistics on real estate prices in the set of indicators and, together with the Fund, sponsored a conference on real estate prices and financial stability in October 2003. I am therefore pleased to note that the Australian Bureau of Statistics is making efforts to improve the timeliness and quality of house price statistics.The three initiatives mentioned by Carson come on top of numerous other IMF initiatives such as the development of manuals for statistics on balance-of-payments and external debt. The Fund undertakes numerous steps to promote the adoption of these methodologies amongst its constituency, for instance through training and technical assistance. The development of the data dissemination systems (SDDS/GDDS) and the inclusion of the key statistical methodologies in the Reports on Standards and Codes are also putting pressure on national agencies to adjust their national methodologies to the recognised international standards.Various comments can be, and have been, made on these worthwhile efforts, by national compilers as well as users. One positive comment is that the Fund consults widely with statistical experts in national, regional and international statistical agencies as well as in the academic community and, where appropriate, in the private sector in developing its methodologies. One criticism, perhaps too easy, is that the efforts made to date have yet to contribute to fully satisfactory international comparability of key economic and financial data. Where full harmonisation with international best practice is not possible, statistical agencies have to be convinced, at a minimum, to describe through appropriate metadata how their data deviate from international methodologies and best practices. One hope is that the e-standards for statistical data and metadata exchange (SDMX) which the IMF is developing in cooperation with other international organisations, including the BIS, will facilitate the dissemination of such metadata. Better disclosure of deviations from international methodologies might exert pressure on statistical agencies to reduce such methodological "gaps".In terms of statistical best practices, a laudable in...