The Result of the closer examination of the "Business and Investment Survey of Incorporated Enterprises" Used in Estimation of Private Non-Residential Investment in the Second Preliminary GDP Figures for October-December 1999, and Related Future Measures

May 2000
Economic Planning Agency
Economic Research Institute

1. As one of the primary statistics to estimate figures for private non-residential investment in the first preliminary GDP figures, "projected amounts" of investments by financial and insurance corporations derived from the "Business and Investment Survey of Incorporated Enterprises (Research Bureau of the Economic Planning Agency)" are used. "Actual amounts" of investments by financial and insurance corporations derived from the above survey are available in time for the second preliminary GDP figures. Therefore, "projected amounts" are usually replaced by "actual amounts" in computing the second preliminary GDP figures.


2. A review of investments by financial and insurance corporations during October-December 1999 as found in the above survey shows that the correction between the "projected" and the "actual amounts" was extraordinary larger than ever before. It was therefore judged that closer examination was needed prior to use the "actual amounts" in computations. As such, it was decided to temporarily postpone the replacement of the "projected amounts" with the "actual amounts" in the computation of the second preliminary GDP figures.


Note: Scale of correction (October-December 1999):

(the "actual amounts" - the "projected amounts") / the "projected amounts" = - 36%


3. Whenever the amount of investment is not reported in a response to the questionnaire of the above survey, it is assumed that the respondent has made no investment during the period in aggregating the data. The closer examination of aggregated data was deliberately implemented with the attention to this treatment.


4. As a result, it was found that there was no significant difference from the "actual amounts" if aggregations were carried out only for those corporations reporting amounts of their investments.


5. In view of this result, it was decided that the figures for private non-residential investment for October-December 1999 would be re-computed based on the usual method of using the "actual amounts" by financial and insurance corporations during October-December 1999 as derived from the above survey. The revised figures shall be announced at the same time as the first preliminary GDP figures for January-March 2000. As for private non-residential investment for January-March 2000, the usual method shall be followed whereby "projected amounts" of investments by financial and insurance corporations in the above survey are used in the computation of the first preliminary GDP figures.


Note: Investments by financial and insurance corporations in the "Business and Investment Survey of Incorporated Enterprises".

October-December 1999: Year-over-year change in "actual amounts": - 38%

January-March 2000: Year-over-year change in "projected amounts": - 34%



(Reference)

What impact would the use of the "actual amounts" of investments by financial and insurance corporations in the above survey have on real GDP for October-December 1999?


1. Regarding GDP figures for October-December 1999, re-computations will be carried out and the results will be announced at the same time as the first preliminary GDP figures for January-March 2000. Before the time, it is not possible to accurately revise the figures that will result.


2. However, rough estimates indicate the following possibilities:

[1] Impact on quarter-to-quarter real GDP change for October-December 1999 : approx. - 0.2%

[2] Impact on year-over-year real GDP change for fiscal 1999 : approx. - 0.05%