equipment) for depreciation. Al- 

 though data are available to make 

 some estimates of Japanese R&D on 

 a current operating cost basis, these 

 estimates would still not be pre- 

 cisely comparable to the various U.S. 

 data series and in some instances of- 

 fer no advantage over disbursement 

 expenditure data. For consistency, 

 all Japanese data are on a disburse- 

 ment basis, with the understanding 

 that disbursement expenditure data 

 may be slightly higher than oper- 

 ating cost data.^ 



In terms of fields of coverage, both 

 Japan and the United States include 

 research in the natural sciences, en- 

 gineering, and the social sciences. 

 However, Japan, like many of the 

 European countries, also includes in 

 its R&D figures research activity in 

 the humanities and education. The 

 Japanese data in this report have not 

 been adjusted to exclude the hu- 

 manities and education, which con- 

 stituted 6 percent of total Japanese 

 R&D expenditures in 1985, 27 per- 

 cent of higher education expendi- 

 tures (this is discussed in more detail 

 in chapter 4, "Higher Education"), 

 and 4 percent of government and 

 nonprofit expenditures. 



Japanese expenditure data for all 

 R&D performers are reported for the 

 fiscal year April 1 to March 31; U.S. 

 data are reported primarily on a cal- 

 endar year basis (industries and uni- 

 versities are allowed to report on their 

 own fiscal year basis, which may or 

 may not coincide with the calendar 

 year). However, any differences be- 

 tween the Japanese and U.S. data 

 series which result from noncoinci- 

 dent periods of time are certainly 

 slight, and do not affect trends over 

 time. 



personnel 



The international convention for 

 reporting R&D personnel data is to 

 include only that portion of a re- 

 searcher's time actually devoted to 

 research. One of the more problem- 

 atic differences between the Japa- 

 nese and U.S. data is that the U.S. 

 numbers of scientists and engineers 

 engaged in R&D are reported on a 

 full-time-equivalent (FTE) basis; cor- 

 responding Japanese data report the 

 gross total of scientists and engi- 

 neers engaged in R&D. While this is 

 generally not a problem when com- 

 paring scientists and engineers in the 

 industrial and government sectors 

 (since they are presumably working 

 full-time on R&D), it does become a 

 problem for the higher education 

 sector where a researcher's time is 

 divided among research, teaching, 

 and administrative responsibilities. 



To compensate somewhat for this 

 difference, this report excludes Jap- 

 anese scientists and engineers en- 

 gaged in R&D in the social sciences 

 and humanities. This adjustment re- 

 sults in a 21-percent understatement 

 of the total number of Japanese sci- 

 entists and engineers engaged in 

 R&D and a 40-percent understate- 

 ment for the university sector.^ 



-An examination of Japanese industry data for the 

 period 1980-85 sfiowed tfiat for all industries com- 

 bined, R&D expenditures based on operating costs 

 were consistently about 7 percent less than dis- 

 bursement expenditures. 



'This method for estimating the number of FTE 

 scientists and engineers engaged in R&D yields sim- 

 ilar results to those obtained by a Japanese study. 

 This study surveyed researchers affiliated with uni- 

 versities and derived an FTE coefficient of 0.6 for 

 teaching staff. See Inshtute for Future Technology, 

 Study on the Situation of Statistics Related to Research 

 and Development Activity in the United States and Eu- 

 rope, and a Comparison with the Situation in japan (To- 

 kyo, Japan, March 1987). 



currency 

 conversions 



This report expresses most mon- 

 etary figures in terms of constant 1982 

 dollars while Japanese official doc- 

 uments publish R&D funding and 

 expenditure data in current yen. For 

 this report, current yen have been 

 converted into current U.S. dollars 

 using OECD purchasing power par- 

 ities (PPPs), and then converted into 

 constant 1982 dollars using the U.S. 

 Department of Commerce gross na- 

 tional product (GNP) implicit price 

 deflators. 



Market exchange rates have not 

 been used for the currency conver- 

 sion because they have become in- 

 creasingly unsatisfactory in recent 

 years for international statistical 

 comparisons. Given the volatility of 

 contemporary international capital 

 markets, the international value of 

 a currency is not a consistent reflec- 

 tion of the price structure for goods 

 and services within a country, and 

 consequently does not fully reflect 

 the currency's purchasing power. 

 Quite simply, PPPs for Japan are the 

 number of yen needed to buy the 

 same quantity of goods and services 

 in Japan as a dollar can buy in the 

 United States."* In this manner, Jap- 

 anese yen are converted into dollar 

 terms representing the same pur- 

 chasing power as a U.S. dollar in the 

 United States. 



While PPPs provide more com- 

 parable data by equalizing how much 

 a respective currency can buy, they 

 are based on a broad index (or bas- 

 ket) of consumer goods and services 



'The OECD has calculated PPPs for its member 

 countries; all PPPs are expressed in terms of units 

 of foreign currency per US. dollar. 



46 



