Gross Domestic Product
This section describes the value added of industry and services. For 1890-1970, the contribution of value added of industry and services to the total GDP are only computed for the reference countries listed in Maddison (1994). For 1970-1990, data are derived mainly from the World Bank (1998). In the Energy Economy model of the Energy-Industry Subsystem of IMAGE 2, the value added of industry and services are used as input for simulating the activity in the industrial and commercial sectors. Historical data on value added of industry and services are scarce.
Most of the data are directly derived from World Development Indicators (WDI, 1998) which present time series for most of the world's countries from 1960 until 1995. All national data should be converted from the national currencies (LCU) into US$ in order to facilitate the calculation of the average values and to enable cross-national comparisons. In addition, all data should be presented in constant prices to deal with inflation.
Box 1. Conversion method for total
GDP_(1995)US$c,y = GDP_(1987)LCUc,y * DEFLATORc * CF(1995)c
Where data were missing, a growth percentage for GDP per capita derived from the Penn World tables (Heston and Summers, 1991) was used instead. The calculation of this percentage for each country c is represented by the equation in Box 2.
Box 2. Conversion method for growth percentage of GDP per capita
Yc = [(GDP1995/GDP1975)(1/(1995 -1975)) - 1 ] * 100
For the Eastern Europe region, WDI (1998) presents complete historical data series for the period 1970 - 1995 only for Hungary. Albania, Bulgaria and Poland have data since 1980, Romania since 1975. Missing data for Albania are estimated with the growth rates for GDP/cap from the PWT (1991). Bulgaria, Poland and Romania are estimated with the GDP/cap data presented by Maddison (1994). WDI (1998) presents data for the Czech Republic and Slovakia for the period 1984 - 1995, which are taken as one because Maddison (1994) only presents historical GDP/cap data for Czechoslovakia.
Box 3. Trend used for former Soviet Republics before 1987
GDPcap,R,y = GDPcap,RF,y * (GDPcap,R,1987 / GDPcap,RF,1987)
As a final last step, the total GDP per region was corrected for the missing countries (see this table for a listing of the number of missing countries and percentages population per region). The calculated regional GDP/cap (based on real data) were assigned to the missing countries, and added to the calculated total GDP, by multiplying it with their population.
Starting point are the 1970 - 1995 GDP per capita estimates as described above. Maddison presents total GDP, population and GDP per capita estimates for 56 countries for the period 1820 - 1992, accounting in 1992 of 90 percent of world product. Units used are 1990 Geary-Khamis dollars. We used the trend of Maddisons GDP per capita for the period 1890 - 1970 and forced them to match our 1970 estimates for the different regions (in 1995 US$).