Documentation for Historical Estimates of World Trends

HISTORICAL ESTIMATES OF WORLD ECONOMIC ACTIVITY, POPULATION AND LABOR FORCE, 1950-2007

A Data Book of annual historical estimates of long-term trends in the world economy is now available. Data in the Data Book include GDP, total population, and total labor force by main world region from 1950 to 2007 and estimates per capita and per economically active person. It may be obtained by sending an e-mail to Douglas O. Walker at dougwal@regent.edu.

The individual country estimates upon which the region and world totals are based may also be obtained upon request.

Documentation on the data are provided below.

General Note on the Data

This Data Book presents tables relating to long-term trends in the world economic activity, population and labor force and its major economic regions and individual countries. It is based on an update of individual country national accounts estimates and demographic data assembled from United Nations agencies and other national and international sources. Estimates presented here are part of a larger data base now under preparation intended to cover the main demographic, economic, and social trends in the world economy from 1950 to the present.

The historical data on gross product by major expenditure component, population and labor force, upon which the statistical tables are based, have been drawn from original data reported by national and international statistical and research units. The national accounts data conform in general with the United Nations System of National Accounts as presented in the publication System of National Accounts 1993, as collected on a United Nations Statistics Division questionnaire and published in its annual National Accounts Statistics publication. Data collected by the Statistics Division on its questionnaire have been supplemented by the Statistics Division, where possible, to cover the period 1970-2006 for those years and countries for which official data are not available. Where available and necessary, the Statistics Division estimates have been further supplemented by estimates reported by United Nations regional commissions, the World Bank, the International Monetary Fund, and other sources so as to arrive at a more complete set of data for the period 1950-2006. Selected series for main world regions have been extended to 2007 on the basis of preliminary estimates and forecasts.

The population and labor force data are based on those reported by the Population Division of the United Nations and the International Labour Organization, as further supplemented and adjusted to cover the period 1950-2007.

It must be emphasized that the present set of the national accounts data base is preliminary in three ways. First, it draws mainly upon readily available statistical material obtained from international agencies in machine-readable form. Statistical data in country publications and agency reports have been used only the to extent they were needed to fill gaps in detail or check on broad trends. Where gaps could not be filled, interpolations were used in some cases. Second, estimates for some countries are based on proxy indicators such as known historical shares, available price series, interpolations and extrapolations of similar series, or other approximations. Third, assumptions were made in the case of some countries about trends in individual components of the GDP, and/or individual components in some countries were estimated by residual. For all of these reasons, and others, the present national accounts data based should be regarded as a first draft, presented here to gain a idea of broad trends in the world economy and elicit comment and suggestions for revision and improvement.

Data presented in this data book are the responsibility of the compiler of the data, Douglas O. Walker, and should not be attributed to the basic sources from which they have been derived or to Regent University. The data have been prepared for use in analytical studies undertaken the compiler of the data bas, his students, and colleagues involved in academic research endeavors. They are not intended for commercial sale and may not be disseminated for any purpose beyond academic research and teaching.


Gross Domestic Product Data

The Data Book presents estimates of total gross domestic product in billions of constant 1990 U.S. dollars for the world and its major economic regions and main geographic areas for the period 1950-2007. The estimates of total GDP are taken from a set of individual country time-series for GDP by expenditure and GDP by main producing sector in current and constant 1990 national currency prepared by the United Nations as part of its National Accounts Main Aggregates Database.

The original country data relating to gross domestic product and its detail in this database have been drawn, wherever possible, from replies by governments to the annual Statistics Division questionnaire on the national accounts. The form and concepts of the data collected on this questionnaire generally agree with the recommendations of the 1993 United Nations System of National Accounts (System of National Accounts, Studies in Methods, Series F, No. 2, Rev. 4) or its predecessor SNAs.

In the first instance, for the period 1970-2006, estimates on the questionnaire have been supplemented by the Statistics Division, where necessary, by estimates prepared by the Statistics Division for missing countries, items, and years on the basis of information from other international agencies, particularly the United Nations regional commissions, the International Monetary Fund, and the World Bank. Further information on the methodology of data estimation used by the Statistics Division may be found on its web site.

In the second instance, the Statistics Division estimates have been further adjusted or complemented, as needed, by the author of the data book, using generally accepted statistical techniques so as to arrive at comprehensive sets of internationally standardized national accounts statistics for the period 1950-2006 (data for some countries and regions are only available beginning in later years). When unavailable from national and international sources, the series for GDP were backdated to 1950 on the basis of data from "The Conference Board and Groningen Growth and Development Centre, Total Economy Database (January 2008)”. This database and information on its sources and methods can be found at http://www.conference-board.org/economics"

In the third instance, key data series have been extended to 2007 on the basis of preliminary estimates and forecasts available from international agencies in August 2008.

Method of preparation of the GDP data. Although in general the basic set of original data reported by countries on the Statistics Division questionnaire conform to international recommendations, in some instances they have gaps in coverage or are in other respects unsuitable for immediate analytical use. In these cases, it was necessary to bring together data from alternative sources and further edit and supplement the questionnaire data. Estimates in other Statistical Division files and from the United Nations regional commissions and agencies and national publications have been used by the Statistics Division and the author to complete the basic set of data obtained from the questionnaire. To ensure comparability, where required and possible, supplementary estimates for each country have been adjusted where possible to conform with the United Nations' SNA.

In some cases, reported data from various sources are in current prices and constant prices in national currency. In others, basic source data are available only in current prices in national currency. For some countries, basic source data are available only for certain years and/or major expenditure components. Supplementary data to the basic set from the questionnaire are available from the other reporting agencies mentioned above in current prices and constant prices in national currency and as price indices for individual expenditure components.

Where data from a supplementary source conformed to that reported by the Statistical Division and was available for a longer period or in greater detail, it was used to complete or extend the primary set of data obtained from the questionnaire. Where available supplementary data did not conform to the primary set of data, adjustments were made in an attempt to prepare a comparable time-series. In addition, in some cases, separate time-series at current and constant prices have been adjusted to insure temporal comparability, and where necessary constant price data have been shifted to a 1990 base in order that data for different countries may be presented at the prices of some common reference year. Further information on sources and methods of preparation of the Statistical Division national accounts data is given on the Statistics Division web site.

In order to prepare the data in the present file, a review was made of the Statistics Division data. These estimates were compared with those available from other international agencies, national sources, and academic research units. In the case of some countries, IMF or World Bank data were substituted for Statistics Division estimates and used as the basic set of data. When identified, miscoding and errors of data entry were corrected. In addition, the resulting set of data were backdated to 1950 or 1960 on the basis of data from other international agencies. In some cases, when data were missing from all reporting sources for some years or some components, they were interpolated by the author on the basis of selected production and trade indicators reported by various international and national sources. In other instances, individual components were estimated on the basis of movements in relevant economic indicators, and in still other instances were determined by residual.


Exchange rate for conversion into United States dollars. Official and market exchange rates as reported by the Statistics Division for the period 1970-2006 were used as the basic set of exchange rates. In general, the Statistics Division used rates reported by the International Monetary Fund, and these series were backdated by the compiler to 1950. For countries with multiple exchange rates, an effective exchange rate was derived from trade data in national currency and United States dollars published in International Financial Statistics of the International Monetary Fund. In several countries in early years a free market rate was used. In a few countries in a few years national exchange rates based on United Nations operational rates were used because foreign trade exchange rates were not available. Because of fluctuations and distortions in estimates of GDP converted to U.S. dollars in some countries, the Statistics Division developed its price-adjusted rates of exchange (PARE), and these have been used.

Reported exchange rates for economies in transition and a few developing countries in a few years have been adjusted to translate gross domestic product levels expressed in domestic prices to levels consistent with estimates and ratios of GDP per capita expressed in U.S. dollars. These rates should be regarded as adjustment coefficients rather than foreign trade exchange rates.

Current price data expressed in U.S. dollars were computed from the current price national currency data by applying a separate exchange rate for each year to the current price national currency data. Constant price data expressed in 1990 U.S. dollars were computed from the constant price national currency data by applying the exchange rate for 1990 to the constant price national currency data for all years.

Aggregation of country data. Estimates for regions of analytical interest have been established by the compiler by aggregation of country data expressed in U.S. dollars. In forming region totals for the national accounts the following modifications were made: National currency data were translated into U.S. dollars on the basis of the exchange rates discussed above; when years covered by data for a country entering a region total did not encompass the entire period desired for the region total, an estimate for the contribution of the country to the region in the missing years was imputed to the region total on the basis of the growth rate for the region during those years excluding that country; and statistical discrepancies were re-computed as the difference between the sum of the separate components of GDP and the estimated total for gross domestic product, or they were allocated on a percentage contribution basis over the components of the grouping table, or a component was derived as a residual. In some countries, the estimate for total GDP was re-computed as the sum of the expenditure components.

Quality of data. A number of reservations must be made about the accuracy and reliability of the data under preparation.

Since the GDP is the most comprehensive estimate of a country's domestic economic activity, achievement of a high degree of accuracy requires a highly developed national statistical organization, a requirement which is beyond the financial and physical resources capacity of many countries. In the case of commodity trade statistics used as supplementary material, although based mainly on more reliable customs declarations made at the frontier, reported data nonetheless suffer from practical difficulties and administrative weaknesses in the statistical collection apparatus of many reporting countries, and in non-reporting countries are often based upon less reliable partner country statistics.

In general, when preparing estimates of economic activity and international trade for many low income countries, consumption, investment, and output originating on the farm, especially in the subsistence sector, tend to be underestimated, and because of sharp disparities in the price and wage structure between the subsistence and market sectors, they probably also tend to be relatively undervalued.

Similar under-estimation and under-valuation occur in the traditional handicraft sector which is rarely integrated fully into the market. Given the magnitude of the farm and handicraft sectors in many low income countries, this under-estimation and under-valuation also affect the GDP total and per capita estimates. In general, the effect of these factors will tend to be greater in countries with low per capita GDP and may be more pronounced in early years than later years, leading to an upward bias in implicit rates of economic growth. In some low-income developing countries the non-monetary component of GDP can exceed 40 per cent.

Conversion of values in current prices to values in constant prices is especially difficult in many developing countries, owing to the paucity of price data, the frequently small scale of the market, and the incidence of high rates of inflation. The revised SNA recommends valuing both exports and imports of goods at the customs frontier at their point of exit, that is, free on board (f.o.b., i.e, exclusive of shipping charges and other costs of transport, insurance and similar charges to their final destination but including costs of distribution to the dock of the exporting country). In fact, many countries value imports of goods at point of entry, that is, including payments made for transport, insurance and similar changes incurred in their onward transportation to the importing country (i.e., c.i.f.). Because of a lack of information, prices used in the estimation of trade volumes are unit values derived from official statistics. In the first instance, price declarations made at the frontier at the time of export or import may differ from actual prices at the time of sale, and are therefore not prices actually obtained in the market. Secondly, basic information from which unit values are derived may be limited to broad categories of products of differing assortments and changing qualities over time, and the average unit values derived from this information are affected by the non-homogeneous nature of the changing mix of items exported and imported. Finally, frequently prices used to compile economic statistics are understated because they do not consider credit conditions surrounding the transactions being measured. For these reasons, unit value indices used in commodity trade statistics and as substitutes for transactions prices when compiling national accounts statistics, especially in the case of exports and imports of manufactured goods and non-standardized production in general, cannot be expected to provide reliable measures of average price changes over time.

In this regard, the problem of quality change in the compilation of economic statistics is a particularly important one. Separate prices for products of different characteristics and qualities should be used to construct an average price index for some given class of goods and services. In practice, however, a limited number of prices are often used to deflate commodity classifications of many different kinds of related products, and frequently changes in commodity composition are not taken into account when constructing the average prices index. When unit values are calculated, the opposite problem may occur when changes in commodity composition in categories which are not strictly homogeneous. In this case, a change in the proportions of different qualities, quantities, grades and sizes of the articles being exported and imported may have taken place, with no adjustment for these changes. In the case of specialized machinery and equipment and other unique products serious distortions of true price relationships can take place. While substitution of price quotations for unit values may be used in the case of some products, it is impossible to adjust the full range of products. Consequently, prices and unit values used to convert values to volumes may contain serious sources of error.

Cross-country comparisons, as well as aggregates and averages for economy groups, suffer further from the failure of the exchange rate -- even in countries with uniform and stable rates -- to reflect with adequate accuracy the relative purchasing power over domestic goods between countries. Studies have shown that in low per capita GDP countries, the exchange rate may understate purchasing power parity by a factor of three. In this respect, likewise, the GDP in countries with lower per capita GDP. The problems entailed in exchange rate conversions are compounded under a regime of managed and variable rates. As a result of large changes in exchange rates, exchange rate conversions for adjacent years sometimes show substantial changes in relative gross domestic products between pairs of countries when no such real change has actually occurred.

Given the changes to which exchange rates may be subject in various countries, cross-country comparisons and region aggregates and averages will also be affected by the particular year selected as the base year for conversion of values in local currencies into values in a common currency.

In summary, although considerable effort has been made to standardize these data both in terms of their temporal comparability and in their conformity with accepted statistical definitions, full comparability is not possible due to weaknesses in the original data and deficiencies inherent in the method used to convert them to a common currency. Estimates for many countries tend to be based on incomplete or unreliable information and are subject to discontinuities and frequent revision, and for this reason should be used with due regard for the approximate nature of much of the underlying basic data as well as the preliminary nature of the estimates for the last two years. No significance should be attached to comparisons involving small statistical differences.

Significance of GDP for measuring material welfare. Though the GDP is intended to measure the market value of all goods and services produced domestically, in fact it measures the value of general government services at cost rather than market price and, except for subsistence farming noted above, it includes the value of household production for the household itself, whether for consumption (e.g., household care and maintenance) or for investment (e.g., in housing) only to the extent that it is provided for payment. On the other hand, it measures the market value of all marketed goods and services, including military expenditure and socially harmful consumption, together with investment for these purposes, as well as expenditures to overcome social disutilities such as increased transportation costs, pollution, or crime, that may be generated by the production process itself or by concomitant social processes such as urbanization. Since household production on own account is larger in countries with lower per capita GDP, while expenditures to overcome social disutilities generated by production or social change is larger in countries with higher per capita GDP, large differences in GDP may be associated with much smaller differences in material welfare.

Among other factors, the composition of output in different countries reflects the small and great differences that consumers place on different kinds of goods and services for reasons of tastes and preferences. The level and composition of output among countries also reflects differences in the distribution of income and other factors bearing upon the state of material welfare. No adjustment has been made for these differences and the effect they many have on inter-temporal and inter-country comparisons.

For all the reasons mentioned above, the GDP can not be used as an index of material welfare, even for measuring the change over time within a country, but especially for measuring differences between countries. It is a reasonably adequate -- although far from perfect -- index only for what it is intended to measure, namely, the output of marketed goods and services, as partially adjusted to include a limited amount of imputed output on the farm and in the household.


Population and Labor Force Data

Population estimates and projections. Population estimates and projections are those prepared by the Population Division of the United Nations under the medium variant assumption as assessed in 2006. Estimates for the years 1950-2005 are those made by the Population Division while projections for the years 2006-2025 correspond to the medium variant projection. In the case of a few countries, estimates from national sources have been substituted for those of the Population Division.

Annual estimates for the total, male and female population and the urban and rural population of individual countries, not shown in the data book but carried in the larger data base of which the estimates of total population have been drawn, are also those of the United Nations Population Division. Interpolations of five-year estimates and projections of the age-structure of the population carried in the larger data base were performed by the compiler for the period 1950-2025 to arrive at estimates (1950-2005) and projections (2006-2025) for individual years.

The medium variant assumption represents an assessment of future demographic trends, expected social and economic progress, ongoing government policies, and prevailing public attitudes towards population questions.

Labor force estimates and projections. Estimates and projections of labor force participation rates for the total labor force correspond to those of the 5th edition of its estimates and projections of the economically active population.

Reported labor force participation rates in the 5th edition provide estimates of labor force participation rates for the total, male and female labor force for the period 1980 to 2000 and projection rates to the year 2020. These rates were applied to the population series to obtain estimates and projections of the total, male, and female labor force.

The labor force estimates for the year 1980 were backdated to 1950 on the basis of trends in the ILO presented in the 4th edition of its estimates and projections of the economially active population.

In making these estimates, the female labor force was derived as the difference between the estimated total labor force and the male labor force.

The labor force comprises the economically active population, including the armed forces and unemployed, but excluding housewives, students and other not economically active groups such as the completely disabled.


Country Coverage

Region and world totals in the Data Book are based on individual country estimates aggregated in accordance with the classifications shown in the annex table: