PROBLEMS OF A CENSUS OF NATIONAL WEALTH

This paper is essentially a summary of the book Measuring the Nation's Wealth (Volume 29, Studies in Income and Wealth, New York: National Bureau of Economic Research, 1964), which is the report of study directed by the author. The purpose o f the study was to assess the problems and possibilities of conducting a national census of real wealth as a basis for continuing wealth and balance sheet estimates for the U.S. economy, by major sector.
It is stressed that the balance sheets and wealth estimates should be designed as a consistent part of an integrated system of national income accounts. Thus, valuation (at market prices and/or depreciated replacement costs), sectoring, and type-of-asset detail in the basic data and derived estimates should be compatible with the flow estimates contained in the economic accounts. Consistency of stock and flow estimates facilitates analysis of inter-relationships, and is helpful in the estimation process.
It is recommended that in the U.S. asset data by broad categories be collected as part of the recurring economic censuses and other reporting systems, but that detail on fixed reproducible assets (construction and equipment) at cost, by year or period of acquisition, be obtained from a small sample of respondents in each industry. The detail would be useful in its own right, and also permit revaluation of the assets by use of price indexes and depreciation rates to a current depreciated replacement cost basis. Where feasible, respondent estimates of market values would also be obtained.
The proposal is thus a compromise between the Japanese 1955 sample survey of assets, and the detailed wealth inventory of the U.S.S.R. which was begun in 1959. Preliminary work is now underway in the U.S. federal statistical agencies to expand collection of asset data, and to prepare comprehensive wealth estimates in the framework of the national income accounts.