I. ESTIMATIONRaymond W. Goldsmith
II. DISTRIBUTIONRobert J. Lampman
Most of the conceptual difficulties encountered in the measurement and analysis of national wealth can be avoided if national wealth and the related concepts of national assets and national net worth are treated within a system of national accounts. The assets, wealth, and net worth of individual economic units and of groups of them should also be treated within this system.
National wealth in a system of national accounts
A system of national accounts, like a system of business accounts, has two main components. The first is a statement of flows of economic objects (commodities, services, money, and other financial assets) during a period of time, usually a year. The second is a statement of the stocks of economic objects, physical and financial, existing at a point in time. These stocks are recorded in the national balance sheet and in the balance sheets of less comprehensive groups. The groups may be broad sectors, such as all households, all business enterprises, or all government organizations within a nation; or smaller subsectors, such as households of a given type or individual industries.
In this approach it is necessary to distinguish, in accordance with business accounting practices, between combined and consolidated balance sheets. The combined balance sheet of a group of units or a sector is the result of adding items of assets and liabilities as they are found in the balance sheet of each constituent unit. The consolidated balance sheet cancels creditor–debtor and holder–issuer relationships among units both of which belong to the group or sector. In the consolidated balance sheet of the American banking system, for instance, the deposits that commercial banks keep with Federal Reserve banks or with other commercial banks disappear. Similarly, in a national balance sheet all claims of one national against another and all holdings by a national of equity in domestic business enterprises are eliminated. Hence, in a national balance sheet the only assets that remain are tangible assets and net claims against foreigners.
National wealth thus is equal to the sum of domestic tangible assets and the net foreign balance. The net worth that appears opposite national wealth and with the same amount is necessarily equal to the combined (or the consolidated) net worth of all ultimate domestic units, that is, of households and governments. Similarly, the wealth of a sector or of a single economic unit is equal to the sum of its tangible assets and its net “foreign” balance, the latter referring to all its claims against and all its liabilities to other sectors or units (including equity securities). The wealth of a sector or unit thus equals the sum of its tangible and financial assets less its liabilities. This difference is called its net worth.
For some purposes, particularly the study of financial structure and development, what is needed is not the consolidated but the combined balance sheets of groups, sectors, or nations. In these balance sheets the sum of tangible and financial assets is equal to the sum of liabilities and net worth, no distinction being made between intragroup and intergroup claims, liabilities, and equities. The footing of a combined national balance sheet may be called national assets.
These basic relations are illustrated in Figure 1, in which it is assumed that foreign assets exceed
foreign liabilities. In each of the three sections of the chart, assets are shown in the lefthand bar and liabilities and net worth in the righthand bar. Hence, net worth exceeds domestic tangible assets. If the same asset is carried at identical values in the balance sheet of creditor and debtor (or the holder and issuer of equity securities), domestic financial assets must equal domestic financial liabilities.
The scope of national wealth
According to the basic definition of national wealth, it consists of the assets remaining in the consolidation of the balance sheets of all economic units within the national boundaries. The concept therefore includes nonreproducible tangible assets, such as land and subsoil assets; reproducible fixed and movable tangible assets, such as buildings and other structures, machinery and equipment, vehicles, and consumer durables; inventories of monetary metals, raw materials, work in process and finished goods; and the excess of foreign assets over foreigners’ holdings of domestic claims, equities, and tangible assets. There remain a few problems of whether to include or omit specific types of assets which are listed below. The decision will usually rest on the uses to be made of the estimates and the availability and reliability of data.
(1) Military assets. There is no reason to omit these if a comprehensive picture of national wealth is wanted, but they should be separated from nonmilitary assets for purposes of analysis.
(2) Works of art and collectors’ items. In principle these should be included at market value, which is fairly well defined for several assets of this type. In practice they are usually omitted because of lack of information on a sufficiently comprehensive basis.
(3) Natural resources. These are excluded insofar as they cannot be separately appropriated or sold, as is the case with sunshine and precipitation. Otherwise they are included, directly or indirectly, as part of the market value of land.
(4) Human resources. These are omitted because human beings are not considered part of national wealth unless they can be appropriated. Where slavery exists, the market value of slaves, which in part reflects their training, constitutes a separate category of national wealth. However, current and cumulated expenditures on education and training are so important in comparison with the usual forms of capital expenditures on structures and equipment that they must be considered and compared with national wealth in the analysis of economic growth and many other problems involving productivity and income distribution [SeeCapital, human].
(5) Rights, such as patents, copyrights, good will, and monopoly profits. Any value that may be put on such rights by the owner disappears in the consolidation of individuals’ balance sheets, since it is reflected in the market value of equity or offset by corresponding imputed liabilities in the balance sheets of buyers or users. For similar reasons the taxing power of governments is not reflected either in national wealth as an asset of the government or in national net worth as a deduction from the net worth of the taxpayers.
The numerous individual objects included in national wealth, or in the balance sheets of economic units or sectors, can be combined or consolidated only if each is given a value that is additive substantively as well as formally. Formal additivity requires valuation in the same unit of account, usually the legal tender of the time and place. Substantive additivity calls also for the use of methods of valuation that are economically uniform for all types of assets. This requires, first, that physically or economically identical assets enter the balance sheet at identical values and, second, that the values for assets of different type, age, physical condition, or marketability be determined by methods that are similar in economic character, although they may differ in sources and technique.
Use of current market prices. Valuation at current market prices, or the nearest feasible approximation to them, is the only method that meets the requirements of both formal and substantive additivity. The use of book values, for instance, would combine different values for economically identical assets. It would also generally produce noncomparable values for different types of assets and for the wealth of different sectors, depending on the time of acquisition of the assets and on their price movements between the time of acquisition and the balance sheet dates.
Valuation of net worth. The problem of valuation arises in connection with the individual physical or financial assets and the liabilities that make up national wealth or appear in the balance sheets of economic units or sectors, and in connection with the net worth of intermediate economic units, primarily corporations and other business enterprises. It does not arise for the net worth of ultimate economic units. This can be determined only as the difference between the sum of their separately valued assets and their similarly valued liabilities, since by definition the net worth of ultimate economic units cannot be the object of a market transaction.
The net worth of intermediate units can often be determined both by the residual method (the only one applicable to ultimate units) and by the direct method of using the market price of corporate stock. These two valuations need not, and generally do not, coincide either for individual corporations or for groups of them. The difference between these valuations often makes necessary special reconciliation items in the national balance sheet, where the net worth of intermediate units (particularly corporate enterprises) may be valued in the balance sheet of the holders by the direct method but in the balance sheet of the issuers by the residual method. Alternatively, it may require the adoption of either the direct or the residual method in the balance sheets of both holders and issuers of corporate stock.
Limitations of market valuation. The principle of valuation at current market price encounters a number of conceptual or practical difficulties.
First, only a fraction (usually a small one) of the total stock of any type of physical or financial asset is actually traded within a period close to the balance sheet date. The market valuation of assets must, therefore, as a rule proceed on the assumption that the price established in these actual transactions is representative of all items of the same type in existence at the balance sheet date. This presumption is generally reasonable for many important categories of assets, particularly securities and fairly standardized types of land, structures, and equipment. Appropriate adjustments must be made where actual transactions are not a random sample of the class of assets they are intended to represent, either because the transactions are of a special character (such as distress sales) or because the assets actually sold differ in composition, age, or condition from the stock of assets of the same class.
Second, market prices do not exist for certain types of assets that are never sold, and cannot be ascertained where transactions are too rare or the assets are not sufficiently homogeneous to justify the assumption that the items sold are a random sample of the class. The main examples are large and specialized structures and equipment. In these cases a substitute method of valuation must be used. The two methods most generally applied are replacement cost and the capitalization of expected future net income. Both methods derive their basic justification from the assumption that the market value of an asset tends, at least in the long run, to be equal both to its current cost of reproduction and to its future net earnings capitalized at the yield established by the market for assets of similar character.
Third, the market value, although established by actual ordinary transactions, may not reflect the evaluations of the owner or of the debtor or issuer. In such cases, the value used in the estimation of national wealth would not be the one that presumably guides the transactor’s behavior.
Fourth, actual market price may not coincide with the price that would obtain in a perfect market. These discrepancies are not peculiar to the evaluation of national wealth. As in other cases of market imperfection, there is usually no practical way to eliminate them even if it were regarded as theoretically desirable to do so.
Undepreciated and depreciated values. In principle there exist two values for each reproducible physical asset, even if the market price criterion is accepted: the value of the asset as it is at the balance sheet date, and the value the asset would have if it were then in the physical condition in which it entered the national economy. Actually, market prices “as is” can be ascertained for only a few categories of fairly standardized old (used) reproducible assets, such as single-family houses and cars. The value of all other reproducible assets must be approximated by reducing the current value of the asset as new (usually the calculated replacement cost new) by an allowance for depreciation. This allowance is necessarily somewhat arbitrary—precisely because of the absence of, or the nonrepresentative character of, prices of used assets. It is determined by the assumptions made regarding the length of useful life of the asset and the distribution of total depreciation (original cost less scrap value) over that life. There are therefore as many alternative estimates of the value of the stock of reproducible tangible assets as there are defensible methods of depreciation, and these estimates may differ considerably from one another.
Deflation of wealth estimates. Because the prices of different types of assets show both cyclical fluctuations and trends, it is necessary to reduce estimates of national wealth to a common temporal price basis if they are to be used in the analysis of economic growth. Reduction to a common local basis is also often necessary for international comparisons. In both cases this process of “deflating” the original estimates expressed in current prices is much less important, if required at all, for the analysis of the structure of wealth or the relation of wealth to income.
Deflation of national wealth estimates and balance sheets is even more difficult than similar operations on national product estimates, both theoretically and practically. The additional theoretical difficulties arise primarily in the case of those financial assets which have a fixed money value. These assets have no obvious physical dimension that can be regarded as the object of the adjustment for price changes, such as exists both for tangible assets and for most of the commodity and service flows of the national income and product accounts. The use of indexes of wholesale or consumer prices, or of other indicators of change in the purchasing power of the currency in which the wealth estimates are expressed, is a conventional but far from satisfactory solution. The additional practical difficulties stem from the paucity, limited reliability, and doubtful comparability of the information available on the prices of physical assets.
Methods of measuring national wealth
In practice, national wealth can hardly ever be measured on the basis of existing balance sheets of sectors or smaller groups of units because such documents are available only for part of the economy (often only for corporate business enterprises) and because they do not apply a consistent system of definition and valuation of wealth. Most existing evaluations of national wealth, therefore, are built up from separate estimates of the components of national wealth. These estimates may cover the entire economy or, more commonly, one of several sectors; and they may use very different sources and methods of estimation. The final national wealth estimates are, then, synthetic, representing a combination of a number of estimates for individual components and sectors. Such synthetic estimates are acceptable so long as reasonably uniform methods of valuation are used for the different components. The main methods used in estimating components of wealth are summarized below.
Exhaustive (item-by-item) enumeration and direct valuation (census method) is usually feasible only for a few components of national wealth, such as agricultural land and single-family houses, where it can be based on the declarations of owners or occupants. The census method is extremely expensive if applied to a large number of nonstandardized properties. Probably its most extensive use is represented by the inventory in the Soviet Union of most of the country’s fixed reproducible assets as of the end of 1959, on the basis of an engineering appraisal of millions of individual structures and pieces of equipment (Ryabushkin 1961).
Sample evaluations, usually by engineering appraisal, of a limited number of properties of a given type may be blown up to the national total for the same item.
Property tax valuation is applied primarily to real estate. Since tax values commonly deviate from current values, an adjustment, usually based on a sample of properties, is required.
Fire insurance values may be used if insured values are close to market values (or if the difference can be ascertained) and if the ratio of insured to uninsured properties is known.
Perpetual inventory (replacement cost) estimates are based on the assumption that the market value of reproducible assets is equal to or close to their replacement cost. In this case the current value can be estimated by adjusting the original cost of classes of reproducible assets (reflected in past capital expenditures on them as given, say, in national income accounts) for changes in the cost of construction and the prices of equipment and for capital consumption. This method has been used increasingly in the last decade to estimate the value of reproducible fixed assets, which in most countries contribute between half and three-fourths of national wealth (Goldsmith 1950). It is particularly important for types of tangible assets that have no market value, such as government structures, military equipment, and large private nonresidential structures. Most of the estimates in Table 4 have been derived by this method.
Personal property tax returns can be used in the few cases where individuals have had to file returns listing their assets in sufficient detail and at sufficiently uniform valuations to permit an estimation of the wealth of the entire personal sector.
Samples of personal balance sheets obtained from a relatively small number of households have become another source of estimation of the value of all or selected assets of the personal sector. This approach, made possible by the development of modern sampling and interview methods, has been widely used in the United States and Great Britain (Goldsmith 1955-1956; Lydall 1955).
Estate tax or probate returns provide a source which is based on the assumption that the detailed statements of assets and liabilities, at prescribed methods of valuation requested for tax or other purposes, of persons within an age bracket dying during the year may be regarded as a sample of the wealth of all living persons of the same age. The wealth of the living having wealth equal to at least the lower limit to which the estate tax or probate rules apply can then be estimated by multiplying the reported figures by the inverse of the mortality rates applicable to the different age groups of decedents (Lampman 1962).
Capitalization of income can be used when the gross or net income of a specific type of wealth is known, say, from the national income accounts, and where a market rate of capitalization exists. This method is particularly important for types of tangible assets to which the perpetual inventory estimation cannot be applied, such as land and subsoil assets.
Table 1 provides a condensed picture of the main components of national wealth and of the balance sheets of the main sectors of the United States economy at the end of 1958, the last year for which a detailed comprehensive estimate has been made.
While the combined assets of all sectors were in excess of $3,700 billion (excluding nearly $90 billion of military assets), the consolidated assets, that is, the national wealth, amounted to only $1,700 billion, consisting of domestic tangible assets of about $1,650 billion and net foreign assets (including monetary metals) of $50 billion. The difference of about $2,000 billion between combined and consolidated assets is represented by domestic financial assets, which disappear in consolidation, where they are offset by domestic liabilities and equities.
There are considerable differences between the broad definition of national wealth used in Table 1 and narrower definitions that are sometimes preferred. Thus, a definition of national wealth excluding military assets and consumer durables results in an estimate of not quite $1,500 billion. A still narrower definition, limiting national wealth to so-called productive assets (denned as the tangible assets of business enterprises plus net foreign assets, and excluding all government tangible assets, residential real estate, and consumer durables), further reduces national wealth to about $850 billion.
Approximately 18 per cent of national wealth in the broad definition (but excluding military assets) consists of nonreproducible assets, chiefly agricultural and residential land. Structures, of which about half are residential, account for nearly 50 per cent. Total equipment contributes fully 20 per cent, of which producers’ durables represent 12 per cent and consumer durables 10 per cent. Inventories are equal to approximately 8 per cent of national wealth, while net foreign assets account for only 3 per cent.
|Table 1 — Combined national and sectoral balance sheet of the United States, end of 1958 (billions of dollars)a|
|NATION||Unincorporated Corporate||State and local||HOUSEHOLDS|
|a. Details may not add to totals because of rounding.|
|b. Includes government insurance and pension funds.|
|c. Excludes military and Atomic Energy Commission assets ($21 billion structures, of which $4 billion outside United States; $52 billion producer durables; $16 billion inventories, according to estimates in Goldsmith 1962, p. 405).|
|d. Included in column 8.|
|e. Includes U.S. government short-term obligations.|
|f. Excludes U.S. government short-term obligations.|
|g. Only U.S. government securities; others included in column 8.|
|h. Includes only bonds and debentures (other than U.S. government short-term obligations) and mortgages.|
|i. Includes net worth of unincorporated nonfarm enterprises.|
|Source: Goldsmith & Lipsey 1963.|
|1. Physical assets||1,653||182||108||490||10||57c||173||632|
|Residential structures and land||472||19||20||26||1||1||6||399|
|Nonresidential structures and land||672||104||44||240||9||48||161||66|
|Inventories (including livestock)||130||26||17||79||-||8||0||-|
|II. Financial assets||2,082||25||30||276||693||58||30||970|
|Insurance and pension claims||200||7||–||_||–||_||–||193|
|Other short-term claims•||307||0||16||107||151||21||6||6|
|Bonds and debenturesf||358||5g||0g||7||240||7||8||91|
|IV. Net worth||2,247||187||97||508||71||-182||140||1,426i|
|V. Total assets (I + II = III + IV)||3,735||208||138||766||704||116||203||1,602|
|Table 2 — The national wealth of the United Statesa selected dates, 7805-7958 (billions of dollars)|
|a. Not including military assets (negligible 1805-1929; $73 billion in 1945; $89 billion in 1958).|
|b. Does not include subsoil assets.|
|c. Includes producer durables.|
|d. Includes monetary metals.|
|e. Summary estimate for 1966 is 2,460 in current and 1,550 in 1947/1949 prices (Kendrick 1967, p. 12).|
|Sources: Lines l-VI-1805 and 1850 from Goldsmith 1952, pp. 307, 310, 317; 1900-1958 from Goldsmith 1962, pp. 117-118.|
|Line VII—1805 and 1850 linked to 1900 figure on basis of estimates in 1929 prices in Goldsmith 1952; 1900-1958 from Goldsmith 1962, pp. 119-120.|
|Line VIII—1805 and 1850 based on Gurley & Shaw 1957, p. 256; 1900 from Kendrick 1961, p. 298; 1929-1958 from U.S. Office of Business Economics 1958, pp. 118-119.|
|Line IX—averages of midyear figures in U.S. Bureau of the Census 1960, p. 7.|
|1. Nonreproducible assetsb||0.70||3.02||31.0||114||122||311|
|V. International assets (net)d||–0.06||–0 . 15||–0.7||17||22||50|
|VI. National wealth, current prices•||1.30||7.17||87.7||439||576||1,703|
|VII. National wealth, prices of 1947/1949•||3.20||29.40||314.0||779||789||1,244|
|VIII. National product (gross), current prices||0.85||2.60||18.7||104||214||445|
|IX. Population (millions)||6.36||23.67||76.8||122||141||176|
Growth and structural changes since 1805
The main function of national wealth estimates is to permit analysis of the growth of national wealth and of structural changes in it. The essential data are shown in Table 2. The estimates for 1805 and 1850 are only very rough approximations. The figures for these two dates do not include the value of slaves, although they then constituted part of private and national wealth. Slaves would add nearly $200 million, or 30 per cent, to the national wealth in 1805, and about $800 million, or almost 20 per cent, in 1850 (Goldsmith 1952, pp. 317-318).
An analysis of the rate of growth of national wealth requires estimates in constant prices, which are shown in line VII of Table 2. Total reproducible wealth, the aggregate most relevant for growth analysis, increased almost six hundred times between 1805 and 1958, or at an average annual rate of slightly more than 4 per cent. The estimated rate of growth was considerably faster during the nineteenth century—5.2 per cent—than in the twentieth, when it declined to 2.6 per cent. The growth rate, however, recovered in the postwar period (through 1958) to nearly 4.0 per cent. More significant is the estimated rate of growth of reproducible national wealth per head. This rate was nearly 2.0 per cent for the whole period 18051958; 2.6 per cent for the nineteenth century; 1.2 per cent for the first 60 years of the twentieth century; and 2.2 per cent for the postwar period 1945-1958.
One of the outstanding changes in the structure of national wealth is the decline of nonreproducible assets, practically without interruption, from 55 per cent in 1805 to 18 per cent in 1958. This reflected primarily the fall in the share of agricultural land from almost half at the beginning of the nineteenth century to 6 per cent at present.
Within reproducible wealth, changes can be followed in detail only since 1900, a period in which they have been relatively moderate. The share of residential structures declined very slightly, remaining close to three-tenths. That of livestock dropped sharply from over 5 per cent to less than 2 per cent, and that of inventories from 12 to 8 per cent. On the other hand, the shares of producer durables and consumer durables together rose from 22 per cent to 27 per cent. (Military equipment, not included in the total, was equal to 6 per cent of it in 1958 and to about one-fourth of civilian equipment.) Closer analysis of structural changes in national wealth requires detailed estimates by sectors and by types of structure and equipment, which are not yet available for long periods of time.
Financial assets and national product
Much of the value of estimates of national wealth lies in the possibility of comparing them with estimates of financial assets and of national product.
The ratio of financial assets to national wealth, which measures the size of the financial superstructure in comparison with the real (tangible) infrastructure, increased considerably during the nineteenth century (Table 3). By 1850 it was
|Table 3 — The relation of national wealth to financial assets and national product in the United States, selected dates, 1850-1958|
|FINANCIAL INTERRELATIONS RATIOa||Including nonreproducible assets||Excluding nonreproducible assets||Including military assets|
|a. Financial assets (including equity in unincorporated business) divided by national wealth.|
|b. Net stock of wealth divided by gross national product.|
|Sources: Col. 1, 1850 roughly estimated, 1900-1958 from Goldsmith 1960. Cols. 2-4 calculated from Table 2.|
probably on the order of half, considerably above the level of the beginning of the century. In 1900 it had advanced to approximately three-fourths. It continued to rise until 1929, when it reached a value of one and one-fourth, partly under the influence of the sharp advance in stock prices during the 1920s. Since then the ratio has not shown a definite trend. Throughout the 1950s, it was close to one and one-fourth, after a sharp but temporary bulge attributable to the repressed inflation of World War II.
The ratio of national wealth to national product increased sharply during the nineteenth century; stayed at nearly the same level from 1900 through 1929; declined sharply until 1945; and during the next 15 years regained only part of the loss. As a result, the 1958 ratio was still well below the levels of 1929 and 1900. The direction of the movement of the ratio is not changed if deflated values are used for the numerator and denominator, if military assets are included, or if nonreproducible assets are excluded. The extent of the change, however, is sometimes substantially affected by these differences in definition. The interpretation of these movements in the capital-output ratio is not easy and requires more detailed figures than are now available. The variations in the ratio reflect in part the shift first toward capital-intensive sectors (railroads and public utilities) and then the opposite shift toward sectors that require relatively little capital (services). The movements also reflect changes in production functions within sectors or industries, particularly the relative importance of capital-saving technology and changes in the degree of utilization of plant and equipment.
The basic characteristics of the structural changes in national wealth that have taken place in the United States since the beginning of the nineteenth century are also evident in the differences that appear in a comparison of the structure of the national wealth of 17 countries in the 1950s (Table 4). Unfortunately, the estimates available for such a comparison differ greatly in scope, detail, and reliability, so that conclusions must be drawn with great caution.
A first characteristic of the comparison is the inverse relationship between the share of nonreproducible assets in national wealth and the level of economic development and industrialization. In India, for example, the value of nonreproducible assets, mostly representing agricultural land, is at least as large as that of all reproducible assets. In the advanced countries the share of land and subsoil assets in total assets is usually less than one-fifth, and much of this represents the site value of urban land.
The differences are much less pronounced or systematic among reproducible assets (excluding consumer durables). On the average, the share of structures is slightly above three-fifths, including about one-fourth of the total for dwellings alone; that of producer durables is close to one-fourth; and that of inventories is in the neighborhood of one-seventh. There is no clear association between the share of the main types of reproducible assets and the level of economic development. Instead, large deviations from the average reflect either climatic differences (e.g., the low share of dwellings in Japan, Colombia, and South Africa) or special features of economic structure (e.g., the high share of inventories in countries with substantial international trade, such as Great Britain, Canada, and South Africa). The differences are, of course, largest in the share of net foreign assets and liabilities. Compare such creditor countries as Belgium and the Netherlands, for which net foreign
|Table 4 — Structure of national wealth of selected countries in the 1950s, reproducible tangible assetsa (columns 3+5+7) =100|
|NATIONAL WEALTH||REPRODUCIBLE TANGIBLE ASSETS|
|(2+3+5+6+7+8)||Structures||Equipment||Invenforiesb||NET FOREIGN ASSETSc|
|a. Military assets excluded.|
|b. Includes livestock.|
|c. Includes monetary metals.|
|d. Includes all government reproducible tangible assets; hence slightly overstated and column 5 understated to some extent.|
|e. Livestock only; hence substantially understated and columns 3-5 proportionately overstated.|
|f. Includes agricultural structures, but excludes commercial land.|
|g. Equipment included with structures,|
|h. Includes producer durables.|
|Sources: Goldsmith & Saunders 1959, pp. 14-16. United States from Goldsmith 1962, pp. 117-118; U.S.S.R. by Goldsmith in Rao 1964, p. 96.|
investment is equal to about 8 per cent of national wealth, with Canada or Australia (not to mention some less developed countries for which no data are available), where a considerable part of the entire national wealth is foreign-owned.
It will be noted that the structure of national wealth is very similar for the United States and the Soviet Union, so far as the relationship of residential and nonresidential structures, producer durables, and livestock is concerned. The main differences are the presumably much higher share of consumer durables in the United States and the higher share of agricultural land in the Soviet Union (no official estimates are available for these two items for the Soviet Union).
Raymond W. Goldsmith
Conference on REsearchin Incomeand Wealth 1964 Studies in Income and Wealth. Volume 29: Measuring the Nation’s Wealth. The George Washington University Wealth Inventory Planning Study. New York: National Bureau of Economic Research.
Gini, Corrado (1914) 1962 L’ammontare e la composizione della ricchezza delle nazioni. 2d ed. Turin (Italy): Bocca.
Goldsmith, Raymond W. 1950 Measuring National Wealth in a System of Social Accounting. Volume 12, pages 23-79 in Conference on Research in Income and Wealth, Studies in Income and Wealth. New York: National Bureau of Economic Research.
Goldsmith, Raymond W. 1952 The Growth of Reproducible Wealth of the United States of America From 1805 to 1905. Pages 243-328 in International Association for Research in Income and Wealth, Income and Wealth of the United States, Trends and Structure. Income and Wealth, Series 2. Cambridge: Bowes.
Goldsmith, Raymond W. 1955-1956 A Study of Saving in the United States. 3 vols. Princeton Univ. Press. → Volume 1: Introduction; Tables of Annual Estimates of Saving, 1897-1949. Volume 2: Nature and Derivation of Annual Estimates of Saving, 1897-1949. Volume 3: Special Studies, by R. W. Goldsmith, Dorothy S. Brady, and Horst Mendershausen. See especially Volume 3.
Goldsmith, Raymond W. 1960 Le bilan national des Éetats-Unis depuis la guerre. Banque Nationale de Belgique, Bulletin d’information et de documentation 35, vol. 2:157-172.
Goldsmith, Raymond W. 1962 The National Wealth of the U.S. in the Post War Period. National Bureau of Economic Research, Studies in Capital Formation and Financing, Vol. 10. Princeton Univ. Press. → Data in tables, copyright © 1962, by National Bureau of Economic Research.
Goldsmith, Raymond W.; and Lipsey, Robert E. 1963 Studies in the National Balance Sheet of the United States. 2 vols. National Bureau of Economic Research, Studies in Capital Formation and Financing, Vol. 11. Princeton Univ. Press. → Data in Table 1, copyright © 1963, by The National Bureau of Economic Research, Inc.
Goldsmith, Raymond W.; and Saunders, Christopher (editors) (1959) 1960 The Measurement of National Wealth. International Association for Research in Income and Wealth, Income and Wealth, Series 8. Chicago: Quadrangle. → First published in England by Bowes.
Gurley, John G.; and Shaw, E. S. 1957 The Growth of Debt and Money in the United States, 1800-1950: A Suggested Interpretation. Review of Economic Statistics 39:250-262. → Data in Table 2, copyright, 1957, by the President and Fellows of Harvard College. Data from this book are reprinted by permission of the publishers.
Kendrick, John W. 1961 Productivity Trends in the United States. National Bureau of Economic Research, General Series, No. 71. Princeton Univ. Press.
Kendrick, John W. 1967 The Wealth of the U.S. Finance , no. 1:10-13, 34.
Kuznets, Simon 1938 On the Measurement of National Wealth. Volume 2, pages 3-61 in Conference on Research in Income and Wealth, Studies in Income and Wealth. New York: National Bureau of Economic Research.
Lampman, Robert J. 1962 The Share of Top Wealthholders in National Wealth: 1922-1956. National Bureau of Economic Research, General Series, No. 74. Princeton Univ. Press.
Lydall, H. F. 1955 British Incomes and Savings. Oxford: Blackwell.
Rao, C. Radhakrishna et al. (editors) 1964 Essays on Econometrics and Planning: Presented to Professor P. C. Mahalanobis on the Occasion of His 70th Birthday. New York: Pergamon.
Ryabushkin, T. V. 1961 Methodological and Organizational Problems of a General Stock-taking of Fixed Assets. Institut International de Statistique, Bulletin 38, no. 2:483-488.
Schultz, Theodore W. (editor) 1962 Investment in Human Beings. Journal of Political Economy 70, (Supplement) no. 5, part 2.
U.S. Bureau of the Census 1960 Historical Statistics of the United States, Colonial Times to 1957: A Statistical Abstract Supplement. Washington: Government Printing Office.
U.S. Office of Business Economics 1958 U.S. Income and Output: A Supplement to the Survey of Current Business. Washington: The Office.
Winkler, Wilhelm (1890-1894) 1923-1928. Volksvermögen. Volume 8, pages 770-786 in Handwörterbuch der Staatswissenschaften. 4th ed. Jena (Germany ): Fischer.
Wealth and income are closely related concepts in economics. Wealth is a stock of natural resources and previously produced goods in existence at a moment in time, while income is a flow of goods and services produced or received during a stated period of time. Wealth arises, to a considerable degree, out of income, and income requires wealth for its generation. This articleis concerned with the distribution of wealth by type of holder and by the degree of inequalityof the personal distribution of wealth.
Distribution by type of direct holder
The ownership of wealth varies by type of economy. In a fully communized economy all wealth would be in the hands of government, while in a fully private economy government would hold none of the wealth. In fact, all national economies fall between these extremes. In 1958 the wealth of the United States was six-sevenths privately owned (Goldsmith 1962; see Table 1).
|Table 1 — Percentage distribution by sector of net tangible wealth (excluding military assets) of the Unted States, selected yearsa|
|a. As of end of year.|
|b. Not available.|
|Source: Goldsmith 1962, table A-19, pp. 145-146, rounded to nearest whole percentage point. Copyright © 1962, by National Bureau of Economic Research.|
|Net foreign assets||3||3||0||b|
|State and local governments||5||8||10||10|
Over one-third of the total private and public wealth was held by households and a little under one-third by corporations; about one-tenth was in the hands of farmers, and an even smaller share was held by nonfarm, noncorporate business. Nonprofit organizations such as churches and charitable foundations owned only 2 per cent of the tangible wealth.
Accounting for indirect ownership
The preceding discussion of the distribution of tangible wealth by sector should be complemented by a discussion of indirect ownership through intangible assets that are claims of one sector on the tangible assets of another. On a net basis the personal sector, along with nonprofit organizations, holds full claim to the tangible assets of the business sector, and the tangible wealth of the governmental sector is substantially offset by liabilities to businesses and persons.
Complex interrelationships among and within sectors are present in a modern private enterprise economy. For example, financial intermediaries give claims in exchange for their own assets, which are in turn largely intangible [SeeFinancial intermediaries]. To take another example, a person may borrow from a financial intermediary (which in turn holds assets deposited by other individuals) to buy securities issued by a corporation that owns other corporations.
Easy conclusions about the significance of the intersectoral distribution of net assets should be distrusted. This is especially true of the government sector, since the net assets of a government may be a most imperfect indicator of its capacity to meet its obligations. Some authors have suggested that a capitalization of government’s current authorization to collect taxes should be entered as an asset of the government and as a liability of the private sectors.
Distribution of personal sector wealth
The meaningfulness of a distribution of wealth within a particular sector is open to some dispute. It may be more important to know who actually controls an asset than to know who owns it either on a gross or on a net basis. Thus, minority owners of a business corporation may be in a position to control it because of their management responsibilities or through their control of the assets of a financial intermediary. However, there is considerable interest in knowing how wealth is distributed within the personal sector, even though this distribution may give only limited information about the distribution of power.
Personal sector wealth is made up of tangible wealth plus claims on corporate and noncorporate businesses in the form of stocks, bonds, mortgages, demand deposits, and insurance rights; and claims on government in the form of bonds, notes, and currency. The tangible wealth may be classified as real estate, producer durables, and consumer durables; the intangible wealth includes, in addition to the types mentioned above, patents, copyrights, and rights to pension funds and personal trust funds. All of these types of wealth confer a right to future income, but the nature of the other rights conferred varies considerably among the several types. Thus, the rights associated with cash are quite different from the contingent rights associated with insurance or pension funds, and still different from the restricted rights of an interest in a personal trust fund. Since all the rights and powers of wealth ownership are the subject of law and contract, a perceptive analysis will indicate by its categorization of personally owned wealth something of the consequences that ownership of each of the several types has in practical affairs. Business equities, though they indicate more influence in some economies than in others, generally have more meaning in terms of power than do, say, consumer durables.
In measuring personal wealth it is not common practice to capitalize the future earnings from one’s own personal services, but it is nonetheless true that a person’s balance sheet of assets and liabilities should be viewed with reference to the value of such earnings. In this light, investments in income-earning assets may be considered as alternative to outlays, such as those on education or health, that may increase the value of one’s personal services. [SeeCapital, HUMAN.]
The total value of a person’s tangible and intangible assets, less his obligations to others, equals his net worth, or equity. The aggregate of all persons’ net worth, less certain intangibles such as patents, equals the value of national tangible wealth less the net assets of governments and nonprofit organizations.
Sources of information. Knowledge about the distribution of wealth among persons is imperfect and incomplete. No full census of wealth has ever been undertaken in any country, and the few existing studies are incomplete both as to the coverage of wealth items and as to the part of the population surveyed. Three basic methods have been followed: censuses or sample surveys, estimates from income or property tax data, and estimates from probate or estate tax data. The most direct method is the census or survey, which involves questioning family or spending units or individual wealth-holders about the amounts of wealth of various types that they hold or have held.
A less direct method is to estimate the value of income-producing wealth from income tax or property tax data. For example, the interest income attributable to a person may be capitalized at the going rate of interest to estimate the value of his debt securities. This income-capitalization method frequently suffers from such inadequacies of data that only a fragmentary picture of the over-all distribution can be developed, giving rankings of asset holdings of particular types by income classes. A related method is to work from property tax returns, which in some countries permit an estimate of the composite total of the assessed valuations of at least limited categories of wealth.
Another indirect method of estimating wealth distribution is to work through records of the wealth held by deceased persons as provided for probate proceedings or for death tax purposes. From these records it is possible to compile, first, a distribution of wealth among the persons who died in a given year. This distribution may then be converted into an estimated distribution of wealth among living persons by use of the estate-multiplier method, developed by Bernard Mallett in England in 1908. Corrado Gini of Italy and G. H. Knibbs of Australia were among the first to follow Mallett’s method. [See the biography of Gini.]
This method involves multiplying the numbers of deceased persons and the amounts of wealth they held by the inverse mortality rate for their age, sex, and socioeconomic group. Thus, if x is the number of men aged 70 who died in a year, each holding $1 million, then x times the appropriate inverse mortality rate is the number of 70-year-old men holding $1 million who were living in that year. This procedure must be carried out for each age-sex group in order to construct a representation of wealth-holdings by size, by type of property, and by characteristics of wealthholders.
A principal difficulty with the method is the uneven quality and limited availability of data. Probate information is locally filed and not centrally recorded. Moreover, laws concerning the information to be recorded vary widely across jurisdictions. Estate tax data are less than ideal materials for the student of wealth distribution because of the relatively high minimum value of estates for which tax returns must be filed. In the United States in the 1950s, the minimum of $60,000 meant that only about 2 per cent of all decedents were represented on such returns. In Great Britain, on the other hand, a relatively low filing requirement of £100 has yielded a much larger sample of decedents on the tax returns. Numerous other difficulties arising out of the nature of tax and property law prevent an easy translation from estate tax returns to a representation of wealth-holdings by living persons.
Socioeconomic distribution. Personal wealth is distributed in a pattern quite different from that of income. The distribution of wealth is much more unequal than that of income, but not all persons in the highest income classes are also in the highest wealth classes. Similarly, the lower income groups may include some moderately wealthy individuals. This is explained to some extent by the life cycle of income and the timing of wealth accumulation and receipt of inheritances, which concentrate wealth in the hands of older persons. Because wives are typically younger than husbands and because of the greater life expectancy of women than of men in many countries, wealth shifts toward women as age increases. Proceeds of life insurance, in which wives frequently are named as beneficiaries, account for the inclusion of younger widows among substantial wealth-holders in some countries.
The Survey of Consumer Finances found that in the United States in 1953 the share of wealth held by the top 10 per cent of spending units, ranked by net worth, was about 60 per cent; their share of income was about 23 per cent. In the same year the top 10 per cent, ranked by income, received about 30 per cent of income, but only about half of top income receivers were simultaneously top wealth-holders. The Survey also reported that selfemployed, managerial, and professional persons were heavily overrepresented in the upper reaches of both income and wealth distributions.
Further information about wealth distribution in the United States is provided by Lampman’s study (1962) of top individual wealth-holders, which was based upon federal estate tax data. According to this study, the top 1.6 per cent of adults (those having $60,000 or more of gross estate in 1953) had a median age of 54 years, and there was a positive association between age and size of estate within the group. Women made up one-third of the group and held 40 per cent of the wealth. This top 1.6 per cent held an estimated 30 per cent of total personal wealth and substantial shares of most specific types of wealth. Business assets were more highly concentrated than consumer assets; while the top group referred to above held over threefourths of corporate stock, they held only 16 per cent of real estate and 13 per cent of life insurance. (See Table 2.)
|Table 2 — Share of the top 1.6 per cent of adults in personal sector wealth, United States, 1953|
|TYPE OF PROPERTY||TOTAL WEALTH IN PERSONAL SECTOR (Billions of dollars)||SHARE OF WEALTH HELD BY TOP WEALTH-HOLDERS (Per cent)|
|* Not included in the total because the assets of such funds are distributed among the preceding categories.|
|Source: Lampman 1962, table 90, pp. 192-193.Copyright © 1962, by National Bureau of Economic Research.|
|State and local government bonds||16||100|
|Mortgages and notes||31||36|
|Life insurance reserves||78||13|
|Pensions and retirement funds||64||6|
|Miscellaneous property, including equity in noncorporate business and consumer durables, etc.||221||18|
|Personal trust funds*||(54)||(70)|
|Debts and mortgages||133||22|
The degree of concentration of wealth varies among countries because of differences in industrial and occupational patterns, the age-sex composition of the population, and many other factors. Although the income distribution is less unequal in Great Britain than in the United States, the reverse is true of the wealth distribution. The latter finding is explained to some extent by the greater importance of agriculture and of owner-occupied housing and consumer durables in the United States (Lydall & Lansing 1959). No fully comparable studies have been made for other countries.
Historical changes in wealth inequality. It appears that in both the United States and Great Britain the trend has been toward diminished concentration of wealth. Langley (1950-1951) reports that the top 1 per cent of adults owned 70 per cent of the wealth in England and Wales in 1911-1913, but only 50 per cent in 1946-1947. Lampman similarly finds a drop in the share of wealth of the top 1 per cent in the United States from 32 per cent in 1922 to 25 per cent in 1953. However, the latter drop is to some extent explained by transfer of wealth within families; the share of wealth of top families did not fall so much.
Factors influencing distribution. The distribution of wealth at any moment is the resultant of many past events. Each person represented in the distribution has experienced some variation in income, saving, dissaving, borrowing, value of assets held, and making and receiving gifts, inheritances, and transfers of other kinds. Each year each person will experience a series of such changes, and his position in the wealth-holder ranking may shift from year to year. Moreover, new persons enter the ranking each year and others die. Over a long period the age-sex composition of the population may change, the relative prices of assets may shift, behavioral characteristics of people may be altered, or institutional arrangements may change so as to restrict or enlarge individuals’ freedom and ability to accumulate or transfer wealth.
The many factors involved in the determination of the wealth distribution may be classified as behavioral, exogenous, demographic, and institutional. Behavioral factors include propensities to save, to choose among assets and debts, and to choose among alternative ways to transfer wealth before and at the time of death. For purposes of this analysis, the level and distribution of income and the levels of asset prices may be taken as exogenous. Changes in these variables will have an independent effect upon the wealth distribution.
Changes in the age-sex composition of the population may influence the wealth distribution even though the behavioral characteristics of each age-sex group remain the same. Institutional changes may also lead to a new pattern in the over-all wealth distribution. Examples of such changes are changes in the progressivity of income, estate, gift, or inheritance taxes and changes in laws or customs concerning methods of holding or transferring wealth. Changes in the rights of women and minors to hold wealth; the development of insurance, pension funds, and trust funds; and modification of governmental loan or subsidy programs are further examples of institutional change.
Changes in the over-all distribution of wealth, then, may be seen as the net result of changes of the several kinds detailed above. In some historical situations price changes may be quantitatively most important; in other periods demographic or institutional changes appear to be dominant. Some of these factors may be identified in explaining the decrease in inequality of distribution of wealth among persons in the United States between 1922 and 1953. In this connection it is illuminating to know that approximately one-third of the market value of wealth in the hands of persons in 1953 may be considered as having been accumulated by previous generations, one-third as the result of rising prices over the previous thirty years, and only one-third as accumulated out of the incomes of persons then living. During the years between 1922 and 1953 the adult population grew markedly older, thus influencing the distribution in the direction of greater inequality. The changes in capital value also tended to accentuate inequality, but a lessening of income and savings inequality and the wider spreading of wealth among family members worked to offset it.
Theory of wealth distribution change. The theory of the determination of wealth distribution is underdeveloped, as is empirical study in this field. The Marxian proposition that a capitalistic society has a tendency toward increasing inequality would seem to be based upon the observation that wealth yields income, which in turn is the source of wealth, and the idea that an original inequality may be reinforced by the institution of inheritance. The plausibility of this proposition should not lead one to ignore the self-limiting and offsetting processes in the picture. A self-limit is apparent in the fact that increases in wealth reduce its average yield, and this in turn controls the share that property income is of total income. Another example of a self-limit is found in the relationship between a top group’s share of wealth and its share of saving. In year-to-year change the group will maintain its share of wealth only if its share of saving is equal to its share of wealth, ceteris paribus. But as the share of wealth increases, it becomes increasingly difficult for the group to maintain a parallel change in share of total saving. Demographic changes also tend to be self-limiting.
An example of an offset to increasing inequality is the fact that rich people transfer wealth to more beneficiaries than do the less wealthy. The “creative destruction” of changing market conditions may also work to offset an existing pattern of inequality, with innovations and obsolescence combining to destroy an initial wealth advantage of one group or another. Finally, social policy may work to offset a tendency toward increasing inequality. Specific areas in which policies may influence the distribution of wealth include education, labor markets, social security, taxation, land tenure, home ownership, credit, insurance, and the law of inheritance.
Relatively little research has been done on either national wealth accounting or the distribution of personal sector wealth. More research is needed on methods of estimating the wealth-holding of persons, on the relationship between income and wealth distributions, on savings and transfers of wealth by estate size, on the use of personal trust funds and foundations, and on capital gains and losses. The theory of changes in wealth distribution should be explored in various ways. One promising method is the use of simulation techniques, in which a hypothetical population is assigned a set of behavioral characteristics and demographic, exogenous, and institutional changes are successively assumed. [SeeSimulation, article onEconomicProcesses.]
Further comparative research is also needed to find how wealth distribution varies by type of economy and by stage of development. Finally, research is needed into the social and political significance of changes in the forms of wealth claims and of altered sharing of wealth among sectors.
Robert J. Lampman
Atkinson, Thomas R. 1956 The Pattern of Financial Asset Ownership: Wisconsin Individuals, 1949. Princeton Univ. Press.
Butters, J. Keith; Thompson, L. E.; and Bollinger, L. L. 1953 Effects of Taxation: Investments by Individuals.Boston: Harvard Univ., Graduate School of Business Administration, Division of Research.
Conference on Research in Income and WealthStudies in Income and Wealth. → Published since 1937. See especially volumes 3, 12, and 14.
Goldsmith, Raymond W. 1955-1956 A Study of Saving in the United States. 3 vols. Princeton Univ. Press. → Volume 1: Introduction; Tables of Annual Estimates of Saving, 1897-1949. Volume 2: Nature and Derivation of Annual Estimates of Saving, 1897—1949.Volume 3: Special Studies, by R. W. Goldsmith, Dorothy S. Brady, and Horst Mendershausen. The Mendershausen study is the most directly relevant. However, all of these volumes and Goldsmith 1962 are the basic works in the field of wealth accounting.
Goldsmith, Raymond W. 1962 The National Wealth of the United States in the Post War Period. National Bureau of Economic Research, Studies in Capital Formation and Financing, Vol. 10. Princeton Univ. Press.
International Association for Research in Income and WealthBibliography of Income and Wealth.→ Published since 1952.
Internationalassociation for Research in Income and WealthIncome and Wealth. → Published since 1949.
Lampman, Robert J. 1962 The Share of Top Wealthholders in National Wealth: 1922-1956. National Bureau of Economic Research, General Series, No. 74. Princeton Univ. Press.
Langley, Kathleen M. 1950-1951 The Distribution of Capital in Private Hands in 1936-1938 and 19461947. Oxford University, Institute of Statistics, Bulletin 12:339-359; 13:33-54.
Lydall, Harold F.; and Lansing, John B. 1959 Comparison of the Distribution of Personal Income and Wealth in the United States and Great Britain. American Economic Review 49:43-67.
Morgan, Edward V. 1960 The Structure of Property Ownership in Great Britain. Oxford: Clarendon.
Sargan, J. Dennis 1957 The Distribution of Wealth. Econometrica 25:568-590.
Survey of Consumer Finances. → Published since 1946 in Federal Reserve Bulletin and in publications of the University of Michigan Survey Research Center.