National Income Accounting and Economic

Welfare: The Concepts of GNP and MEWby KENNETH STEWART

HE MOST comprehensive indicator of economicperformance in the nation in a given year is gross na-tional product (CNP). Changes in CNP reflect bothchanges in prices and changes in the physical volumeof output. CNP adjusted for price level changes isgenerally accepted as a reliable indicator of growth inthe nation’s total production and is used by economicanalysts to indicate whether the economy is expand-ing or contracting. Policymakers use GNP data, alongwith other measures of economic activity, in the for-mulation and subsequent evaluation of stabilizationpolicy.

A growing GNP is generally associated with ex-panding opportunities for employment and an increas-ing amount of material welfare. Economic policy fa-cilitating GNP growth is formulated, in part, as ameans of reducing both unemployment and poverty.But a growing GNP has also been accompanied byurban decay and pollution, which are not accountedfor in national income data. Critics of economicgrowth, as measured by national income data, arguethat such data tend to emphasize the growth of ma-terial welfare while ignoring what is happening to the“quality of life” or “social welfare.” CNP has beengrowing, but what has been happening to totalwelfare?

William Nordhaus and James Tohin recently pro-posed an indicator to obtain a measure of “economicwelfare” or “standard of living” to complement GNP.’This indicator, referred to as “Measure of kconomic

~William Nordhaus and James Tobin, “Is Crowth Obsolete?”,Economic Growth, Fiftieth Anniversary Colloquium, Vol. 5(New York National Bureau of Economic Research, 1972).

Welfare” (MEW), would modify the present CNPmeasure primarily in three ways: 1) by subtractingestimates of certain costs or “bads”, such as pollution,from the national income totaL 2) by excluding someservices, such as police services, since it is possible thatincreased police budgets to combat rising crime do notindicate an increase in welfare; and 3) by adding toCNP some activities, such as household activities

housework, home repairs, etc. ) and leisure, whichare not included in the CNP total -

This-article discusses the Nordhaus-Tobin measureof economic welfare. Since they use GNP as a pointof departure, the concept of ONP is reviewed in thefirst part of this paper and then compared with theproposed MEW concept.

Gross national product can be defined as the marketvalue of domestic current final output.2 It provides ameasure of the nation’s aggregate economic activity —

income or output— measured in terms of current mar-ket pi-ices over a given period of time, usually a year.

Two methods can be used in measuring the nation’sincome or output — the income approach and the ex-penditure approach. The income approach determinesgross national income by totaling the various incomeshares of tire factors of production, such as compensa-tion of employees, rental income, proprietors’ income,

2For fmirther discussion ot the GNP concept - see Anncn A -

Alchian and William B. Allen, University Economics, 3rd ed.(Belmont, California: Wadsworth Publishing Company, Inc.,t972). especially pp 529-533.

Page 18

FEDERAL RESERVE BANK OF ST. LOUIS

net interest, and corporate profits (and adding in anallowance for depreciation, indirect business taxes,and other smaller items). The expenditure approachdetermines the current value of production basicallyby totaling all expenditures for final goods and serv-ices based on type of purchase and expenditure (plusthe net change in business inventory). Expendituresin the national income accounts are classified as per-sonal consumption expenditures, gross private domes-tic investment, government purchases of goods andservices, and net exports. The two approaches provideapproximately the same total, for expenditures onfinal goods and services provide income to the factorsof production which produced these items.

In general, nonmarketed goods, such as goods andservices produced and consumed by the household(which would include meals prepared in the homeand home repairs) are not included as part of the na-tion’s measured inconre. The exclusion of such produc-tive work performed by household members limits thevalidity of the GNP concept as a measure of thenation’s total product.3

It also should be stressed that not all market trans-actions are included in determining GN1’, for this

APRIL 1974

would involve double-counting. Final products are notnormally resold; intermediate products are resold insome form. For example, flour sold by a miller to abaker is resold in the form of bread. To count theflour sold by the miller and the bread sold by thebaker as part of GNP would involve double-countingthe value of the fiour.~

The concept of CNP then necessarily implies selec-tion of what one considers “productive activity”.In determining GNP, one must use some criteria ofproduction which are based on an implicit or explicitvalue judgment. To quote Simon Kuznets, a pioneerin developing national income accounting concepts:

if no criteria of social productivity are used,national income becomes a mechanical total of allnet receipts of individuals and business agencies,regardless for what activity or even \vhether there isany activity. It would include the compensation ofrobbers, murderers, dnsg peddlers, and smugglers,differential gains from the transfer of claims, andpure transfers such as gifts and contributions, which,in the absence of a productivity criterion, cannot bedistinguished from payments for services. Such ajudgmentless estimate would be of little use, since, tomeasure all market transactions, some gross rather

~At the end of an accounting period, any increase in the in-ventory of raw materials (or intermediate products) is in-cluded as part of the total product of that period. Double-counting can be avoided by totaling only the market value of“final” products, such as bread (plus an allowance forchanges in inventories), or by totaling the sum of the “valueadded” by all finns. Value added by a firm equals the marketrevenues received by the firm minus the cost of the raw mate-rials. In the above example the value added by the bakerwould he the revenues received through the sale of the breadminus the cost of the flour and other ingredients.

Page 19

Market transactions involving the exchange ofwealth or claims to wealth are also excluded in thedetermination of GNP. Exchange of stocks on thestock market and exchange of bonds in the securitiesmarkets only shift osvnership of claims to existing as-sets from one person to another. For the most part, thesale of a used car has a similar effect. In both cases, noincrease in production or productive capacity is di-rectly related to the exchange of these assets. Includedin GNP, however, are some of the dealer costs asso-ciated with these transactions. These costs include,among other things, the salaries and the commissionsof the stock and security brokers and used ear sales-men, since they provide a current service in the ex-change of existing assets. In determining what is in-cluded in GNP, the emphasis is on current economicactivities which are “productive” in the sense of creat-ing income. A sale of a new car would be included inGNP for this is an end item of current productiveactivity.

IA leading authority on national income determination, SimonKuznets, considered this problem when discussing issues in-volved in defining national income. He argued against theinclusion of nonmarket activities in general, but cautioned inthe interpretation of data which exclude such activities: “Thenational income estimator must choose between comprehen-sive definition — with the consequence that large sectors ofthe economy either cannot be measured on a continuous basisor cannot be included with mnore precisely measurable sec-ton because the errors are so enormous — and a narrowerdefinition that confines economic activities to those market-bound — for which tolerably reliable estimates can be made.In current national immcome measurement in this country, thedecision is usually in favor of the second alternative. And itfinds support in the argument that the activities so segregatedfor measurement are the ones subject primarily to economiccriteria and rationale; whereas those that are not directed atthe market are much more a part of life in general. One mayand does discharge a housekeeper for inefficiency in nanag-ing a household, hut by itself this is rarely a ground fordivorce

“The national income estimator cannot do much about suchomissions, since scarcity or lack of data is inherent in thenature of the omitted areas- But in interpreting national in-come movements in terms of satisfying consumers’ wants, thelimitation of national income largely to noncasual market-bound activities most be stressed. In this country as in manyothers where the market is always being extended, the rela-tive importance of the household as a source of consumergoods is declining. Many activities fonnerly perfonned by thehousewife or other members of the family and not measured(baking, sewing, camring, etc.) have progressively been takenover by business enterprises and gone into market-boundactivities; other household functions have vaoished withoutleaving a direct substitute in business activity. Hence, na-tional income totals tend to exaggerate the upward movementin the supply of goods to consumers, if such supply is cons-prehensively defined as coming from both market-bound andfamily activities.” [Simon Kuznets, National Income: A Sssm-mary of Findings (New York: National Bureau of EconomicResearch, Inc., 1946), pp. 124-125.]

FEDERAL RESERVE BANK OF ST. LOUIS APRIL 1974

than net total is requisite. It would measure neitherthe positive contribution of the country’s economicsystem to the needs of its members for purposes ofconsumption or capital formation nor the sum totalof what the inhabitants of the country think theirincome is.5

Kuznets favored a policy of making any underlying“scheme of values or social philosophy” explicit andallow it to guide the selection of the data.

The concept of “production” or “productive activity”in the measurement of national income has been givendifferent meanings by various writers and goveim-ments. In The Wealth of Nations, xvhich was firstpublished in 1776, Adam Smith distinguished produc-tive activities as the making of material goods only;all services, such as those provided by churchmen,lawyers. doctors, musicians, etc., were considered un-productive since “the work of all of them perishes inthe very instant of its production.””

Smith’s concept of productivity was perpetuated inthe writings of David Ricardo and John Stuart Milland formed the basis of the primary national incomeestimates in England and France for neaiiy a centusy.It was not until Alfred Marshall identified the produc-tion of goods and services with the creation of utilityin the latter part of the nineteenth century that esti-mators in these two countries returned to a broaderconcept of production.7 This broader concept in-eluded services as sn-eli as material commodities in themeasurement of output. Karl Marx accepted Smith’sdistinction, and consequently, the Soviet Union andother communist countries of Eastern Europe adopteda concept of national product that basically excludesall those services svhicls do not contribute to materialproduction,5

5Simon Koznets, National Income and Its Composition,1919-1938, Vol. 1 (New York: National Bureau of EconomicResearch, 194i), p. 4. In 1971, Kuznets received the NobelPrize in Economics, which was awarded, in part, for his workon developing measurements of national income.

6Adam Smith, The Wealth of Nations (New Yerk: TheModern Library, 19371, p. 315.

~Earlierestimators of national income had used a more com-prehensive production concept. See Intemational Encyclo-pedia of the Social Sciences, s-v. “National Income andProduct Accounts: Developments up to World War I.”

5See Moshe Yanovsky, Social Accounting Systems (Chicago:Aldine Publishing Company, 1965), ;sp. 112-115. Other as-pects of national income accounting in the Soviet Union arealso influenced by the writings of Marx. For example, follow-ing Marx’s theory of value, income is relater1 to only onefactor of production — social labor.

In the United States, studies on the measurementof national income appeared in the mid-nineteenthcentury, and the National Bureau of Economic Re-search published several studies in the l920s, Spurredby the economic depression and increasing govern-ment involvement in economic affairs, the Depart-uncut of Commnerce established a National Income Di-vision in the late 1930s which prepared estimates ofnational income data on an official basis. Official fig-ures of U.S. national income and product first ap-peared in the Survey of Current Business in 1942 andwere published in accounting fonn for the first timein 1947. Various revisions and refinements have beenmade since, but the basic structure of national incomeaccounting has not been altered greatly.’

National income or GNP in the United States todayis basically a measure of the market value of goods andservices produced during a given period of time.’°As two proponents of an indicator to mueasure eco-nomic svelfai-e, William Nordhaus and James Tobindo not question the usefulness of the GN1’ data as ameasure of production. They consider GNU data in-dispensable for shom-t-run stabilization policy and forassessing the economy’s long—run growth in productivecapacity. They- do question, however, the usefulnessof GNP data in evaluating tIme growth of economicwelfare.

Nordhaus and Tobin would like to see the develop-xnent of a new concept to measure the growth of eco-nomic welfare, and their argument for the develop-uncut of such a concept is as follows:

An obvious shortcoming of ON? is that it is an in-dex of production, not consumption. The goal of eco-nomic activity, after all, is consumption. Althoughthis is the central premise of economics, the profes-sion has been slow to develop, either conceptuallyor statistically, a measure of economic performanceoriented to consumption, broadly defined and care-fully calculated. We have constnmcted a primitiveand experimental ‘measure of economic welfare’(MEW), in which we attempt to allow for the more

°The present U.S - national income accounting systesis consistsof five interlocking accounts: National Income and ProductAccount, Personal Income and Outlay Account. CovernmentReceipts and Expenditures Account, Foreign Iransactions Ac-count, and Cross Savings and Investment Account.t0The major exceptions concerning production of goods andsen’ices which are not marketed hot included in the sneas-nreunent of CNP are estimates of food produced and con-sumed on farms, financial services of commercial banks andother financial intenrieci iari s- and the rental value of owner—occupied houses -

Page 20

FEDERAL RESERVE BANK or ST. LOUIS APRIL 1974

obvious discrepancies behveen GNPand economic welfare.11

To construct their measure of wel-fare or consumption, Nordhaus andTobin make several modifications tothe existing national income accounts.These modifications fall into three gen-eral categories: 1) reclassification ofGNP expenditures as consumption, in-vestment, and intermediate; 2) imputa-tion for the services of consumercapital, leisure, and household activ-ides; and 3) correction for some ofthe disamenities of urbanization andindustrialization.’2

hji.Si:aflWf4i>C

These modifications are shown inTable I. In essence, this table providesvarious additions and subtractions togross national product, or net nationalproduct, to arrive at what is labeled sustainableMEW.13

Cciit~’l Ccz~utnniion Sustainable MEW as ameasure of consumption is somewhat similar to theconcept of net national product (NNP) as a measureof production. Part of the output included in GNP willbe used to repair and replace the existing stock ofcapital goods. This portion of output is classified as thecapital consumption allowance. The subtraction of thecapital consumption allowance from GNP gives NNI’.NNP tells us how much current income or productioncan be consumed consistent with the maintenance ofproductive capacity or income potential.

In a similar manner, the Nordhaus-Tobin conceptof sustainable MEW provides a measure of “theamount of consumption in any year that is consistentwith sustained steady growth in per capita consump-tion at the trend rate of technological progress.” Thesustainable MEW concept then considers not only theamount of capital which must be replaced in a periodto maintain consumption at the existing level, but alsohow much additional investment or abstention fromconsumption in the current period must be made inorder to keep consumption per capita growing at some

‘1Nordhaus and Tobin, “Is Growth Obsolete?”, p. 4.1mIbid., l~-5.

‘3Another concept, labeled actual MEW, consists only of totalconsumption for a given period and does not take into ac-connt any investment expenditures,

rate which is based on technological progress.’4 Afterestimates for both the capital consumption allowanceand the growth requirement are made, these esti-mates are subtracted from GNP.

Ie~eieo~nnl iit~&’n~UmthGc-~-Some output,classified as final output for GNP purposes, is reclassi-fied as regrettables and intermediates by Nordhausand Tobin and is excluded from MEW.

By intermediate product, Nordhaus and Tobinmean “goods and services whose contributions to pre-sent or future consumer welfare are completelycounted in the values of other goods and services”;’5

they are “not directly sources of utility themselves butare regrettably necessary inputs to activities that mayyield utiity.”’~ Regrettables represent expendituresfor national security, prestige, or diplomacy, which inthe judgement of Nordhaus and Tobin, do not directlyincrease the economic welfare of households. No sharpdividing line exists between what is classified as in-termediates or regrettables.

Some private expenditures and some Governmentexpenditures are reclassified as intermediate products

“Sustainable MEW omits capital expenditures required tomaintain the capital-output ratio. According to the authors,“It allows for capital depreciation, for equipping new mem-bers of the labor force, and for increasing capital per workerat the trend rate of productivity change.” See Nordhaus andTobin, ‘Is Growth Obsolete?”, pp. 24-25.

iSIbid., p. 5.

~°Ibid.,p. 7.

Table I

GROSS NATIONAL PRODUCT AND MEASURE OPECONOMIC WELFARE (MEW): 1929 AND 1965

Billion, oF DollarS, 1958 Prices I1929 1965

Gross National Product $203.6 S 617.8Less: capital consumption, NIPAI 20.0 - 54.7

Net National Product, NIPA 183.6 563.1

Less: NIPA finol output reclassified asreqrettables and intermediates

a) Government 6.? 63.2

b) Private 10.3 309Imputottons to’ items not included in NIPAPlus, a) Leisure 339.5 626 9

hI Nonmarket activity 87.5 295.4c) Services of public and private capital 29.7 78.9

Less: d) Disamenities 12.5 - 34.6Less, Additional capital consumptio.i 19.3 92.7Less, Growth equirenlenl 46.1 101.8

Sustoinoble MEWS 543 6 1241.1

‘x II~~ii:,:. t.. b.tu.ijs.itl l,is.-,-snc- its..,! 1:’.,d.,,t ,ts.’L’,,:iiit,in: i,. eli: ‘a, hint. B

- ~ViilisssnN,,i-ilh,s,n. uses j:un,-s ‘I ,sbss.. ‘Ia lirosesh Oh,.,,!, . Kru,.us,,,r firo,,-th.1- UtaIts .%nnn, rsns~ ~.,Uuuussssn.\.,sl 5 Nt-v. is,’ : Nati,,sml H:srossu,-li_nun-ns c Mt-st-a’. cli, I a72 j . i.,

Page 21

FEDERAL RESERVE BANK OF ST. LOUIS APRIL 1974

or regrettables. Private expenditures, such as personalbusiness expenses and a part of transportation expen-ditures in the GNP accounts, would be reclassified asintermediate products. A major portion of Govern-ment purchases, such as national defense, spaceresearch and technology, international affairs andfinance, veterans benefits, general government, andcivilian safety (police, fire, and correction) are reclassi-fied as regrettables or intermediate products and sub-tracted from GNP.’7

The authors impute an estimate formany activities which they feel have a positive ornegative effect on social welfare but are not consid-ered in the determination of GNP. Specifically, im-putations are made for leisure, nonmarket activity,disamenities, and services of public and privatecapital.

The most substantial modifications to GNP in ob-taining a measure for sustainable MEW are the resultof the imputations for leisure and nonrnarket activity.Leisure is important to a welfare index, for welfarecould rise (consumption of leisure) while GNP fallsif employees voluntarily decide to work less. An esti-mate for nonmarket activity or household productionand consumption, such as meals, cleaning, and homerepairs, is also added to GNP to obtain MEW.

An estimate for the disamenities of urbanization issubtracted from the GNP data in determining MEW.This estimate considers social costs which are not in-eluded in the costs of producing consumption goodsand services.18 These costs would include pollution,litter, congestion, noise, and insecurity. The estimate

~~Ibid,,p. 7. Kuznets supported the notion that many gov-ernment services should be treated as intermediate goodsrather than final product. He argued that services to busi-nesses such as economic legislation and the maintenance ofinternal and external security is not a direct service to con-smners but a cost of maintaining society at large: “acondition of economic production rather than an activitydirectly yielding final economic goods.” He supports theexclusion of these government activities from a country’soutput by emphasizing that the total which is sought is “thatof product, of end-result of activity — not of the volume ofactivity itself.” See Simon Kuznets, “Discussion of the NewDepartment of Commerce Income Series: National Income:A New Version,” The Review of Economics and Statistics(August 1948), pp. 156-157. Also see MartinJ. Bailey, “Ap-pendix: The Concept of Income,” National Income and thePrice Level, 2nd ed. (New York: McGraw-Hill Book Com-pany, 1971), pp. 272-274. For an argument against suchexclusion see Milton Gilbert, George Jaszi, Edward F. Deni-son, and Charles F, Schwartz, “Obiectives of National In-come Measurement: A Reply to Professor Kuznets.” TheReview of Economics and Statistics (August 1948), pp183- 189.

~For an economic analysis concerning problems of social costs,see H. H, Coase, “The Problem of Social Cost,” TheJournal of Law and Economics (October 1960), pp. 1-44-

of these costs is based on the income differentials be-tween large cities and smaller towns and rural areas.Assuming that people can choose residential locations,a portion of the observed income differential can beconsidered a “disamenity premium” which compen-sates individuals for unpleasantness associated withliving in urban areas,’°

Services of public and private capital is the lastcategory of imputations for items not included in GNP.The only imputation made for the services of capitalin determining GNP is the addition of an estimate forthe services received from owner-occupied housing.The MEW concept would extend imputations fromcapital to include services from Government struc-tures (excluding military) and services from con-sumer durable goods (under the MEW concept, con-sumer durables are reclassified as investment goodsrather than consumption).

Nordhaus and Tobin state that they are after ameasure of consumption, “broadly defined and care-fully calculated,” but then label this measure a“measure of economic welfare,” However, consump-tion and welfare are two different (although related)concepts. Welfare would depend on the amount oftotal satisfaction one receives from total consumption,and, among other things, would depend also on thedistribution of income. Nordhaus and Tobin realizethe problelns involved in trying to measure welfareand state that they “cannot ... estimate how well in-dividual and collective happiness are correlated withconsumption,”2° In a comment on the Nordhaus-Tobin MEW concept, Robin C. 0. Matthews pointsout that debates in the 1940s recognized such distinc-tions between consumption and welfare, and arguesthat the MEW concept is a measure of consumptionnot a measure of welfare.21

Obtaining reliable estimates of various economicactivities which are not included in the national in-come accounts poses a serious problem in computingMEW, The problems involved in obtaining an accu-

‘°According to the authors, the disamenity premium wasabout 8 percent of average family disposable income in1965. Since income differentials have tended to inducemigration to urban areas, only a portion of the estimatedincome differential is subtracted from the CNP accounts asa disamenity premium,

20Nordhaus and Tobin, “Is Growth Obsolete?”, p. 25.

2lflobin C. 0. Matthews, “Discussion,” Economic Growth,Fiftieth Anniversary Colloquium, Vol. 5 (New York: Na-tional Bureau of Economic Research, 1972), p. 91.

Page 22

FEDERAL RESERVE BANK OF ST LOUIS APRIL 1974

rate measure of household activities is one of the rea-sons why such activities are not included in measuredGNP. The authors of MEW recognize this problemand attempt to estimate the reliability of various com-ponents of MEW.

Nordhaus and Tobin rank the reliability of thecomponents of MEW as having a low error, mediumerror, high error, or very high error, Data in the na-tional income accounts, such as GNP, are used as abenchmark in determining reliability and are put inthe low error category. Components in the very higherror category are judged to have about ten timesthe percentage error of GNP. The imputations forleisure, nonmarket activities, and disamenity fall intothis very high error category. The imputations forthese activities, however, account for much of thedifference between GNP and MEW.

The imputations for leisure and household activities,in terms of constant prices, vary greatly depending onhow current price estimates are deflated. The authorsobtained constant price estimates of both activities bydeflating current prices by a consumption deflator andby deflating by wage rates. The accompanying chartpresents three different growth paths of sustainableMEW which depend on how leisure and nonmarketactivities are deflated. The authors indicate a pref-erence for variant B which deflates leisure by a wageindex and which deflates nonmarket activity by aconsumption deflator.22

All three variants of MEW show a positive rate ofgrowth over time, which indicates that real consump-tion per capita has been increasing. According tovariant C, per capita sustainable MEW grew at a 3.6percent average compound annual rate from 1929 to1965, which is slightly faster than the 3.1 percent ratefor NNP. In the same time period, variant B grew ata 2,3 percent rate and variant A at a 1.8 percent rate.

Nordhaus and Tobin have provided an estimate ofsustainable consumption over time, After allowing forsome of the disamenities of modern production tech-niques and urban congestion, their estimates showthat net consumption has been growing, but probablyat a slower rate than total measured output. Theyrecognize that many unsolved problems are posed bytheir MEW concept, but view the measure as an at-tempt to obtain an indicator of the growth in eco-nomic welfare. Perhaps the intent and conclusions of

~2

Variant A deflates both leisure and nonnlarket activity by awage index, and variant C deflates both activities by aconsumption deflator-

their study can best be summed up by the authorsthemselves:

We recognize that our proposal is controversial onconceptual and theoretical grounds and that many ofthe numerical expedients in its execution are dubi-ous. Nevertheless, the challenge to economists toproduce relevant welfare-oriented measures seemscompelling enough to justify some risk-taking. Wehope that others will be challenged, or provoked, totackle the problem with different assumptions, morerefined procedures, and better data, We hope alsothat further investigations will be concerned withthe distribution, as well as the mean value, of ameasure of economic welfare, an aspect we have notbeen able to consider,23

23Nordhaus and Tobin, “Is Growth Obsolete?’, p. 26.

Per Capita Net National Productand Per Capita Sustainable

Measure of Economic Welfare (MEW)1929-1965

1958 I‘ric~nI

PatIo StoleIhoasoad, of Dollarsa

Ratio ScaleThoasoads of Dollors

a

7

6

S

4

3

2

1925 30 35 40 45 50 55 60 19655,,’,,, William Nardha,, and Jome, Tot]’. In maclb Ob,okM’. E’o,omi,

G,a’e,h, #ittiett, A,,],, ra’ry to]loqvim, Vol.5 Ne, York, National B~,ea,at Eta,,,]’ Re,ea,ct,, l9?2ipp. 52, 57~

No*~Data on MEW or, a,aiI,ble only a, ‘he yea” 929. 35, 45, 47, 54, 58, and‘65 For ‘amporabilit’, Ike NNP series in planed only for 1k,,, yeans. eec,tho,gh thin ned,, I, avoi]able on a “eli’,,,, ban],,,,, thin period.

Page 23

SUMMARY

GNP, in general, provides a measure of currentproduction in the United States for which a moneyincome has been received. As a production or incomemeasure, it has been used in forming policy goals foroutput, employment, and price level changes.

Criticism has been directed against the GNP con-cept as emphasizing the quantity of goods and serv-ices produced while ignoring what is happening to the“quality” of life. Proponents of a new measure to ob-tain such a quality indicator, or an economic welfaremeasure, would modify the present national incomeaccounts in several ways. They would subtract fromGNP any costs incurred in maintaining clean air andwater in the production of goods and services, sub-tract an estimated cost of urban congestion, subtractexpenditures for police and national defense, and in-clude an estimate for the value of nonmarket activi-ties such as household activities and leisure time.

In obtaining their measurement, however, Nordhausand Tobin have had to resort to a number of crudeestimates and rely considerably on their own valuejudgments concerning the classification of goods andservices as consumption or intermediate product. Inparticular, Nordhaus and Tobin make estimates of

many activities, such as household activities, thatofficial national income estimators have avoided be-cause of lack of data. Their judgments concerningwhat is or is not an intermediate product also play asignificant part in this measure. The problem of de-fining intermediate products is not unique to tlusmeasure of consumption, however, as the problem ofdefining intermediate goods remains an unsettled is-sue among various national income estimators. Forexample, Simon Kuznets supports the notion thatmany governmental services should be treated as in-termediate products rather than final pi’oducts in thenational income accounts,

Although the proponents of this new concept referto it as a measure of welfare, it more accurately pro-vides a broad measure of consumption. A welfaremeasure would quantify the amount of satisfaction orutility received from consumption and would depend,in part, on the distribution of income. As a measure ofconsumption, however, the MEW concept attempts toprovide an indicator more closely associated with theconcept of welfare than that provided by a productionmeasure such as GNP. Unfortunately, in view of thehigh error content associated with this measure, theproposed MEW concept does little more than breakthe ice in an attempt to provide an accurate estimateof economic welfare.