the economic value of human life

To establish the economic value of a human life, lifetime earningsdiscounted at a 4 per cent rate are presented by age, sex, color, andeducation. These estimates are intended for use by economists,program planners, and others. Various specific findingsare reported.

THE ECONOMIC VALUE OF HUMAN LIFE

Dorothy P. Rice, F.A.P.H.A., and Barbara S. Cooper

IntroductionTHE value of human life expressed in

terms of lifetime earnings is a basictool of the economist, program planner,government administrator, and otherswho are interested in measuring the so-cial benefits associated with investmentsin particular programs. For public pro-grams, such as the control and eradica-tion of disease, highway construction,accident control, education, vocationalrehabilitation, welfare, housing, andflood control, the valuation of humanlives is a basic requirement for theproper calculation of the benefits to bederived. The recent emphasis on cost-benefit analysis in all of these areas re-quires that adequate tools be providedfor analysis. Like the carpenter, whosework is generally facilitated and productimproved by the availability of good ma-terials and equipment, the economistmust be equipped with the tools of histrade in this case, basic data for valua-tion of human life.The aim of this report is to provide

improved, refined, comprehensive, andup-to-date estimates of the present valueof lifetime earnings in considerable de-tail according to age, sex, color, andeducational level. Presentation of thedata in this form will enable the econo-

mist to choose the most appropriate se-ries of data for program evaluation. Forexample, a health program to reducemortality in a specific age and colorgroup can be evaluated by use of thelifetime earnings data developed for thatgroup. Likewise, basic data are providedfor measuring the benefits from invest-ment in various educational programs.The data presented are limited to the

quantification of the value of human lifein terms of lifetime earnings. These areby no means the only measures of thevalue of human life. Schelling notes thatvaluation may be in terms of the worthof one's life to oneself or to whoever willpay to prolong it, and the amount willvary accordingly.' However, in this re-port the value of a person is defined interms of his economic worth as a pro-ductive member of society and theamount will vary according to age, sex,color, and degree of educational attain-ment.

Historical SummaryQuantification of the human life

values in economic terms is not a newconcept. In their "Money Value of aMan," Dublin and Lotka traced thisprocedure from the valuation of slavelabor in ancient times through 'the vari-

VOL. 57. NO. 11, A.J.P.H.1 954

ECONOMIC VALUE OF HUMAN LIFE

ous estimates made by Sir William Pettyin the 17th century and Adam Smith acentury later, on up to the time of theirfirst edition, in 1930, and later, to 1946,when their revised edition was pub-lished.2 In the revised volume, theseauthors defined the money value of manas the present value of his net futureearnings, i.e., gross future earnings lessthat part which he consumes or spendson himself. Detailed data were presentedfor ages 21 through 64 according to spec-ified annual levels of earnings in 1934,employing a 2.5 per cent discount rate.The human life value concept has also

been applied commercially in the fieldof life and health insurance. In 1927,Huebner, in his "Economics of Life In-surance,"3 stated his thesis that optimumlife insurance protection should equalthe capitalized monetary worth of theindividual's earning capacity. This ideaof employing potential lifetime earningsas a measure of adequate life and healthinsurance protection is now used exten-sively in the field.

Lifetime Earnings in Health Studies

The cost-benefit studies in the healthfield during the past few years includevarious estimates of the value of humanlife. Some are based on incomes; someon earnings; some impute a value tohousewives; some account for consump-tion; and various discount rates are em-ployed. Prest and Turvey point out intheir comprehensive survey of cost-bene-fit analysis: "Some of the differences be-tween authors in the way they estimatebenefits stem from differences in theavailability of statistics rather than fromdifferences in what the authors wouldlike to measure if they could."4The following passages briefly exam-

ine the different basic assumptions, tech-nics, and data employed by varioushealth researchers in their presentationof earnings foregone.

In assessing the cost of mental illness

in 1958, Fein presented estimates of thepresent value of future expected incomefor ten-year age groups at various dis-count rates ranging from 2 to 5 percent. AIedian income in 1952 was usedfor males and females, ignoring the eco-nomic value of housewives.5

Weisbrod, in "Economics of PublicHealth," used fundamentally the sameformula as Dublin and Lotka for cal-culating the present value of net futureearnings. However, there were some dif-ferences in the meaning of consumptionand the earnings components of theformulas.6 Weisbrod developed the con-cept of marginal consumption the addi-tional consumption associated with anadditional person-and imputed a valuefor nonmarket household services of fe-males in terms of units of family re-sponsibility. Data were presented forthe present values of net future earningsby single years of age and by sex,using cross-sectional 1949 earnings andtwo discount rates-4 and 10 per cent.These data have been used recently byseveral economists in the development ofcosts of illnesses.7Klarman calculated the present value

of future earnings for syphilis cases andfor those who died of cardiovasculardiseases. For syphilis, the calculationwas based on 1961 average earnings bysex and color, with no adjustments forage and the value of household services.Klarman used a net discount rate of ap-proximately 2 per cent, having adjustedfor productivity increases.8 In the laterwork on cardiovascular diseases, thepresent value of lifetime earnings wascalculated for ten-year age groups onthe basis of 1962 average earnings foremployed males and females, and sepa-rate calculations were made for the valueof housewives' services for each ageclass. In this case, a 4 per cent discountrate was used.9

In the recent study of one of theauthors, "Estimating the Cost of Illness,"lifetime earnings were presented for

NOVEMBER, 1967 1 955

males and females at five-year age inter-vals, based on full-time, year-roundearnings in 1963 and taking into ac-count an imputed value for housewives'services. Two discount rates were em-ployed-4, and 6 per cent.10

Lifetime Earnings in Education Studies

Lifetime earnings have been used inmany studies relating, to the benefits ofeducation." As in the health studies,various estimates have been made. Someemployed annual data, others, lifetimeearnings; some were based on earningsand others on income; some presentedseveral discount rates, others applied nodiscounting at all. All the studies wereconfined to males and the income orearnings data were for all men whoworked, including those employed parttime. The following summarizes thebasic data and technics used in severalof these studies.

In his 1960 study, Mliller presentedaggregate lifetime earnings by years ofschool completed by males for selectedyears from 1939 to 1958. Included weredata for two ages-18 and 25. The esti-mates were based on mean annual earn-ings and no discounting was employed.'2In his testimony on equal employmentopportunity before the Senate Commit-tee on Labor and Public Welfare in1963, Miller presented estimates of ag-gregate lifetime earnings for males 18years of age, by level of education,color, and region, and for selected occu-pations, using the same bases as in hisprevious study and 1959 earnings data.13

Houthakker presented estimates oflifetime income at age 14 by years ofschool completed, based on estimated1949 mean income, and employing threediscount rates: 3, 6, and 8 per cent.14Bridgman presented aggregate life-

time incomes for males aged 25, by twolevels of schooling completed (highschool and college). These estimateswvere based on 1949 and 1956 mean and

median income anid discounting was notemployed.15

In his article on the "Valuation ofHuman Capital," Weisbrod presentedessentially the same data as were de-veloped for his study on the "Economicsof Health." The data here were pre-sented for males at five-year age in-tervals, based on 1950 earnings and dis-counted at 4 and 10 per cent.'6

Recently Weisbrod estimated the dif-ferential in lifetime incomes betweenhigh school dropouts and graduates. Heused 1949 median incomes and calcu-lated the present value of the differen-tials at ages 16 and 18 according to sex,color, and major geographic region, em-ploying 5 and 10 per cent discountrates.'7

Basic Assumptions

Despite the fact that estimates of life-time earnings have been developed bythese many researchers, all were devisedfor a specific use and are not readilyadaptable for other purposes. The econ-omist engaged in analysis of programsrequiring lifetime earnings data musttherefore develop his own set of esti-mates to meet his needs. Furthermore,the available estimates are not current,having been based on obsolete income orearnings data. Finally, these estimatesare generally limited to males, becauseof the problem associated with measur-ing lifetime earnings for females, mostof whom move in and out of the labormarket while spending a considerableportion of their productive lives ashousewives.The estimating procedure for the de-

velopment of the lifetime earnings aspresented here is described in detail inthe study by Rice, "Estimating the Costof Illness." The method developed takesinto account life expectancy for differentage, sex, and color groups, varyinglabor force participation rates, the cur-rent changing pattern of earnings at

VOL. 57. NO. 11. A.J.P.H.1956


successive ages, imputed value of house-wives' services. and the discount rate.The following is a brief review of thebasic assumptions and economic con-cepts employed.

Life Expectancy

The lifetime earnings data were de-veloped on the assumption that each co-hort will follow his or her pattern of lifeexpectancy as reported for 1964 at suic-cessive ages. The National Center forHealth Statistics publishes life tables bysex and color.18 Cohort data wvere ob-tained for four groups: white and non-white males and white and nonwhitefemales. No adjustments were madefor variations in life expectancy byeducational level. The assumption isthat life expectancy at any given age isthe same regardless of educational at-tainment.

Labor Force ParticipationThe estimate of lifetime earnings takes

into account varying labor force par-ticipation rates at different ages. Theassumption is that an individual wvill bein the labor force and productive duringhis expected lifetime in accordance withthe current pattern of labor force par-ticipation for his sex, color, and educa-tional level. For this calculation. the Bu-reau of Labor Statistics provided un-published data for 1965 on the numberof employed persons by age. sex, color,and years of school completed. Use ofthe number employed in 1965 assumesconditions of relatively high employ-ment (approximately 96 per cenit of thelabor force employed).

EarningsThe appropriate measure of output

for individuals is year-round. full-time earnings and the proper measure ofexpected earnings is the arithmetic aver-age or mean. In many of the stuidiescited above. median earnings (or in-come) for all wage earnings were used

because they are generally available inpublished form. For this report, full-time, year-round mean earnings esti-mated from mean incomes were em-ployed. Unpublished data from the Biu-reau of the Census provided the meanearnings for all male workers at five-year age intervals. The equivalent earn-ings data for females and by color weredeveloped from 196-1 mean income data.The estimated mean earnings data wereadjusted upward to take cognizance ofwage supplements-employer contribu-tions for social insurance and privatepension and welfare funds-on the as-sumption that these should be includedas a measure of total output.The only data presently available on

earnings or income by educational levelare from the 1960 census. For each ageclass, the ratio of median income at eachlevel of school completed to the medianfigure for the entire age group wN-asapplied to the 1964 estimated year-round, full-time mean earnings to obtainfinal figures for the program s input foreach age group.

In applying these cross-section surveydata to the estimates of lifetime earnings.it is assumed that the future pattern ofearningrs for an average individualwithin a particular color and eduica-tional attainment group will remain thesame as that estimated for the base year,1964. This model recornizes that theaverage individual may expect his ownearnings to rise as he ages and gainsexperience in accordance with the cross-section survey data for 196-4.

Housewives' Services

As indicated in the brief review of theliterature in the field, very few re-searchers have developed lifetime earn-ings data for women, owing to the dif-ficulties associated with measurinr theirearnings when they move in and out ofthe labor market. To omit the value ofservices of housewives, however. wouldseriously underestimate the value of the

NOVEMBER, 1967 1 957

lifetime productive contribution of fe-males.

Housewives' services are estimated inthis report at the average earnings rateof a domestic worker-$2,767 in 1964.This imputed value is clearly on the lowside for it makes no allowance for thehousewife's longer work week or thesize of the household cared for.

Rates for women keeping house wereavailable for 1964 by age and color fromunpublished Bureau of Labor Statisticsdata, but not by educational level. Theestimating procedure involves the appli-cation of the available rates to the num-ber of females expected to be out of thelabor force in each age cohort, andrecognizes that the keeping-house ratevaries with age. The use of the samerates for each educational level assumesthat the keeping-house rate does notvary by degree of educational attain-ment. This is clearly an erroneous as-sumption, but lack of data prevented useof an alternative procedure. However, inpractice, when the rates were applied tothe varying number of women at eacheducational level estimated to be out ofthe labor force, the numbers did changefor each cohort. The effect of the ap-plication of a uniform value to womenkeeping house will be discussed later.

Discounting

The value of money changes withtime so that in order to calculate thepresent monetary value of man, his fu-ture expected earnings must be con-verted to their worth today. Banks andKotz state that "a given sum is normallyworth more today than an equal sum atsome future date, because the money (orresources) can be profitably invested (orconsumed) in the interval between to-day and the future. Interest is the pre-mium paid to reflect the fact that anygiven sum or resources could be putto profitable uses over a period of time.. . .It follows that the value of moneywhich is not currently available, butwhich will become available (or spent)

some years hence must be discountedfor the interest which could be earnedin the interim, which is why the presentvalue of a dollar to be received in thefuture is always less than 100 cents."19

For the valuation of public healthor recreational programs, where thebudgets are measured in terms of dollarvalues of expected lifetime earnings, theaggregate earnings must be discountedat an appropriate rate of interest to de-termine their value today.

Although there is general agreementamong economists that discountingshould be employed, there is no agree-ment on which discount rate, i.e., rateof interest, to use. Yet the selection ofthe discount rate is most importantsince its effect is considerable. Thehigher the discount rate, the lower thepresent value of future earnings. Witha high rate of discount, earnings far inthe future yield only a small presentvalue. Conversely, lowering the discountrate increases the present value of theseearnings far in the future.The selection of an appropriate dis-

count rate in the valuation of publicprograms is especially important as theuse of too low a rate for discountingfuture benefits may lead to uneconomicadoption of projects. Banks and Kotzdeplore the lack of a central governmen-tal authority to prescribe the discountrate to be used despite its importancein arriving at public investment deci-sions.The authors have prepared data on

the present value of lifetime earningsemploying four discount rates: 2, 4, 6,and 8 per cent. Only the data based onthe 4 per cent discount rate are pre-sented here; the data for the other threemay be obtained from the authors.

Productivity Increases

While future earnings must be dis-counted to reflect lost interest, averageannual earnings must be increased to re-flect gains in productivity. It is anunderstatement of lifetime earnings to

VOL. 57. NO. 1 1. A.J.P.H.1 958


assume that a person ten years fromnow will earn the same amount as aperson of the same age, sex, color. andeducational level earns today. In orderto adjust for the gain in productivity.an average annual gain can be projectedand applied to the annual earnings. Thisrate of increase may be incorporatedinto the discounting calculations to ob-tain a net effective discount rate. Forexample, assuming a rise in productivityof 3 per cent a year.20 a discount rateof approximately 7 per cent will be re-duced to a rate of approximately 4 percent (1.07 / 1.03 =1.039). the rate usedin this report.

Allowance for ConsumptionThere is a diversity of opinion re-

garding the treatmeint of consumption.Insurance companies treat consumptionas a deduction from a person's contribu-tion to output. Dublin and Lotka andWeisbrod deduct consumption from totaloutput in their calculations of the earn-inr losses.21

Fein and Klarman, on the other hand.make no such adjustment. Fein sum-marizes his views as follows:

"Certainly the net figure (gross valtue lessconsuimption) derived by Dublin anid Lotkato indicate the imioney value of a mnan to hisfamily is correct for their purposes. It is notat all apparent, however. that the net coinceptis the correct one w-hen w%ve deal wvith theeconomic value of a man to society. It istrue that man coinsumes partly in order tomaintain himself, and in this seinse some ofhis consumption may be considered as a grossinvestment to take care of depreciation: itis also true, howvever, that constumption is anend in itself anid can be viewved as a final,rather than an intermediate, step in the crea-tion of other products. The question involvedconcerns the purposes for which an economyvexists." On anl individual's incoine, "the in-dividtial enjoys life, and it is for this pur-pose that the social econiomy exists."22

In accordance with the above -iewpointand because we are measuring the eco-nomic value of man to society and notto his family. no allowance for consump-tion is made.

Findings

Data are presented on the present

value of lifetime earnings of males andfemales by age, color, and years ofschool completed, discounted at 4 per

cent. Table 1 presents the data by sex

and color, and Table 2 presents the datafor males and females by color and years

of school completed. Data for three levelsof education are presented: completionof 8 years. 12 years. and 16 or more

years of school. (Data may be obtainedfrom the authors for three additionallevels of education-completion of lessthan 8 years, 9-11 years, and 13-15 years

of school.)The following highlights some of the

relationships and differentials found inlifetime earnings by age, sex, color, andeducational level.

Pattern of Lifetime Earnings

The value of lifetime earnings variesat different ages. For men. discountedexpected lifetime earnings increaserapidly and sharply, peaking in theyoung adult years ages 25-29-andthen decreasing at an even faster ratebleginning in middle age. Without dis-counting, the aggregate lifetime earningswould be highest at the youngest age

groups, decreasing with age. The dis-counting procedure involves the appli-cation of increasingly higher rates toeach year's earnings. At the under-one-year age group, in which a person is notconsidered productive until at least age

14, his first year's earnings are dimin-ished by the rate for the 13th year. The13th year of discounting at the rate of4 per cent yields a figure of 60 per cent,reducing that first year's earnings bytwo-fifths.At the young adult ages, many years

of expected earnings are also discountedheavily, but the earnings are countedimmediately and these earnings over a

long work span result in peak earningsat these age groups. At age 20. whenmany persons have already begun work-

NOVEMBER. 1967 1 959

Table I-Present value of lifetime earnings, by age, sex, and color, discounted at 4per cent, 1964

Age Males Females(in years) All White Nonwhite All White Nonwvhite

Under 1 $ 58,879 $ 61,810 $34,228 $36,234 $37,166 $28,6551-4 64,733 67,815 38,058 39,697 40,646 31,6975-9 79,026 82,769 46,527 48,417 49,563 38,70710-14 96,336 100,882 56,769 58,996 60.386 47,19415-19 114,608 119,921 67,546 68,269 69,890 54,549

20-24 128,268 134,191 74,581 71,639 73,419 56,89225-29 131,416 136,121 73,767 70,272 71,789 54,77430-34 125,499 129,623 68.338 67,772 69,201 51.88435-39 114,379 118,787 61,499 64,087 65,694 48,03840-44 99,303 102,614 52,978 59,536 61,074 43,044

45-49 81,224 84,901 42,799 53,633 55,386 38,28850-54 62,502 65,581 30,507 46,962 48.444 34,28555-59 44,667 46,381 20,005 40,187 41,102 29,72360-64 26,013 27,213 12.263 32,732 33,362 25,00965-69 13,530 14,023 6,887 26,038 26,367 21,149

70-74 11,005 11,346 5,971 20,628 20,801 17,68275-79 8,207 8,436 4,608 14,722 14,803 13,08080-84 4,634 4,775 2,513 7,751 7.797 6,83485 and over 751 776 396 1,241 1,251 1,067

ing, first-year earnings are counted atfull value, the next year at 96.2 percent, and so on. This individual is age34 before the factor reaches 60 per cent.Older persons have relatively few re-maining years of working life and usu-ally at lower earnings levels. Althoughthese years are counted at more nearlytheir full value, the present economicworth of these individuals is substan-tially less than that of young adults.

Male and Female DifferentialsUnder this discounting procedure,

what is the economic value of humanlife? For males, the value of discountedlifetime earnings ranges from about$59,000 for infants, to a peak of ap-proximately $131,000 at ages 25-29, anddecreases to $750 at ages 85 and over(Table 1).Female earnings reveal an entirely

different pattern. The level is lower, therange narrower, and the decline frompeaks more gradual. The discountedearnings range from about $36,000 forinfants, to nearly $72,000 at ages 20-24,and down to about $1,200 at ages 85

and over. The slower rate of descentin earnings for females is the result ofattaching a value to housewives' serv-ices. As women age and leave the laborforce, they are still credited with a dol-lar value if they stay home and keephouse, which prevents their earningsfrom falling as rapidly as those of menwho have left the labor force.

It is not surprising to learn that malesgenerally have higher lifetime earningsthan do females. In the young adultyears, the discrepancy is quite large.Lifetime earnings of males at ages 20-24are nearly four-fifths greater than thoseof females of the same age. However, inthe older years, when many men havestopped working and women (who havea longer life expectancy) are keepinghouse, the relationship reverses. At ages60-64, the expected lifetime earningsfor females are considerably greaterthan those for males.

Educational Differentials Among MalesNumerous studies have shown that

education pays off economically-per-sons with more schooling earn more

VOL. 57, NO. 11. A.J.P.H.1960


Table 2-Present value of lifetime earnings, by age, color, sex, and years of schoolcompleted, discounted at 4 per cent, 1964

Years of school completedMlales Females

16or 16or8 years 12 years more years 8 years 12 years more years

$ 52,34157,42770.09085,428

102,036

112,820112,928108,674101.13289,568

75,90559,79743,00225,04312,714

10,2877,6484,330703

36.04440,07848,99759,78271,545

79.51479,46175,24469.81863,007

53,18339,32927,07617,24110,361

8,9836,9323,780596

$ 73,36280,48998.237119.736139,541

150,454151,086145,602135.726120.477

102.86482,61061,09137,54221.071

17,04912.6767,1751,166

46,80352,04063.62177.62691,200

99,516100,31196,51790,16280,612

68,18952,05637,38324,85616,003

13.87510,7085,839920

$ 99,074108,701132,669161,702187,999

209,265221,738223,471215,539195,328

170,300139.643106,93971,06345,077

36,47221,11715,3502,494

57,07463,46077,58394,661110,160

123,069127,975122,480114.789104,518

88,91166,12046,36231,25420,537

17,80613.7417,4931,181

831,08934,00041,45950.51359,981

64.88164.06462,20759,28055,025

50.04444,25338,11131,62725,468

20,09114.298,.5311,208

26,72729,56536.10344,01951,597

54.41752,85651,26748.57244,515

40,18635.90130.93825.66821,441

17,92613,2616.9291,082

$45,12649,35160,17973,32081,340

80,56678,16175,93872,84368,697

63,30156,42047,97937,52828.695

22,63716,1108,4851,361

39,42743,61353,25864,93572,868

73,57172,18270,74667,23761,097

54,45447,68339,44429,88422,856

19,10914,1367,3861,154

$ 59,14664.68478,87596,099

108,928

112.327112.351112.023111,162109,707

104,85895.70880.98358,70841,612

32,82823.36212.3041,974

53,18858,83471,84587,59899,948

105.372106.7882105,698101,78694.296

84,22471.37355.93537,84525,262

21,12115,6248,1641,275

Age(in years)and color

W'hiteUnder 11-45-910-1415-19

20-2425-2930-3435-3940-44

45-4950-5455-5960-6465-69

70-7475-7980-8485 and over

NonwhiteUnder 11-45-910-1415-19

20-2425-2930-3435-3940-44

45-4950-5455-5960-6465-69

70-7475-7980-8485 and over

NOVEMBER. 1967 1961

money over their lifetimes. However, itis the variations in extent of the pay-off according to educational attainment,color, and sex that are most revealing.At each age level within each sex andcolor group, additional schooling is as-sociated with substantial increases inlifetime earnings.The lifetime earnings of a white male

ages 20-24 who graduated from highschool are currently worth one-thirdmore than those of one who only com-pleted the eighth grade. This differen-tial increases with age to 50 per centat ages 60-64 (Table 2) .

Comparison of white male collegegraduates and high-school graduatesshows even more substantial economicgains from education. The lifetime earn-ings of the 20-24-year-old high schoolgraduate are $150,000, compared with$209,000 for the college graduate-a 39per cent differential. And the earnings ofthe 60-64-year-old, similarly educated,show an enormous differential-89 percent!

For the nonwhite males, educationalso pays off but the monetary returnsare far lower than for whites. Lifetimeearnings for 20-24-year-old nonwhitemale high school graduates are one-fourth greater than for elementaryschool graduates. The differential be-tween college and high school graduatesis about the same. At ages 60-64, the col-lege-high school disparity for nonwhitemales is one-third that of whites-26per cent compared with 89 per cent.

Educational Differentials Between White andNonwhite MalesWe have seen that for both white

and nonwhite males, it pays to obtainmore schooling. But how do lifetimeearnings of nonwhite and white malescompare, assuming the same educationallevel? Figure 1 shows the data on thepresent value of male lifetime earningsby age, color, and years of school com-pleted.

There are wide disparities betweenlifetime earnings of white and nonwhitemen having the same education, a find-ing that is neither new nor surprising.Nevertheless, the magnitude of this dis-parity is disconcerting. The level of ex-pected earnings for the nonwhite malewith a college education is just abovethat of the white male with only eightyears of school completed, and consider-ably below that of the white male highschool graduate.The economic value of human life of

the nonwhite male ages 20-24 with eightyears of school is presently $80,000. Forhis white contemporary the value is$113,000-42 per cent greater. The dif-ferential for high school graduates is 51per cent and for college graduates ashigh as 70 per cent!The differentials between white and

nonwhite males do not vary substan-tially for different age groups of ele-mentary school and high school gradu-ates, but the economic value of agedwhite college graduates (60-64) is morethan double that for nonwhites-a differ-ential of 127 per cent.

Educational Differentials Among FemalesHow do educational levels affect the

lifetime earnings of a female, when herservices as a housewife are included?It is clear that education also pays offfor women (Figure 2). Although thelevel of earnings for women is consider-ably lower than that for men at everyeducational level, there are significantmonetary gains attached to higher edu-cation. At ages 20-24, the earnings ofthe white female high school graduateare one-fourth higher than those of onewho only completed elementary school.And the differential between college andhigh school graduate at that age is con-siderably larger-39 per cent.

For the nonwhite female, the gainsare even greater. At the peak years(20-24), lifetime earnings of the highschool graduate are presently worth 35

VOL. 57, NO. 11. A.J.P.H.1962


per cent more than the eighth gradegraduate and those of the college grad-uate are worth 43 per cent more thanthe high school graduate.

Other DifferentialsIt is true that the levels of lifetime

earnings for men are higher than those

for women. But how far apart are theyfor the same level of education? At ages20-24, the future earnings of the whitemale eighth grade graduate are presentlyvalued at $113,000, or nearly three-fourths higher than those of his femalecounterpart. For high school and col-lege graduates, the differentials are 87

Figure 1-Present value of lifetime earnings of males, by age, color, and years ofschool completed, discounted at 4 per cent, 1964

c

z0

vn

a

z0I-

z7F.

UINDER 1 5-9 15-19 25-29 35-39 45-49 55-59 65-69 75-79

AGE GROUP854

NOVEMBER. 1967 1 963

Figure 2-Presenlt value of lifetime earnings of females, by age, color, and years ofschool completed, discounted at 4 per cenlt, 1964

1200

z

z80

w

.<

LaoLli ~ v *

WLo 40 7-

UNDER 1 5-9 15-19 25-29 35-39 45-49 55-59 65-69 75-79 85+

AGE GROUP

per cent and 86 per cent, respectively(Figure 3). These are certainly sizabledifferences, and they are even greaterat ages 30-34.Comparison of white male and fe-

male lifetime earnings at the oldest agesdiscloses a completely different picture.At ages 60-64, lifetime earnings ofwvomen with eight years of school com-pleted are greater than those of menbecause of the value imputed to house-wivies' services. But for college grad-uates in this age group the men's earn-ings are higher. Men with higher edu-cation remain employed longer thanthose with a lesser education and theirhigher earnings contribute to make theirremaining lifetime earnings higher thanthose for females with the same educa-tion.

Still another picture emerges whenexpected earnings of male and femalenonwhites are compared. Here, the earn-ings differentials decrease as the educa-

tional levels increase. At ages 20-24, thedifferential is 46 per cent for those withonly eight years of school completed and17 per cent for college graduates.Another set of differentials of inter-

est is between lifetime earnings of whiteand nonwhite females. Here, at peakearnings (ages 20-24), the differentialsfor college graduates of different colorare small indeed-whites are only 7 percent higher than nonwhites. For whiteand nonwhite males, the comparable dif-ferential was 70 per cent-lO timeshigher (Figure 3). The earnings differ-ential between white and nonwhite fe-males with the same level of educationincreases with age, but is substantiallyless than that for males throughout.

Summary

Lifetime earnings at a 4 per cent dis-count rate were presented by age, sex,color, and three levels of education.

VOL. 57, NO. 11. A.J.P.H.1 964


These were developed for use by theeconomist, program planner, and othersin studies requiring such up-to-date anddetailed estimates.Major findings are:Peak discounted lifetime earnings are

to be found at the young adult ages.Peak lifetime earnings for men are

nearly double those for women, evenwith an imputed value for the servicesof the housewife.

Education pays off at each age levelwithin each sex and color group, but theextent of the payoff differs for eachvariable.The payoff of additional schooling for

Figure 3-Present value of lifetime earnings at age 20-24, by sex, color, and years ofschool completed, discounted at 4 per cent, 1964

.1

fu.L5Lt. hA

NOVEMBER. 1967 161965

white males is greater than for non-white males.At each educational level, the value of

lifetime earnings for white males is fromtwo-fifths to one and one-half timesgreater than for nonwhite males, de-pending on level of education and age.The level of expected earnings for

nonwhite males with a college educationis just above that of the white male withonly eight years of school completed andconsiderably below that for the whitemale high school graduate.A college education compared with a

high school education pays off more fornonwhite females than for white fe-males.

REFERENCES

1. Schelling, T. C. "The Life You Save May Be YourOwn." A paper presented at the Second Confer.ence on Government Expenditures, Brookinp Insti-tuition. Washington, D. C. (Sept. 15-16), 1966.(processed) .

2. Dublin, Louis I., and Lotka, Alfred J. The MoneyValue of a Man. New York: Ronald Press, 1946.Chap. 2, pp. 6-21.

3. Huebner, S. S. The Economics of Life Insurance.New York: Appleton-Century-Crofts (3rd edition),1955.

4. Prest, A. P., and Turvey, R. Cost-Benefit Analysis:A Survey. Economic Journal 75:683-735 (Dec.).1965.

5. Fein, Rashi. Economics of Mental Illness. NewYork: Basic Books, 1958.

6. Weisbrod, Burton A. Economics of Public Health.Philadelphia. Pa.: University of Pennsylvania Press,1961.

7. See, for example:The Royal Commission on Health Services. Ottawa,Canada: Queen's Printer, 1964, V. 1, p. 511.American Medical Association. Report of the Com-mission on the Cost of Medical Care, Vol. III:Significant Medical Advances. Chicago, Ill., 1964,pp. 61 and 64.

8. Klarman, Herbert E. "Syphilis Control Program."In: Measuring Benefits of Government Investments.Ed., Dorfman, Robert. The Brookings Institution,Washington, D. C., 1965, pp. 367-410.

9. . "Socioeconomic Impact of Heart Dis-ease." In: The Heart and Circulation. Second Na-tional Conference on Cardiovascular Diseases, Wash-ington, D. C., 1964. Vol. II, chap. 2, pp. 693-707.

10. Rice, Dorothy P. Estimating the Cost of Illness.Health Economics Series No. 6. US Department of

Health, Education, and Welfare, PHS Publ. No.947-6 (May), 1966.

11. See, for example:Becker, Gary S. Human Capital. National Bureauof Economic Research, No. 80, General Ser. NewYork: Columbia University Press, 1964.

. Underinvestment in College Education?American Economic Rev. Papers and Proceedings,American Economic Association 50:346-378 (May),1960.Bridgman, D. S. Problems in Estimating the Mono.tary Valuie of College Education. Rev. of Economicsand Statistics. XLII :180-184, 3, Supp. (Aug.), 1960,pt. 2.Hansen, Lee W. Total and Private Rates of Returnto Investment in Schooling. J. Polit. EconomyLXXI:128-140, 2, (Apr.), 1963.Houthakker, H. S. Education and Income. Rev. ofEconomics and Statistics 41:24-28 (Feb.), 1959.Miller, Herman P. Annual and Lifetime Income inRelation to Education: 1939-1959. American Eco.nomic Rev. 50:963-986, 5 (Dec.), 1960.-_________. Statement before the Senate, 88thCongress. Hearings, Committee on Labor and PublicWelfare. Vol. 6.Renshaw, Edward F. Estimating the Returns toEducation. Rev. of Economics and Statistics 42:318-324 (Aug.), 1960, pt. 1.Schultz, Theodore W. Investment in Man: An Eco-nomist's View. Social Service Rev. 33:109-117 (June),1959.

. "Rise in the Capital Stock Representedby Education in the United States, 1900-51." In:Economics of Higher Education, US Department ofHealth, Education, and Welfare, DE-50027, Bulletin1962, No. 5, pp. 93-101.Weisbrod, Burton A. "Preventing High School Drop.outs." In: Measuring Benefits of Government In-vestments, Ed., Dorfman, Robert. The Brookings In.stitution, Washington, D. C., 1965, pp. 117-171.

12. Miller, Herman P. "Annual and Lifetime Incomein Relation to Education: 1939-1959." Op. cit.

13. . Statement before the Senate. Op. cit.14. Houthakker, H. S. Education and Income. Op. cit.15. Bridgman, D. S. Problems in Estimating the Mone-

tary Value of College Education. Op. cit.16. Weisbrod, Burton A. "The Valuation of Human

Capital." J. Polit. Economy LXIX:425-436, 5 (Oct.),1961.

17. "Preventing High School Dropouts."Op. cit.

18. US Department of Health, Education, and Welfare,Public Health Service, National Center for HealthStatistics, Life Tables, Vital Statistics of the UnitedStates, 1964, V. 11, Sec. 5.

19. Banks, Robert L., and Kotz, Arnold. The ProgramBudget and Interest Rate for Public Investment.Public Administration Rev. XXVI :283-292, 4 (Dec.),1966.

20. Klarman, Herbert E. The Economics of Health. Co.lumbia University Press, New York, 1965, p. 168.

21. Dublin, L. I., and Lotka, A. J. The Money Valueof a Man. Op. cit., p. 70.Weisbrod, B. A. Economics of Public Health. Op.cit., p. 49.

22. Fein, R. Economics of Mental Health. Op. cit.,pp. 18-19.

Mrs. Rice and Mrs. Cooper are medical economists, Office of Research andStatistics, Social Security Administration, U. S. Department of Health, Education,and Welfare, Washington, D. C. 20201.

This paper was presented before the Medical Care Section of the AmericanPublic Health Association at the Ninety-Fourth Annual Meeting in San Fran-cisco, Calif., November 1, 1966.

VOL. 57, NO. 11. A.J.P.H.l1966

the economic value of human life

Documents