Affordability & Access;Costing, Pricing&Accountability Course Goals: 1) Apply theoretical and empirical methods of economics to education policy.

Economics is a powerful framework for thinking about education policy. Economics focuses us on the incentives created by a policy, allowing us to predict its intended and unintended consequences. 2) Distinguish good empirical research from bad. There are thousands of empirical education studies. This course helps you separate the wheat from the chaff by teaching you to read closely and critically. We will discuss the key challenges in education research and learn the best-practice methods that overcome them.

Economics of Education
Education economists analyze both what determines or creates education and what impact education has on individuals and the societies and economies in which they live. Historically at the World Bank a great deal of emphasis has been placed on determining outcomes to educational investment and the creation of human capital. The primary mission of the economics of education group is to identify opportunities for improved efficiency, equity, and quality of education and promote effective education reform processes; to help improve, among both World Bank staff and clients, knowledge of what drives education outcomes and results; to better understanding how to strengthen the links of education systems with the labor market; and to build and support a network of education economists and build bridges to all those who are interested in their work. The World Bank addresses key themes of Economics of Education through work in these five topic areas:

Why is economic analysis important? All countries face the problem of allocating limited resources such as capital, skilled and unskilled labor, land and other natural resources to a variety of different uses such as production of consumer goods, investment in industry, infrastructure, education, health etc. with an aim to reach a more fundemantal goal of reducing poverty, accelerating economic growth and/or reducing income inequalities. Given limited resources, choices have to be made between alternative uses of these resources such that the benefit to the economy and society is as large as possible. Therefore, all World Bank projects must meet certain selection criteria, including economic justification. in this regard the education sector specifically seeks to answer the following questions in a country's public expenditure analysis:

How much is spent on education and what is the share of the government's expenditure? How do governments finance the education sector and what do they finance? Is there equitable distribution of the public resources? Is the public getting its money's worth?

Is the spending adequate and sustainable?

Public-Private Partnerships in Education

What is school-based economic management?\ Goals Investment in Education COST ANALYSIS IN EDUCATION DemandandSupplyofSchooling-AccesstoEducation Economic and Social Externalities in Education The course aims to provide students with a perspective on the pressing issues concerning the education sector in countries -theoretical, methodological and applied- and to bring them towards the research frontiers in certain selected topics. It will also try to shed light on the evolution of economics of education as a sub eld and its contribution particularly in developing economies. Education is central to development and the course will focus on emphasizing this relationship. Broadly, it will assert the importance of education as an investment, adding to the stock of human capital which becomes an essential primoever for economic growth. It will then introduce students to the various methods used in analyzing both the inputs and outputs that go into the provision of good quality education services

Education economics
From Wikipedia, the free encyclopedia

Not to be confused with Economics education.

Economics

Economies by region [show]

General categories
Microeconomics Macroeconomics History of economic thought Methodology Mainstream & heterodox

Mathematical & statistical methods

Mathematical economics Game theory Optimization Computational Econometrics Experimental Statistics National accounting

Fields and subfields

Behavioral Cultural Evolutionary Growth Development History International Economic systems Monetary and Financial economics Public and Welfare economics Health Education Welfare Population Labour Managerial Business Information Industrial organization Law Agricultural Natural resource Environmental Ecological Urban Rural Regional Geography

Lists
Journals Publications Categories Topics Economists Economy: concept and history Business and Economics Portal
This box: view talk edit

Education economics or the economics of education is the study of economic issues relating to education, including the demand for education and the financing and provision of education.
Contents
[hide]

1 Demand for education 2 Financing and provision 3 Education production function 4 Marxist critique of education under capitalism 5 See also 6 Notes 7 Referencecs 8 External links

[edit]Demand

for education

The dominant model of the demand for education is based on human capital theory. The central idea is that undertaking education is investment in the acquisition of skills and knowledge which will increase earnings, or provide longterm benefits such as an appreciation of literature (sometimes referred to as cultural capital).[1] An increase in human capital can follow technological progress as knowledgeable employees are in demand due to the need for their skills, whether it be in understanding the production process or in operating machines. Studies from 1958 attempted to calculate the returns from additional schooling (the percent increase in income acquired through an additional year of schooling). Later results attempted to allow for different returns across persons or by level of education.[2] Statistics have shown that countries with high enrollment/graduation rates have grown faster than countries without. The United States has been the world leader in educational advances, beginning with the high school movement (19101950). There also seems to be a correlation between gender differences in education with the level of growth; more development is observed in countries which have an equal distribution of the percentage of women versus men who graduated from high school. When looking at correlations in the data, education seems to generate economic growth; however, it could be that we have this causality relationship backwards. For example, if education is seen as a luxury good, it may be that richer households are seeking out educational attainment as a symbol of status, rather than the relationship of education leading to wealth. Educational advance is not the only variable for economic growth, though, as it only explains about 14% of the average annual increase in labor productivity over the period 1915-2005. From lack of a more significant correlation between formal educational achievement and productivity growth, some economists see reason to believe that in

todays world many skills and capabilities come by way of learning outside of tradition education, or outside of schooling altogether.[3] An alternative model of the demand for education, commonly referred to as screening, is based on the economic theory of signalling. The central idea is that the successful completion of education is a signal of ability.[4]

[edit]Financing

and provision

In most countries school education is predominantly financed and provided by governments. Public funding and provision also plays a major role in higher education. Although there is wide agreement on the principle that education, at least at school level, should be financed mainly by governments, there is considerable debate over the desirable extent of public provision of education. Supporters of public education argue that universal public provision promotes equality of opportunity and social cohesion. Opponents of public provision advocate alternatives such as vouchers.[5][6][7]

[edit]Education

production function

An education production function is an application of the economic concept of a production function to the field of education. It relates various inputs affecting a students learning (schools, families, peers, neighborhoods, etc.) to measured outputs including subsequent labor market success, college attendance, graduation rates, and, most frequently, standardized test scores. The original study that prompted interest in the idea of education production functions was by a sociologist, James S. Coleman. The Coleman Report, published in 1966, concluded that the marginal effect of various school inputs on student achievement was small compared to the impact of families and friends. [8] The report launched a large number of successive studies, increasingly involving economists, that provided inconsistent results about the impact of school resources on student performance.[9][10] The interpretation of the various studies has been very controversial, in part because the findings have been directly entered into policy debates. Two separate lines of study have been particularly widely debated. The overall question of whether added funds to schools are likely to produce higher achievement (the money doesnt matter debate) has entered into legislative debates and court consideration of school finance systems.[11][12][13] Additionally, policy discussions about class size reduction heightened academic study of the relationship of class size and achievement.[14][15]

[edit]Marxist

critique of education under capitalism

Although Marx and Engels did not write widely about education the social functions of education, their concepts and methods are theorized and criticized by the infuence of Marx as education being used in reproduction of capitalist societies. Marx and Engels approached scholarship as "revolutionary scholarship" where education should serve as a propaganda for the struggle of the working class. [16] The classical Marxian paradigm sees education as serving the interest of capital and is seeking alternative modes of education that would prepare students and citizens for

more progressive socialist mode of social organizations. Marx and Engels understood education and free time as essential to developing free individuals and creating many-sided human beings, thus for them education should become a more essential part of the life of people unlike capitalist society which is organized mainly around work and the production of commodities. [16]

ECONOMICS OF EDUCATION INTRODUCTION J O HN VAIZEY T h e rise in interest in the economics of education whi ch has taken place in the last few years is d ue to a concatenation of several forces. O n e is the rising cost of education itself. It has be en s h o wn that education at early stages of e conomic deve lopment takes be twe en i a nd 2 per cent of the gross national product. In advanced economies it n o w takes at least 4 , a nd sometimes 5 or 6 per cent. Projections for future expenditure suggest that the proportion will b e c o me higher. T h e reasons for this are not far to seek. 1 A s economies g r ow richer there is a switch a w a y from primary industry to secondary industry, a nd then from secondary industry to the service trades. A s individuals b e c o me richer the proportion of their i n c o me devoted to food a nd other basic necessities drops. T h e a m o u n t that they spend on health care, transport, holidaysand educationrises. N ot only this, but increasingly the provision of skills for the e c o n o my at all levels b e c o m es m o re dependent u p on formal arrangements for education a nd training. Doctors in the nineteenth century learnt their trade largely by apprenticeship. At that time the requirements for medical registration we re formalized, a nd the mo u n t i ng knowl edge of the causes of sickness a nd concerning the fundamental sciences lying behind the functioning of the h u m a n b o dy led to the incorporation into the medical curriculum of a great deal wh i ch c an only be imparted by formal education. T h e s a me is true of a host of other professions a nd occupations from dentistry to architecture. 2 T h e accumulation of knowl edge itself, especially of the physical world, has be en the basis of the technological transformation of the world e c o n o m y. T h e accumulation of knowl edge a nd its transmission to n e w generations represents an increasingly important part of e conomic activity. For these reasons alone, therefore, as a nation grows richer it tends to increase m o re than proportionately its expenditure on education. But there are also arguments for m o r e educational expenditure springing 1. I have reviewed this evidence in The Economics of Education, 1962; see also Friedrich Edding, Internationale Tendenzen in der Entwicklung der Ausgaben fr Schulen und Hochschulen, Kiel, 1958. 2 See Jean Floud, A . H . Halsey and C . Arnold Anderson (eds.), Education, Economy and Society, Glencoe, 1961. 619 I N T E R N A T I O N A L S O C I A L S C I E N C E J O U R N A L from concepts of justice a nd h u m a n decency; the a r g ume n t, for instance, that underprivileged groups of the populationworking people, girls,

racial minoritiesshould have access to educational opportunity in the s a me degree as their m o re fortunate contemporaries. T h e aspiration for d emo c r a cy a nd equality expresses itself largely in a desire for education. 1 Another force m a k i ng for interest in the economics of education has b e en the growing concern of economists with problems of economic growth. This subject, a m o n g We s t e rn economists at least, h ad be en largely neglected during the years of ma ss u n e m p l o y m e nt before the second world w a r. Immediately after the second world w a r, h owe v e r, there w as a surge of interest in the subject. Because of the relationship in Keynesian economics be twe en the m o v e m e n ts of gross investment a nd gross national product the preoccupation of growth economists has be en with the extent of investm e n t in physical capital. T h e re has be en a deve lopment of techniques, such as capital output ratios, designed to s h ow h o w m u c h extra output could be expected from any given increase in the proportion of gross national product devoted to physical investment. 2 T h e substantial truth of this teaching about e conomic growth cannot be denied. Nevertheless, it has b e en k n o w n for m a n y years that it is not the whole truth, a nd m o re a nd m o re information has be en accumulated to suggest that the significant variables m a y well lie outside the physical plantthe factories a nd ma chine rywhi ch is erected in an e c o n o m y. T h e re has be en evidence, for e x amp l e, that countries like N o r w a y with high rates of physical investment h a ve g r o wn no faster than those with comparatively l ow rates. T h e re has been evidence that countries w h i ch have increased their proportion of physical investment over a period of a few years have not substantially increased their rate of e conomic growth. Increasingly specialists in the economics of growth, like Professor W . Arthur Lewi s, have suggested that w h at is perhaps m o st significant is the wh o le nexus of social attitudes to growth wh i ch underlie the e conomic process itself. 3 These attitudes change through time in both favourable a nd perverse w a y s. It is obvious that education c an reinforce both tendencies. O n the o ne h a nd m u c h education is concerned with handing on the beliefs a nd values of society, that is, with preserving a nd not with modifying its culture. If the process of g r owth is liable fundamentally to break u p the f r amework of any society (and that is very probable), then it will run directly counter to the aims of the existing education system. This has certainly be en the case with almost all forms of traditional religious education, certainly those wh i ch we re established before m o d e rn forms of capitalist social a nd economic organization b e c a me d omi n a n t. O n the other h a n d, an education system or curriculum wh i ch is based u p on the findings of m o d e rn science i. See Seymour Harris (ed.), Higher Education in the United States, N e w York, 1961. 2. See, for example, F. M . Meier and R . E . Baldwin, Economic Development: Theory, History, Policy, N e w York, 1957.

3. W . Arthur Lewis, Theory of Economic Growth, London, 1955. 620 E C O N O M I C S OF E D U C A T I O N a nd psychology will be concernedalmost by definitionwith the process of c h a n g e. It will p r omo te a nd accelerate it, not necessarily restrictively: evidence has s h o wn that such a system could also be used to diminish the heart-ache a nd the h u m a n cost of a process of transition from a stable but stagnant to a stable but growing society. T h u s , in these a nd other w a y s, the a r g u m e nt has developed about education in particular, a nd social expenditure in general, that outlays on productive investment are not in themselves likely to be successful unless they are supported by a substantial expenditure on the provision of social capitalon the infra-structure of a developing society. B ut w h at is the nature of 'social' capital ; h o w does it differ from 'ordinary' investment? 1 At its m o st material, this social capital consists of things such as roads a nd docks w h i ch yield n o , or very little, direct return to their builders, w h i ch are constructed by governments or other public agencies, a nd are essential auxiliaries to productive investment in factoriesin this case for transportation purposes. At its least material, social investment is concerned with the i m p r o v e m e nt of the health a nd welfare of the people. T h e se are desirable ends in themselves, but because without such investment the process of g r owth will be held ba ck, it is legitimate to class this as productive rather than unproductive expenditure. T h e spread of malaria in Italy, for e x amp l e, is frequently given as o ne of the reasons for the b r e a k d o wn of the R o m a n E m p i r e. T h e l ow productivity of m a n y workers in tropical areas is directly attributable to their poor physical state. Insufficient m e a n s of education is another direct cause. W i t h o ut s o me system of m o d e rn education an advanc ed e c o n o my w o u ld not w o r k, because it depends on general a nd specific skills w h i ch education alone is able to dispense. Another force wh i ch has b e en m a k i ng for the growing interest in the economics of education has therefore b e en the shortages of skills w h i ch appear to h a ve held ba ck e conomic g r owt h. This aspect of the subject c a m e to the fore mainly in m o d e rn fully emp l o y ed economies. In the U . S . S . R . , the Uni t ed States a nd We s t e rn E u r o p e, it has b e en often s h o wn that at points w h e re the e c o n o my is growing fastest o ne of the crucial limitations o n forther g r owth is frequently the supply of skilled people. In S w e d e n, for e x a m p l e, a ny decision by the G o v e r n m e nt to e x p a nd the social services is limited directly by the n u m b e r of doctors, social workers a nd other trained personnel likely to be available in the next ten years. W h e n r e a r m a m e nt started in the early 1950s m a n y countries reported acute shortages of scientists a nd engineers, w h i ch held ba ck their p r o g r a m m e s. T h e rate of road-building in Britain has be en limited by the availability of compe t ent civil engineers. T h e se facts h a ve led m a n y economists to the view that the supply of skills is itself a factor in e c o n omic g r owt han important factor w h i ch is quite distinct from the question of

1. See Michael Kser, 'Needs and Resources for Social Investment', International Social Science Journal, Vol. X I I, N o . 3, p. 409-33. 621 I N T E R N A T I O N A L S O C I A L S C I E N C E J O U R N A L attitudes to e conomic growth a m o n g the population at large, wh i ch w e have already discussed. In ma k i ng projections of the needs for skilled m a n p o w er of all sorts, it has been possible to calculate back into the education system the n u m b e rs of educated people required in order to produce the necessary skills to service the e c o n o m y. This c omp l ex art has been the basis of the French planning system a nd plays a major role in the U . S . S . R . It is c omi ng to the fore in countries like the Netherlands, Yugoslavia a nd S w e d e n. It is n o w seen to be of critical importance in the developing countries of Asia, Africa, and Latin Ame r i c a. In Nigeria, for exampl e, the A s h by Report suggests that in order to produce the skilled people needed to maintain the rate of growth of the Nigerian e c o n o my wh i ch is desirable, the expansion of the education service will have to be considerable. T h e s a me has be en s h o wn in reports on Pakistan a nd Greece. Consequently a great deal of w o rk on m a n p o w e r forecasting has a direct relevance to the education system itself. T h e link between skills and the schools seems to be at the heart of the subject. T h e specialized role of m a n p o w e r forecasting has thus been another force bringing the economics of education into prominence. It has both a wi de and a na r row interest. In m a n y countries, for instance, the growth of the educational service is limited by the supply of teachers. All forecasts of the likely d e m a n d for their services a nd of possible supply suggest that over the next ten or twenty years there are likely to be serious shortfalls. Wi thin education itself the economist has a special role to play in analysing these shortages. Certain conclusions follow from his w o r k. It has be en felt increasingly, for e x amp l e, that investigation is needed into the technology of education. Education itself is an industry like a ny other. It has certain inputs a nd certain outputs. T h e se outputs are not easily defined, but neither are the outputs of m a n y service trades because they do not readily produce series of concrete objects wh i ch can be presented as a statistical series, as happens, say, for the output of electricity or motor cars a nd also because education is not normally bought a nd sold a nd consequently has no definable monetary output. T h e consequences of this are twofold. It is believed of education in particular and of the service trades in general that their technology is stagnant a nd that, therefore, as productivity rises in the rest of the e c o n o my they tend to b e c o me relatively m o re expensive. O n the other h a n d, because their output cannot be measured, because there is not a time series showing h o w the productivity of education has changed, there has be en little w o rk d o ne on the possibilities of reducing the unit cost of education. A n d since the technology of education is not investigated, it does b e c o me stagnant. Since education is the largest single user of skilled m a n p o w er in an e c o n o my a nd since skilled m a n p o w e r is likely to remain scarce in mo st

economies, it follows that the education system is likely always to be short of teachers. Consequently any substantial attempt to step up the output of education mu st d e m a n d a substantial change in its productivity. 622 E C O N O M I C S O F E D U C A T I O N In the articles wh i ch follow the question of the remuneration a nd status of teachers is not u n d er discussion. It w o u ld s e em that because of the likelihood of continuous scarcity of their services their incomes will tend to rise relatively to those of i n c ome earners in general. If steps could be taken to increase the productivity of teachers by the use of ancillary workers a nd m o d e rn technological devices, it is highly probable that their professional status w o u ld be enhanc ed, a nd that this in turn wo u ld tend to raise their remuneration. W a g e s are affected not only by relative scarcity in the labour ma rke t, but also by the social standing of different groups of workers, since the labour ma rke t, as Barbara W o o t t on has indicated, is exceedingly imperfect. 1 T h e study of the relative status of the teaching profession throughout the world is a matter in wh i ch economists a nd sociologists could well c omb i ne so as to m a k e effective a nd authoritative contributions to one of the mo st interesting social questions of our time. T h e status of teaching is indisputably b o u nd up with the status of education a nd on the status of education a great deal in society depends. All these forces, then, c o m b i ne to m a k e the economics of education an urgent, exciting a nd important subject. T h e papers w h i ch are presented here deal with a n u m b e r of topics. First, there is the contribution wh i ch education c an m a k e to the econo m y . Professor Strumilin s h ows that differences in the earning p o w er of people with a nd without education are very great, a nd that this reflects differences in their physical productivity. A similar point about earning has b e en m a d e in n u m e r o us studies in the United States. This has led, for e x amp l e, to the slogan that a college education is wor th $100,000 (and is n o w said to be wor th 1250,00o) . 2 T h e m e t h od w h i ch is adopted is to take earnings through out the working life, discounting for sexual, social a nd morbidity characteristics, a nd the result shows the difference w h i c h, as far as c an be ascertained, is attributable largely to education. It is true, of course, that m u c h of this w o rk suffers from multiple correlations; that is to say, a m a n with a g o od education is likely to be also s ome o ne brought up in favourable family circumstances, favourable geographical circumstances, a nd w h o has other advantages, physical a nd psychological, w h i ch w o u ld in a ny case enable h im to do well throughout life. It could be said, perhaps unfairly, that exponents of this w o rk are discounting the inequalities of the e conomic system, a nd no m o r e . 3 Nevertheless, if y ou could take identical twins a nd

give t h em clearly favourable circumstances, but allow o ne of t h em to have educational advantages while the other w as m a d e to begin w o rk earlier, it seems highly probable that the twin with educational advantages 1. Barbara Wooton, The Social Foundations of Wage Policy, L o n d on 1956. 2. Here a variety of sources are available for study. See, for example, Theodore W . Schultz 'Investment in M a n : an Economist's V i e w ', The Social Service Review, University of Chicago, Vol. 33, June 1959. 3. See m y The Economics of Education, Chapter III. 623 I N T E R N A T I O N A L S O C I A L S C I E N C E J O U R N A L w o u ld be the m o r e highly paid of the t w o. This extra i n c o me w o u ld m e a s u re to s o me extent at a ny rate the contribution w h i ch he has m a d e to society because of his education. Nevertheless, as Professor B o w m a n points out, there are differences b e t w e en the individual's r ewa rd a nd the r ewa rd w h i ch society as a w h o le m a y get. T h e se differences m a y be that the advantages of his education accrue to s o m e b o dy else, as m i g ht h a p p en in a system of slavery or exploitation. T h e re m i g ht be advantages to a c o m m u n i ty in having educated people in it w h i ch d o not necessarily a c c rue to the individual. S u ch people m i g ht be ready, for e x a m p l e, to accept voluntary responsibility in n o n remunerative activities; they m i g ht be less prone to c r ime, less likely to m a k e d e m a n d s on the social services, w h i ch are excessively used by the thoughtless a nd incompetentall this w o u ld be a s o m e w h at imme a s u r a b le 'return'. F u r t h e rmo r e, the m a r k et for labour is a highly imperfect o n e. Differences in w a g es a nd salaries at m o st only indicate differences of marginal physical productivity a nd do not m e a s u re t h e m, a nd they d o not necessarily s h ow the contribution w h i ch is being m a d e to society b y the individual concerned. In a characteristically brilliant contribution M r . Debe auva is s h ows h o w the d e v e l o pme nt of e c o n o m ic thought o n the subject of the returns to skill w a s de t e rmined b y the structure of the labour m a r k et in the nineteenth century. T h e imperfection of the labour m a r k et as it w as then m a d e it difficult to d e d u ce the connection b e t w e en the inputs of skill a nd the w a g es of labour; a nd furthermore the classical theory (whi ch w a s fundamental b o th to the neo-classicists a n d to the Marxists) treated labour as a statistically simple series of units. W e see h o w the elementary education, w h i ch w as all the producer required, w as u n d e rtaken by the c o m m u n i t y, a n d that the professional training of skills w a s accepted by business. This led the industrialists a nd the economists of the nineteenth century to neglect the costs of training, w h i ch did not enter into the entrepreneur's calculations of profits. F or these a nd other reasons, therefore, the simple correlation of earnings through a lifetime wi th educational b a c k g r o u nd is not necessarily a decisive indicator of the c o ntribution w h i ch education m a k es to e c o n o m ic life. It certainly is not, of course, a ny indicator of the contribution w h i ch education m a k es to the culture. T h e reason w h y calculations of this sort are undertaken is to enable policy decisions to be m a d e a b o ut the degree to w h i ch education should be e x p a n d ed relative to expenditure o n physical plant a n d buildings

in productive business. Therefore the yield of education has to be expressed as a rate of return o n its costs. 1 T h e w h o le question of h o w to m e a s u re educational expenditure has c o m e to the fore in recent years. Until very recently the statistical calculations represented only that part of G o v e r n m e nt expenditure w h i ch i. O E C D Policy Conference on Economic Gr owth and Investment in Education: Paper II, Targets for Education in Europe in igyo (by Ingvar Svennilson, in association with Friedrich Edding and Lionel Elvin). 624 E C O N O M I C S O F E D U C A T I O N w as labelled 'educational expenditure' a nd therefore excluded gove rnment expenditure on education w h i ch appeared u n d er other budgetary heads, a nd excluded private expenditure. Furthermore, there w as no attempt to break d o w n expenditure on education into its c o m p o n e nt parts a nd to ensure that these we re internationally compa r abl e. N o r w as there any attempt to express this expenditure as a proportion of the gross national product. T h e first attempt to do this in a systematic w a y in a single country took place in 1958. 1 Since that time a substantial a m o u nt of progress has been m a d e in the systematic me a s u r eme nt of educational expenditure, w h i ch allows valid international comparisons to be m a d e . This improvem e n t in financial a nd e conomic statistics has be en a c compani ed by an attempt to bring s o me sort of order into educational statistics generally. A rapid imp r o v eme nt is n o w taking place throughout the world in this respect. For the first time it has b e c o me possible to s h ow h o w far the percentages of age groups at school in different parts of the world are capable of being validly c o m p a r ed with o ne another. This imp r o v eme nt in the statistical basis of knowl edge about educational realities will enable a substantial a m o u n t of progress in theoretical w o rk to take place. 2 O n e such conceptual issue whi ch divides scholars is the question of 'earnings foregone'. H o w far should the costs of education include not only outlays on salaries a nd goods and services, but also w h at the students might have earned h ad they b e en at w o rk a nd not studying? In national i n c ome accounts as they are c o m m o n l y presented it seems probable that there is little or no place for the concept of earnings foregone. T h e national i n c o me is tied fairly closely to monetary flows, a nd a ny departure from these flows has to have serious justification. It is argued by those w h o support the view that earnings wh i ch students mi g ht have m a d e h ad they not been studying represents a real e conomic loss, that not to mention or calculate the extent of this loss m e a ns underestimating the cost of education. T h at there is s o me truth in this point of vi ew cannot be denied, but m a n y w o u ld feeland I wo u ld have thought that their a r g ume nt carried considerable weightthat this loss is best presented in terms of physical quantities, that is to say, the degree to w h i ch the labour force is reduced by

the withdrawal of students. T o do otherwise is to launch out u p o n a sea of assumptions about w h at earnings w o u ld be if people w h o are not at w o rk we re at w o r k, a nd the figures b e c o me extremely unreliable. In Professor B o w m a n ' s article an attempt is m a d e to suggest me thods by wh i ch the social returns of education c an be discerned. Professor B o w m a n does this in an informative w a y by distinguishing be twe en returns to the individual a nd returns to society at large. It m a y well be felt, howeve r, that in societies wh i ch have a socialist system a nd particularly in such as are underdeveloped a nd proceeding towards higher levels of i n c o me by a centrally directed e c o n o m y, this basic assumption about private earnings 1. See m y The Costs of Education, 1958. 2. Manual of Educational Statistics, Paris, Unesco, 1961. 625 I N T E R N A T I O N A L S O C I A L S C I E N C E J O U R N A L measuring in s o me sense contributions to society is not necessarily the best guide to the mo st practical policy for g o v e r nme nt to pursue, whatever its theoretical justification m a y b e. M a n y of these countries h a ve found that m a n p o w er planning is for t h em the most appropriate form of discerning a connection between their e conomic development p r o g r a m m es a nd the h u m a n skills whi ch are largely the product of the education a nd training system. T h e whole field of m a n p o w e r planning has b e en the subject of c o nsiderable research a nd writing. T h e first systematic examination on an international basis took place at a conference held by the O E C D in the H a g u e in N o v e m b e r 1959, 1 a nd w as the subject of a high-level expert meeting whi ch w a s held at the International L a b o ur Office in G e n e v a, in October 1962. T h e details of this m a n p o w e r planning are not relevant here; it is sufficient to observe that the m a n p o w e r p r o g r a m m es in s o me developing countries have be en formulated in such a w a y as to enable their targets to be translated into d e m a n ds on the education system, a nd that such d e m a n ds require the mo st substantial changes in the structure, organization, content, a nd m e t h o ds of education. 2 S o m e of these points form the subject of M r . Benson's contribution, wh i ch raises the wh o le question of the nature of the production function in education. M r . Benson uses A m e r i c an data in an attempt to distinguish be twe en different costs of different w a ys of teaching. H e points out that there is a secular tendency towards i m p r o v e m e nt in the quality of education w h i ch puts pressure u p on local resources. In the poor countries of the world this tension b e twe en growing d e m a n d a nd available supply is so great as to put intolerable strains u p on the existing education system. M r . B e n s on has three important suggestions to m a k e for dealing with this situation. H e proposes a revision of curriculum, the redeployment of staff, a nd the use of n e w physical devices in the education system. 'It is too early', he says, 'to me a sure the returns of these various changes. A

difficulty, of course, aside from the general one of estimating yield of a service activity, is that the change in process of whatever kind in the schools is likely to yield a s o m e w h at different set of products than formerly we re h a d .' This shows that it is difficult to me a sure the usefulness a nd effectiveness of n e w developments in education. Nevertheless, the general point remains that changes in techniques h a ve taken place in education a nd will continue to take place. Because they are the result of pedagogical research, these changes have often be en insulated from the pressure occasioned by the relative scarcity of different factors, a nd have therefore tended frequently to be cost-raising rather than cost-reducing. It follows that the basis of expansion in education, assuming (as w e mus t) that resources are limited, m u st be a form of technological change w h i ch is cost-reducing; such an 1. Alanpower Needs in an Age of Science, Paris, O E E C , i960. 2. See Investment in Education, Nigerian Federal Ministry of Education, i960 (The Ashby Report). 626 E C O N O M I C S O F E D U C A T I O N expansion could both throw open education to m o re people and m a k e it m o re effective, if future research a nd development on an operational basis we re to relate to k n o w n economic scarcities. It is certain that educational research has for the most part m e a nt investigation of psychological data. This has profoundly changed teaching practice throughout the world by revealing m o re the nature of children a nd by revealing m o re about the nature of the learning process itself. T h e problems in education w h i ch are n o w c o m i ng to the fore s e e m, howeve r, to transcend these psychological problems. T h ey are problems of scarcity a nd shortage, problems of social adjustment, problems of poverty a nd illiteracy. T h e y are in other words problems in economics a nd sociology. It m a y be taken as an accepted fact that the elimination of poverty is the major probl em facing mo st of the nations in Asia, Africa, a nd Latin Ame r i ca at present. This m e a ns that the fundamental intellectual problem facing workers in the social sciences is h o w to aid the acceleration of economic growth. M a n y economists in recent years have turned their attention to the problem of economic growth in the poorer areas of the world. T w o contributions here are concerned with this issue in relation to educationthose of Professor W . Arthur Lewis a nd M . Michel Debeauvais respectively. It is wor th pointing out w h at education can do a nd w h at it cannot d o. As has be en observed above, a great deal of education wh i ch is cultural in content m a y be harmful to growth a nd certainly need not necessarily promote it. O n the other h a n d, certain parts of education are an essential preliminary to growth. T h u s the whole question of the content of education a nd of the me thods by wh i ch instruction is given is raised by any consideration of the place of education in economic growth. Economists have a bias towards subjects wh i ch appear to promote rational me thods of thought, like mathematics a nd the natural sciences, towards subjects wh i ch break u p accepted attitudes, like sociology and other social studies, and

towards practical subjectscrafts a nd elementary agricultural skills wh i ch enable people to develop their capacity to earn their o w n living. T h e whole question of the appropriate balance of education is raised by the paradox of intellectual u n e m p l o y m e n t . 1 In country after country throughout the developing nations there is a thirst for education w h i ch exists alongside severe intellectual u n emp l o ym e n t. Prima facie this is evidence of an imbalance in the content of the courses. At a recent meeting, for e x amp l e, the Swedish Minister of Education d r ew attention to the shortage of scientists and engineers, at the s a me time to the apparently growing surplus of those educated in the humanities. 2 In countries like India this imbalance takes the acute form of there being i. Report of the Committee on National Education, Karachi, Pakistan Ministry of Education, i960. 2. O E C D , Ability and Educational Opportunity in a Modem Economy (Report on the Conference organised by the Office of Scientific and Technical Personnel in collaboration with the Swedish Ministry of Education, Kunglv, June 1961). 627 I N T E R N A T I O N A L S O C I A L S C I E N C E J O U R N A L no jobs for those with qualifications in the humanities. Nevertheless, this is a superficial explanation of the reasons w h y intellectual u n e m p l o y m e nt exists. T h e real reasons lie deeper. The re is, first of all, the unwillingness of people w h o have achieved educational qualifications to w o rk in jobs wh i ch they feel fall below their appropriate social status. T h u s a graduate will often be unwilling to teach in a village school and he is able to maintain himself without w o rk because the extended family system m a k es it possible for h im to live on his family's resources. E v en m o re profoundly the w a ge a nd salary structure of a society m a y inhibit the e m p l o y m e nt of qualified people because the conventional salary for a graduate m a y be such that it is substantially greater than he is worth to the firm whi ch w o u ld otherwise emp l oy h i m. Gradually, however, the existence of u n e m p l o y m e nt and pressures for social change will tend to eradicate the social snobbery whi ch prevents people from taking jobs which they consider beneath t h em, and, as Professor Lewis points out, the growing pressure of n u m b e rs of educated people will tend to drive d o w n their income differential. This will of itself tend to raise the level of employm e nt a m o n g intellectuals relative to that of other people. Nevertheless, a fundamental problem still remains. In a country with substantial reserves of m a n p o w e r and w h i ch is short of physical capital and this is almost a definition of an underdeveloped countrya choice m a y be m a d e between developing the e c o n o my along conventional lines a nd keeping substantial labour reserves underemployed or unemployed, and developing the e c o n o my by ma k i ng use of these labour reserves for intensive techniques of production. T h e latter is in fact regarded by m a n y economists as the most appropriate form of economic development for m a n y Asian nations. This m e a ns that the use of labour p o w er to develop

the e c o n o my m u st not be inhibited by an inappropriate w a ge and salary structure. T h e consequences of such a view in the matter of educational policy are profound. In the first place, the skills whi ch are necessary for an e c o n o my which is developing by labour-intensive me thods are different from those whi ch are appropriate to one w h i ch is developing by capital-intensive me thods. A different sort of worker a nd a different sort of supervisor are required. This m e a ns that the imitation of overseas models of education is even m o re inappropriate than it seems on purely sociological a nd anthropological grounds. B ut it also m e a ns that the development of a wa g es a nd salary policy whi ch leads to the greater e m p l o y m e nt of people (and particularly of skilled a nd qualified people) is inescapable. N o w h e re is this dilemmafor d i l emma it is, in terms of pratical politicsmore evident than in the educational system itself. T h e basic factor wh i ch education m a k es use of is m a n p o w e r . By definition, in mo st poor countries m a n p o w er is relatively cheap (though, of course, skilled m a n p o w er m a y be relatively expensive). Also in country after country there are u n emp l o y ed intellectuals, while the schools are short of teachers. It follows that the possibility of developing the educational 628 E C O N O M I C S OF E D U C A T I O N system depends to a considerable extent u p on driving the cost of teachers d o w n to the levels at whi ch the u n e m p l o y ed intellectuals can be fully absorbed, or on devising m e a ns by w h i ch they can be employed. This is a pa r adox indicating that the existing money/pr i ce system does not reflect the prevailing scarcities of real factors in an e c o n o m y. It is also a probl em in public finance, since the education system is usually largely financed from the Excheque r, a nd therefore there is a genuine fiscal barrier to the expansion of the education systema fiscal barrier wh i ch distorts the choices of paths of development a nd prevents their being taken in the light of the prevailing real scarcities a nd abundancies of e conomic factors. T h e re wo u ld s e em to be three courses o p e n. T h e first (which is appropriate to a socialist e c o n o m y) w o u ld be to determine the wages a nd salaries in the light of prevailing e conomic realities a nd the social values wh i ch should be attributed to different professions. In this context the technique of shadow-pricing (which wo u ld enable the salary of a teacher in the gove rnment accounts to appear very low, whether or not he w as paid a higher salary) w o u ld on our assumptions of intellectual u n e m p l o y m e nt clearly indicate a substantial expansion in the education system. In other wo r d s, on this assumption a nd in a given context, the strength of the socialist system lies in that it r emoves the barrier to educational (and economic) growth wh i ch exists because of the difference between the path dictated by reference to the real forces in a situation a nd the path wh i ch is indicated by the mone t a ry distortions of those forces a nd the prevailing fiscal barrier. T h e practical difficulties in pursuing a policy by reference to real rather than mone t a ry forces are of course great, especially in an o p en e c o n o m y. T h e second course w o u ld be to adopt a policy of laissez faire. It could well be argued that this w o u ld not by itself lead to such a pattern

of mone t a ry prices as to enable growth to take place as rapidly a nd as effectively as a direct calculation in terms of real scarcities, by the technique of shadow-pricing, for exampl e, assuming that this could be accurately a nd effectively m a d e . T h e third possibility whi ch offers an alternative to both full-blooded socialism a nd laissez faire is not really an intellectual alternative to either, but a pragmatic c omp r omi se or an evasion of the issue. Nevertheless, it reflects the acuteness of the choice of paths of development (with all its social a nd political consequences) wh i ch any developing nation has to m a k e . This is the use of substantial overseas aid in order to develop the education system. W i t h regard to this, there exist t wo different schools of thought. At regional meetings in Asia, Africa a nd Latin Ame r i c a, the poorer countries have called u p on the wealthier countries to devote s u ms running into billions of dollars to overseas aid in developing education. T h e a r g ume nt advanced is that the s u ms required to develop education are so large that the poorer countries cannot by themselves provide t h e m. Nevertheless, it is difficult to see h o w overseas finance can directly help in the development of an educational system. T h e real problem is of course the supply of teachers. H o w e v e r , the possibility of using a substantial 629 I N T E R N A T I O N A L S O C I A L S C I E N C E J O U R N A L n u m b er of overseas teachers in a developing e c o n o my is not great, not only because teachers are scarce already in the richer countries, but because they do not speak the language of the countries to wh i ch they mi g ht be sent, a nd because the development of an educational system by foreign nationals is not socially or educationally desirable. Consequently, the supply of teachers in large n u m b e rs from the richer to the poorer countries is not really a practical policy. T h e supply of books, teaching e q u i pme nt a nd temporary buildings from overseas nations is, of course, a possible a nd fruitful material contribution, though it represents a small part of the total costs involved. It is therefore difficult to see h o w the millions of dollars whi ch have be en asked for can be turned into those real factors wh i ch are necessary for the development of education. In a country, howeve r, with considerable intellectual u n e m p l o y m e nt the major real factor already exists a nd lies unused, a nd the limitation of growth in m a n y sectors of the e c o n o my (including education) is m o re often than not the balance of p a yme n t s. A n y attempt to accelerate economic growth within the country reveals itself first in an inflation a nd then, instantly, in a serious deficiency in the balance of p a yme n t s, w h i ch has to be covered by overseas aid. T h e greater the degree to wh i ch this deficiency is covered the m o re rapid the possibility of e conomic growth within the country, subject, of course, to real limitations of certain crucial items of supply. Therefore a ny measure w h i ch leads to the greater support of the balance of p a y m e nt can permit m o re rapid growth of a particular industry or service, if the growth of that service is not itself held back by a crucial bottle-neck. (In a country with intellectual u n e m p l o y m e nt this bottleneck should not exist for education, because the supply of pupils a nd

teachers is, by definition, available for rapid expansion.) T h e a r g ume n t, therefore, for overseas support of education is a special case of the a r g u m e nt for general overseas support for the balance of payments in developing countries. It has never yet been m a d e clear h o w far this special support of a particular service is likely to increase the total of overseas aid available, or h o w far it is a device to divert activities inside the country towards that servicein this case, education. If it is the first, then propaganda for greater overseas aid is entirely to be w e l c o m e d. If it is the second, then it m a y be that the development of education will accidentally be given priority w h e n it should not. T h e case of special support for particular purposes is on the verge of external interference in internal affairs, wh i ch has been strongly objected to by the receiving countries. As suming h owe v er that the growth of education will take place as rapidly as possible, with or without overseas aid, Professor Lewis raises a series of questions of the greatest interest a nd importance. O n e of these is the extremely high cost of education in the poorer countries. M r . Benson's a r g ume nt for technological development, w h i ch w as referred to above, is even m o re appropriate in this context than in the wealthier countries of the world, especially as the poorer countries are, in fact, using an imported technology in their education systems wh i ch is clearly inappropriate to 630 E C O N O M I C S OF E D U C A T I O N their real needs, not only economically, but culturally a nd socially as well. It is clear that a 'technological breakthrough' is essential at this stage if education is to develop rapidly a nd effectively. Professor Lewis raises a further question w h i ch is just as important: the appropriate shape of the education pyr amid. In all developing countries pressure exists (for democratic reasons alone) for expansion of primary education. This in turn produces candidates for admission to secondary education (and ultimately to tertiary education). A n y expansion of primary education leads to a consequential expansion of the other sectors. B ut these sectors are affected by forces tending to m a k e t h em expand disproportionately. In order, for e x amp l e, to provide the skills necessary for the e c o n o my there is also a tendency for the secondary a nd tertiary sectors to e x p a n d, independently of the supply of candidates c o m i ng from the primary schools. T h e most striking instance of this is the supply of teachers. Primary education can only expand as fast as the n u m b er of primary school teachers. It follows that the expansion of secondary a nd tertiary education is linked functionally (by the pupil-teacher ratio a nd the rate of attrition of the existing stock of teachers) to the growth of the primary education system. This relationship determines to an extraordinary extent the pattern of educational growth within a country, a nd without a fundamental re-thinking of the whole structure (such as that proposed by Professor Strumilin) it is very difficult for policy decisions to be reached on matters of expansion a nd restriction. T h e study of the fundamental relationships be twe en different parts of the education systemwhat might be termed an experiment in educational

logisticsdemands a considerable degree of e conomic sophistication since it is an exercise in choice with limited resources. So far w o rk wh i ch has been d o ne in this field tends to rely u p on two different sorts of approach. O n e is to investigate the prevailing numerical relationships between different sections a n d, on the basis of such information as is available, to project w h at these numerical relationships are likely to be for the future. For e x amp l e, the rate of growth of children in the sixth forms of English schools is thought to be increasing by something like i y2 to 2 per cent per year, a nd plans have therefore to be m a d e to a c c o m m o d a te t h em on the a s s u m ption that this growth rate will continue. M o r e profound investigation might s h ow the forces underlying these relationships a nd enable predictions to be m a d e about likely alterations in their nature. This leads on to a second m e t h o d, wh i ch is the construction of hypothetical models derived quite frequently from international experience. The se sorts of charts are used to great effect in Une s co's handbooks on world education. T h ey nevertheless imply that certain prevailing m o d es of organization c an easily be imported or exported. Professor Lewis's w o rk is a reminder that this kind of a r g ume nt is a dangerous one, even though it m a y be inevitable in the present situation w h e re there is a dearth of statistics a nd organizational capacity in m a n y of the developing nations. It is dangerous because the tendency to copy a form of organization grows 631 I N T E R N A T I O N A L S O C I A L S C I E N C E J O U R N A L into the tendency to copy that form in its most minute details, a nd this has led to palpable absurdities, such as the widespread use of the classical curriculum of Fr ench lyces a nd English public schools in tropical areas, while the study of the local culture is totally neglected. M r . Debeauvais' suggestion of a similar p y r amid of h u m a n capital for each major e conomic sector is a strikingly original parallel deve lopment to this m e t h od of analysis of educational systems, wh i ch might well have important practical application. T h e reform of education a nd its realignment with the needs of the e c o n o my is of great importance for the developing nations, a nd M r . Debeauvais' proposed m e t h od suggests a fruitful w a y in wh i ch the contribution of the economist to the analysis of education might be advanc ed. In this field, as in m a n y others, the economist acts as a craftsman, or as a critical c omme n t a t o r, a nd his contribution should perhaps be less on the highly theoretical planeas, for instance, a discussion of the relationship between future i n c ome a nd present a nd past outlays on education a nd should consist m o re in m a k i ng pragmatic a nd helpful suggestions to statesmen a nd administrators grappling with the i m m e n se task of bringing education to poor countries. 632 A. S T U D I E S THE ECONOMICS OF EDUCATION IN THE U.S.S.R. STANISLAV STRUMILIN In the U . S . S . R . the education a nd upbringing of the younger generation

are closely associated with life a nd socially useful w o r k. Adults m a y , while actually engaged in production, continue their studies according to their o w n personal vocation a nd preferences. In conformity with the objectives of the n e w p r o g r a m me of the C o m m u n i st Party of the Soviet U n i o n, this system of education promotes the all-round development of the m e m b e rs of communi st society, thus contributing towards the solution of one of the most important of all social problems: the elimination of the essential differences between intellectual a nd physical w o r k. T h e educational system also promotes the achievement of the country's immediate economic tasks since it is a potent factor for progress in this sphere, a nd also for the rapid increase of productivity. In the U . S . S . R . education at all levels, from primary to higher, is free. It is financed mainly from the accumulation of collective socialist productions; and from 1965, w h e n taxes will no longer be paid by the population, it will be supported entirely by the resources of the e c o n o m y. Increasing attention is being given to estimating the returns to be expected from investments in this field of cultural construction. Private entrepreneurs are usually guided in m a k i ng investments by the a m o u nt of profit they h o pe to derive from t h e m, without considering the interests of other people or of society as a whol e. Profit rates, however, depend also on various factors affected by the general situation, such as spontaneous fluctuations in market prices, b o o ms a nd slumps, competition of monopolies, a nd speculations on the stock exchange, strikes, etc. In the socialist countries, on the other h a n d, wh e re private capital a n d, consequently, variations in the general situation, play no part, such a criterion as a high profit rate cannot provide a sufficiently accurate or objective yardstick for measuring the return from investments, mainly because the interests of the whole society must be taken fully into consideration w h e n the expenditure of sums derived from collective saving is involved. T h e profit rate never represents m o re than a part of the returnonly that part received by the owner of a given enterprise. T h e total return from a pa rticular investment can be measured only by the resulting increase in the social productivity of labour since this not only raises the rate of profit for 633 I N T E R N A T I O N A L S O C I A L S C I E N C E J O U R N A L a given enterprise but also reduces the cost of a certain category of products for the whole of society. T h e productivity of labour is determined by various factors, the most important of wh i ch are a high standard of technology and the extent to wh i ch h u m a n labour is assisted by p o w e r. At the present time automation a nd remote control are being m o re a nd m o re widely introduced a nd constant advances are being m a d e in the use of cybernetics a nd electronics. T h e application of science is be coming a decisive factor in the development of the productive forces of society a nd the time will c o me w h e n science itself will be a direct productive force in the fullest sense- of the term. But, if this is to be brought about, m o re a nd m o re of the population mu st be familiarized with the achievements of science a nd this is possible only if large

investments are m a d e in school a nd adult education. T h e most advanced technology can produce no results until it has been assimilated by the workers, a nd the m o re complex a nd expensive technical equipment is, the wider the sphere of knowledge without wh i ch it is easier to break it than to use it properly, the m o re important it is to have a wise economic policy with regard to education, a nd the m o re fruitful capital investments in this field b e c o m e. This has been confirmed by the experience of the U . S . S . R . Prerevolutionary Russia w as notorious for its almost entirely illiterate population. For this reason the n ew workers' a nd peasants' government h ad to begin its cultural w o rk by measures to do a w ay with illiteracy a m o n g the population at large. O n 26 N o v e m b er 1919, while still involved in the difficult struggle against counter-revolution a nd foreign intervention, the Soviet G o v e r n m e nt published a decree concerning the abolition of illiteracy a m o n g all adults under fifty years of age. This w as achieved within twenty years. As early as 1924, however, the country w as confronted with an e n o rmo us n e w task of the s a me order. A draft ten-year school development plan designed to provide universal free a nd compulsory education for all children, starting with not less than four years' primary education, w as submitted to the State Planning Commi s s ion ( G O S P L A N ) for consideration. It w as at this time that G O S P L A N specialists first began to give serious attention to the economic returns from expenditure on school education. Consideration of the salaries a nd productivity of different categories of m a n u al a nd other workers in relation to their educational qualifications, s h owed conclusively that even the most elementary school education is m u c h m o re beneficial to a worker than a similar period of practical training at the bench. In m a k i ng these calculations, w e took into account the effects of several factors, such as age, professional experience a nd technical qualifications. For exampl e, the rudimentary instruction gained in one year of primary education increases a worker's productivity on the average by 30 per cent, whereas the imp r o v eme nt in the qualifications of illiterate workers and the increase in their output, resulting from a similar period of apprenticeship at a factory, is only 12 to 16 per cent a year. T h e improve634 E C O N O M I C S OF E D U C A T I O N m e nt in qualification resulting from one year's education at school is, on the average, 2.6 times greater than that resulting from one year's apprenticeship. After four years' primary education, a worker's output a nd wages are 79 per cent higher than those of a first-category worker w h o has had n o Xf schooling. After seven years' study (incomplete secondary course), an office worker's qualifications m a y be as m u c h as 235 per cent above the lowest level; after nine years' study (complete secondary course), as m u c h as 280 per cent a b o v e; a nd after thirteen or fourteen years' study (higher education), as m u c h as 320 per cent 1 above.

Highly qualified workers lead to high productivity of labour a nd increase not only their o w n earnings but also the social product a n d, consequently, the national income. T h e return from education m a y also be determined by compa r ing a society's expenditure on school education with the resulting increase in the national i n c ome. For exampl e, according to the calculations m a d e in 1924, the s u ms required to carry out the proposed reform in primary education (increase the n u m b er of children enrolled in schools from four to over eight million in ten years) we re estimated at 1,622 million (old) roubles. T h e increase in the national income resulting from the rise in the productivity of labour of workers educated during those years w as already by the end of the period (after only five years' e m p l o y m e n t ), as m u c h as 2,000 million roubles, whi ch m o re than covered the expenditure. T h e active life of a worker, between leaving school a nd retiring, is, however, not five but thirty-five to forty years a nd the greater competency acquired by study at schoolthe cost of whi ch is broadly offset after the first five years of employmentcontinues to serve without placing a ny burden on the budget. According to the s a me estimate, the enormous rise in the productivity of labour d ue to primary education will increase the national i n c ome, over this period of years, not by 2,000 million but by 69,000 million (old) roubles. 2 S u ch are the returns from primary education, which is the least expensive to provide. T h e first major objective of the Soviet educational development plan w as in fact already practically achieved by the beginning of 1934. C o m pulsory primary education w as provided everywhere a nd the length of the course h ad be en extended to seven years in the towns. T h e next object w as to introduce secondary education for all. Al though progress w as temporarily delayed by Hitler's invasion, the wo rk is n o w well on the w a y . T h e n e w p r o g r a m me of the C o m m u n i st Party of the Soviet U n i on sets forth the following objectives: the introduction of eleven years' compulsory secondary general a nd poly technical education, by 1970, for all children of school age in both t own a nd country, a nd of eight years' education for all y o u ng people already employed in the national e c o n o my w h o have not h ad sufficient education. Be twe en 1970 and 1980 all will have the opportunity 1. Details of these calculations will be found in Stanislav Strumilin, Problem? ekonomiki truda (Problems of Labour Economics), published in the U . S . S . R. in 1957 (p. 155, 156, 175, 178). 2. Ibid., p. 165. 635 I N T E R N A T I O N A L S O C I A L S C I E N C E J O U R N A L of completing a full secondary education. T h e development of secondary a nd higher education will continue as n e w advances are m a d e in science a nd technology. Those so wishing will be able to pursue higher or specialized secondary education (either full-time or while working in production) in higher technical education establishments or people's universities. Soviet achievements in the development of the educational system s h ow

the following results. Be twe en 1924 and 1934 alone, enrolments for the shorter primary school course (four years) rose from 8.3 million to 18.3 million, the latter figure covering nearly all children in the appropriate age group. After a considerable falling-off during the w ar years, the s a me level w as again reached a n d, by i960, there were 18.6 million pupils enrolled for the first four years. T h e n u m b er of older children receiving general a nd specialized education enrolled in grades V to X in secondary schools, a nd in technicums, a nd the n u m b er of students attending higher educational establishments, is increasing even m o re rapidly (see Table 1). T A B L E I. Increase in enrolments within the U . S . S . R. since Tzarist times (in thousands of pupils) Type of school 1914 1930 1 1940 1950 i960 Secondary general (grades V X) Specialized secondary Higher educational establishments Total Percentage 1. 1930 is taken as the base year, repre! 2. 2.6 million in 1961 [Pravda, 25 Octob 506 36 112 654 22 renting 100 er 1961). 2 04a 587 288 2 9 17 100 13796 975 812 15583 534 14 612 1 298 1 247 17 157 588

17392 2 060 2 3g6 2 21 848 749 Be twe en 1930 a nd i960, the population of the U . S . S . R . increased by 39 per cent a nd the n u m b er of enrolments in educational establishments above the first level increased 7.5 times, or 22 times in comparison with 1914. As a result, the specialized workers mo st urgently needed for the national e c o n o my we re rapidly trained. T h e n u m b e r of graduates of secondary a nd higher educational establishments also increased from year to year, as did the n u m b er of teachers, doctors, engineers a nd scholars. Table 2 shows the post-revolutionary growth in this field (within the Soviet U n i o n ). Dur ing the early years of the planned reconstruction of Soviet industry, m a n y people thought that technology w as the solution to all problems. B ut the first attempts to get uneducated workers to master technical me t h o ds proved so difficult that another slogan h ad to be adopted: specialized workers provide the solution to all problems. It w as for this reason that the training of specialized workers at educational establishments expanded so 636 E C O N O M I C S OF E D U C A T I O N " T A B L E 2. N u m b er of graduates Type of school Secondary general Secondary technical Higher general and technical education establishments Total Percentage (in thousands) 1914 no data available 4-9 10.7 no data available 1930 48 61.4 43-9 ! 53-3 100 1940 277

237 126 640 417 1950 284 314 177 775 506 i960 1 055 484 343 1 882 1 SS3 considerably be twe en 1930 a nd 1940. T h e w a r slowed d o w n this process but the years of peace be twe en 1950 a nd i960 s aw it continuing a nd developing once m o r e. B e twe en 1918 and i960, higher educational establishments' alone produced 4 , 7 8 1 , 0 00 specialists, a nd secondary technical schools 7,744,000, m a k i ng a total of 12,525,000, of w h o m 6 , 7 5 5 , 0 0 0, or 54 per cent, graduated during the last ten years. 1 M u c h w as also d o ne to eliminate illiteracy. B e twe en 1920 a nd 1939, special schools a nd courses gave training to over 96 million illiterate or semi-literate adults. Furthermore, over a period of only twelve years (from 1924 to 1 9 3 5 ), 30.9 million children completed primary school (four-year course) a nd incomplete secondary school (seven-year course). Table 3 indicates the rate of increase in the n u m b e r of specialists with higher a nd secondary education in the U . S . S . R . over the twenty-year period 1939-59. 2 In twenty years (1939-59), the population of the U . S . S . R . rose from 190.7 million to 2 0 8 .8 million, representing a n increase of 18.1 million, or 9.5 per cent, while the economically active population rose from 88 to 99 million persons, an increase of 12 per cent. T h e n u m b er of persons having received higher or secondary education increased by 43 million, or 2 70 per cent, despite the losses d ue to the w a r . A n even greater increase in this proportion is to be observed a m o n g the gainfully emp l o y ed population (335 per cent) : in the case of industrial a nd collective-farm workers engaged mainly in m a n u al w o r k, the proportion of those with higher a nd secondary education rose from 4 .3 to 31.6 per cent; for industrial workers as a wh o le the figure is as high as 38.6 per cent, reaching 53.5 per cent in the case of the metallurgical a nd metal-working industries, 64 per cent in that of the printing a nd publishing industries, a nd still higher percentages for certain other categories of workers.

3 1. In 1913, in Tzarist Russia, about 290,000 persons had received a higher or secondary education. 2. Narodnoe hozjajstvo SSSR v ig6o godu [The Soviet E c o n o my in i960], p. 21, 28, 29-36, 33 38, 393. In addition to the general education facilities available, there are also vocational training schools and courses of further training for persons already in employment. Between 1951 and i960 alone, 5,649,000 workers completed 637 I N T E R N A T I O N A L S O C I A L S C I E N C E J O U R N A L T A B L E 3. Rate of increase in the n u m b er of specialists with higher a nd secondary education in the U . S . S . R . (1939/59) Total for the U.S.S.R. (in thousands) Hi g h er education Secondary, general a nd specialized education C omp l e te secondary (ten years) Incomplete secondary (seven years) Total (secondary education) Total (secondary a nd higher education) 1939 1 177 2 200 12489 14689 15866 Number 3 778 19544 35 386 54930 58708 1959 Percentage increase over 1939 321 888 283 374 370 Per thousand of the po 1939 6

12 65 77 83 pulation 1959 18 94 169 263 281 Number of persons having received higher and secondary education A m o n g the active population M a n u al workers (in millions) Intellectual workers (in millions) Total (active population) (in millions) A m o n g the rest of the population (in millions) Number of engineers and agronomists (in thousands) Engineers Agronomists a nd zootechnologists 3-2 6.7 9-9 6.0 247 295 2 8 .4 18.2 43-o 15-7 834 477 775 272 435 262 338 162 43 498 113

 316 884 434 A supply of highly qualified specialists is of particular i m p o r t a n ce for e c o n o m ic d e v e l o p m e n t. T h e U . S . S . R . h a s, for a long t ime, b e en the c o u n t ry w i th the m o st engineers. At 1 D e c e m b e r i 9 6 0, the Soviet U n i o n h a d m o r e t h an 1 , 1 1 5 , 0 00 engineers a n d over 1 , 9 3 0 , 0 00 qualified technicians or m o r e t h an 3 million engineering a n d technical specialists. Nevertheless, h a v i ng n o fear of ove r -produc t ion of s u ch specialists, the Soviet educational sys t em trained 1 2 0 , 0 00 engineers a n d 2 2 0 , 0 00 technologists in i 9 60 a l o n e, a n d this effort is to be intensified in the future. It m a y be n o t ed that, in the s a me year, the U n i t ed States of A m e r i c a h a d 5 2 5 , 0 00 engineers a n d courses at vocational a nd technical schools a nd 50.6 million either trained for n e w professions or improved their qualifications by studying individually or in groups, while actually engaged in production. 638 E C O N O M I C S O F E D U C A T I O N that, over the past few years, the n u m b er of n e w graduates of engineering schools has been no higher than 38,000 per a n n u m . Literacy campaigns a nd the introduction of compulsory general primary education have provided a wide foundation for secondary a nd higher education. W h e r e a s, in 1940, as m a n y as 62 million of the population (excluding school children) h ad h ad less than seven years' schooling, in 1960 the n u m b er had fallen to 38 million. T h us the section of the population with a low standard of education is contracting rapidly while the average level of education is rising quickly. In ten years' time, howeve r, w h e n the elevenyear course of secondary education will be generally available, the level will be even higher. O v er 20 million workers in the higher age-groups with a low standard of education will have retired by that time a nd their places will be taken by m o re than 30 million y o u ng people from secondary a nd higher educational establishments. T h e majority of m a n u al labourers, even, will not only have h ad a sound technical training but also hold secondary school-leaving certificates. T h e annual n u m b e r of graduates from higher a nd full secondary educational establishments has m o re than doubled over the past ten years a nd at present (1961) it has already reached about 2 million. H a v i ng regard to the rise in population a nd the continuing increase in the graduation rate, the n u m b er of Soviet intellectual workers will thus rise to at least 30 million in the next ten years a nd to 70 to 80 million in the next twenty. In the Western countries, such a high rate of expansion would be liable to result in a heavy over-supply of workers in all the intellectual professions, an increase in u n e m p l o y m e n t, a nd a sharp decline in wa g e s. But in a

planned e c o n o my there are no such dangers. All Soviet citizens are g u aranteed the right to w o r k. T h ey need fear neither u n e m p l o y m e nt nor a fall in wa g e s. T h e country's current production plans provide, twenty years ahe ad, for the increase in the productivity of labour, the shortening of the working d ay whi ch this will necessitate, a nd an increase in nominal wa g e s, as well as an imp r o v eme nt in the real standard of living of all workers, m a d e possible by the growth of the national income. A s the public education system expands, the State funds allotted to it are systematically increased. F r o m 1932 to i960, the proportion of the budget assigned to education rose from 10.6 to 14.1 per cent, the actual a m o u nt involved being about twenty-five times as great. Wh e r e as in 1940 the appropriations a m o u n t ed to 2,250 million roubles, in i960 they rose to 10,300 million roubles (103,000 million old roubles) or approximately 11,500 million dollars. A n d , of course, parents p ay nothing for their children's education. In i960, the unit cost of education (in n e w roubles) w as about 3,000 roubles in the case of the full secondary education course (ten years), a nd no m o re than 3,600 roubles for the seven-year course of secondary education followed by training in a technical school, a nd 8,000 roubles for higher a nd secondary education combined (fifteen years' study in all), including scholarships. It w as naturally lower in the case of the correspon639 I N T E R N A T I O N A L S O C I A L S C I E N C E J O U R N A L dence courses provided by technical schools and higher educational institutions. It will be seen that expenditure on education is relatively l ow, a nd the direct economic return from it is thus all the m o re significant. This return is reflected first of all in the increase in workers' w a g e s. In 1924-25, the w a ge level of Soviet workers w as still very low, although its real value w as no lower than in pre-revolutionary times w h e n the additional social insurance payments are taken into account. Later on the w a ge level rose greatly, with the increase in the productivity of labour, to which the higher standard of education a m o n g the population m a d e a great contribution. It has already been pointed out that, other things being equal, rudimentary education leads to an increase in wa g es of 30 per cent, a nd a full primary education to an increase of nearly 80 per cent. O n the basis of these figures, it can be said that by 1934 the elimination of illiteracy alone resulted in an increase of income of approximately 2,400 million roubles per a n n u m for 57.5 million workers. In addition, 23.8 million pupils completed the primary-school course be twe en 1924 a nd 1935, w h i ch represents a combined annual w a ge increase of approximately 2,650 million roubles. In other words, considering these two categories alone, without taking account of secondary or higher education, the total annual returns from education already exceeded 5,000 million roubles. By 1935, the additional i n c ome of the population, in wa g es and labour benefits as a whole, d ue to the rise in educational standards w as not less than 1 o per cent. Since that time, the w a g e level a nd average standard of qualification

of Soviet workers have risen to three or four times w h at they we re before. By the time they start w o r k, all y o u ng people have had at least four years' education, a nd m a n y of t h em have h ad seven or ten years' secondary education or a higher education. Engineers a nd technicians are, of course, m o re highly paid than ordinary workers. For each year of study over a nd above the four years' primary education, a worker's annual salary rises by a given a m o u n t. In a single year the increase is not great, but over a period of thirty-five to forty years the total of this additional i n c ome represents, in terms of present standards, not less than 12,000 (old) roubles for technicians and 18,000 (old) roubles for engineersamounts far greater than the total State expenditure on their education. Every Soviet worker playing a part in the production of material goods not only completely offsets the expenditure on the 'production' of his labour, but also creates a surplus product 'for society' wh i ch can be used for social consumption a nd saving. T h e improvement of workers' qualifications results not only in higher remuneration for their labour but also in an increase in their 'social' product, whi ch goes into the country's reserves for public consumption a nd saving. Therefore, in studying die efficiency of school education a nd the profit-earning capacity on investments in this field, account mu st also be taken of the portion of the surplus product resulting from improvement in the qualifications of workers. Between 1940 and 1960, the national income of the U . S . S . R . , expressed 640 E C O N O M I C S O F E D U C A T I O N in constant prices, increased from 33.5 to 146.6 thousand million roubles, representing an increase of 338 per cent or 113,100 million roubles (at 1961 prices), whereas the n u m b er of workers taking part in the production of this income rose, over the s a me period, from 54.6 million to 6 8 .4 million only, or by as little as 25 per cent. But the improvement in the qualifications of the labour force resulting from secondary a nd higher education has yet to be considered. If w e equate a unit of complex labour with a certain n u m b er of units of simpler labour, according to the wages scale, i.e., if w e consider, for instance, that one worker with a higher education diploma equals two workers w h o have completed primary school only, the figures given in the above calculation for the n u m b er of workers employed in production m a y be increased by 19 per cent for 1940 a nd 30 per cent for i960. S u ch adjustments, reflecting the e n o rmo us expansion of higher a nd secondary education in the U . S . S . R . , substantially change the dynamics of labour u p on whi ch the social product a nd national income depend. In the Western countries, whe re the standard of education a m o n g the workers has not risen so sharply, adjustments of this kind m a y not be necessary. In the U . S . S . R . , however, they are so considerable that to ignore t h em would be to discount the whole cultural revolution wh i ch is taking place. Another major factor accounting for the increase in the national income is the equipment of the labour force with m e a ns of working, whi ch m a y be measured approximately in terms of the increase in investment per unit of labour, prices remaining constant.

Between 1940 a nd 1960 fixed capital for purposes of production in the U . S . S . R . rose frome 55,700 to 173,900 million (new) roubles, i.e., it increased m o re than threefold. If the n u m b er of workers remains constant, a ny increase in such capital is, as a general rule, reflected in a corresponding increase in production due to the introduction of the additional equipment. If, however, there is a simultaneous increase in the quantity of labour a nd the equipment at its disposal, then the resulting increase in production is proportional to the product of these factors. Table 4 indicates the degree of importance of each of these factors in the increase in the national i n c ome of the U . S . S . R . over the past twenty years. Contrary to Western practice, only that part of the national income derived from primary sources (production of material goods) is considered in this table ; the incomes of persons employed in the provision of services, even the most useful a nd necessary, such as doctors, teachers a nd other scientific a nd cultural workers, are not included, since these persons exchange their labour for that of the workers in the production sector so that their entire income is in fact d r a wn from the s a me primary sourcethe production of material goods. Ha v i ng thus limited the definition of the national i n c ome, in order to avoid counting the s a me items twice, w e h a ve included in the labour force directly contributing to the production of this income only those workers actually engaged in production, thus excluding the considerable n umb e rs of scientific a nd cultural workers, and m e m b e r s of families engaged in 641 I N T E R N A T I O N A L S O C I A L S C I E N C E J O U R N A L T A B L E 4 . T h e dynamics of the Soviet e c o n o my 1940 1950 i960 Amount Percentage Amount Percentage Amount Percentage Fixed capital in thousand millions of roubles 55.7 100 70.7 127 173.9 3 ^ Labour in the production sector N u m b e r of workers (in millions) 54.6 100 56.0 103 6 8 .4 125 Adjustment for standard of qualification We i g h t ed total Adjustment as a percentage of total (weighted) labour force National income of the U.S.S.R. in thousand millions of roubles D u e to increase in labour force D u e to the equipment available to labour: D u e to increase in fixed capital D u e to the combination of these factors

Discrepancies Total Total productivity of labour 10.5 65.1 16.1 33-5 1 ' = 33-5 100 100 100 100 100 14-5 7-5 20.6 36.2 9.0 3-3 6 .4 54-9 138 108 108 27 10 19 164 152 20.4 88.8 23.0 45-6 71.0 25-7 4-3 146.6 194

136 136 212 77 13 438 322 d ome s t ic w o r k , w h o h a ve h a d a s e c o n d a ry a n d higher educ a t ion. N e v e r theless, the role of highly qualified l a b o ur is clearly v e ry imp o r t a n t. T r u e , if the increase in the national i n c o me is considered f r om the c o n s u m e r aspect, its g r o w th over the past t w e n ty years is seen to b e d u e m a i n ly to the great expans ion of fixed capital a n d technical e q u i p m e n t. T h e slight 'discrepancies' s h o w n in the table a b o v e indicate that w e h a ve n ot t a k en sufficient a c c o u nt of the effect of higher standards of educ a t ion o n t he g r o w th of the national i n c o m e . T h e s e 'discrepancies' a p p e ar to s h o w in particular that the return f r om capital investments o v er the period consid e r ed increased slightly m o r e rapidly t h an did fixed capital itself. W i t h a rise in the productivity of l abour, so long as prices r e m a i n constant, o n ly the pur cha s ing p o w e r of the roubl e, dollar or p o u n d varies; there is n o c h a n g e at all in the actual proportions of l a b o ur w i th different types o f qualification u s e d. Sinc e, in i 9 6 0, the total national i n c o me a m o u n t e d to 1 4 6 , 6 00 million roubles at current prices, 23 per cent of this s u m b e i ng d u e to i m p r o v e m e n t in the qualifications of the l abour force, in m o n e t a ry v a l ue this fraction represents n o less t h an 3 3 , 7 0 0 million roubles. S u c h is the contribution m a d e b y s e c o n d a ry a n d higher e d u c a t i on to 642 E C O N O M I C S O F E D U C A T I O N T A B L E 5. Return from, and profit-earning capacity of, the educational system in the U . S . S . R . (in thousand millions of roubles, prices remaining constant) 1 Fixed capital in the educational and cultural sector Return from education Increase in the national income Current expenditure per a n n um Net increase in income As percentage of fixed capital 1. Narodnoe hozjajstvo SSSR v igo g 9 0 0. [Statistical yearbooks.] 1940 A m o u n t Percentage 4-37 5-39 3-13

2.26 52 odu, p . 7 1 6, 100 100 100 100 1950 A m o u n t Percentage 5-39 11.31 4-25 7.06 131 8 44 ; Narodnoe hozfaj 123 210 136 312 i960 A m o u n t Percentage 16.24 33-72 10.30 23.42 I 144 stvo SSSR v 1958 godu 372 626 329 036 , p- 7 7 0, the national e c o n o my in the U . S . S . R . It m a y be c omp a r ed with current State expenditure on education, a nd with capital invested in the building a nd equipment of every kind of educational, scientific a nd cultural institution, for schools, theatres, art galleries, zoological gardens a nd polytechnical m u s e u ms all play their part in the education of workers a nd the improvement of their qualifications. According to the Central Statistical Office, such investments a m o u n t ed to 1,205 million (new) roubles in 1930, 4 , 3 70 million in 1940, 5,388 million in 1950 a nd 16,237 million in i960, at present values. This expansion is striking, but the rise in productivity d ue to secondary a nd higher education has been even m o re rapid. T h e comparison in Table 5 between the increase in productivity a nd the total expenditure on all types of education in the country shows the constantly increasing rate of return. It will be seen that, between 1940 a nd i960, fixed capital in the sphere of education, science a nd the arts increased even

m o re rapidly than in the production sector, a nd that, in constant prices, current expenditure on education m o re than trebled. O v er the s a me years, howeve r, the addition to the national income d ue to higher a nd secondary education increased m o re than sixfold, a nd the net i n c ome, after deduction of current expenditure, increased tenfold. T h e average returnsfor the national e c o n o my as a wholefrom investments in this sector of cultural development has thus broken all k n o w n records, increasing from 52 to 144 per cent per a n n u m . T o m a n y , such a return from the public education system m a y well s e em surprising, especially as it is not included in any form of official accounting. N ot everyone perfectly understands h o w such a high rate of increase in the national income in the productive sector c omes about through the steady 643 I N T E R N A T I O N A L S O C I A L S C I E N C E J O U R N A L expansion of the n u m b e rs of workers with increasingly high educational standards. Lenin, the founder of the Soviet State, s h o w ed as early as 1918 that the educational a nd cultural d e v e l o pme nt of the population w as o ne of the m a i n requisites for raising the productivity of labour. Since then, there has b e en a real cultural revolution in the U . S . S . R . T h e educational a nd cultural development of the whole of the Soviet population has attained a level never previously reached a nd is continuing energetically. It is not only those directly concerned in the production of goods w h o play an active part in this deve lopment, but also those w h o 'prepare, m o u l d, develop, preserve or merely maintain the labour force'. 1 T h e school teacher 'prepares', the professor 'develops', the doctor, to a lesser extent, 'preserves', the labour force of the productive sectors a nd enterprises. A n d if w e do not take account of their w o rk in calculating the value of the country's material production, it is only in order to avoid counting the s a me items twice, since the results of their w o rk are already s h o wn in the i m p r o v e m e nt in the qualifications a nd the increase in the productivity of the workers directly engaged in the productive sector. For the s a me reason, in estimating the national i n c ome, w e also leave out of account the w o rk of pupils a nd students engaged in acquiring the knowl edge a nd mastering the m e t h o ds of thought w h i ch they will later apply in productive w o r k. Their n u m b e r s, h o w e v e r, are e n o r m o u s. Du r i ng the last few years, a total of at least 50 million persons a year, of all ages, h a ve been attending schools a nd following courses of study. T o this n u m b e r w e should a dd yet another 20 million 'industrial innovators' w h o vie with e a ch other in discharging their undertakings to 'study, w o rk a nd live according to the principles of c o m m u n i s m ' , a nd w h o devote every free h o ur a nd every free minute to study, while continuing to w o rk in production. T h e Ai l -Union Society of Rationalizers a nd Inventors n o w n u m b e rs m o re than 4 million workers engaged in production. In i960 alone, the a m o u n t

saved by the application of 2 , 5 3 6 , 0 00 innovations proposed by these frontrank workers is estimated at almost 1,500 million roubles. B ut these inventions represent something m o re than material advantage to the inventors themselves a nd savings for the State. Ac c o u nt m u st also be taken of w h at this ma ss m o v e m e n t in w h i ch they are taking part m e a ns to the Soviet workers. 'It appears to m e ' , writes o ne of t h e m, 'that the chief advantage of proposals for rationalization is not the savings thereby m a d e in roubles a nd kopeks, increasingly substantial as these are. T h e greatest value of this system is that it kindles enthusiasm a nd does a w a y with b o r e d om a nd indifference. . . . Life is g o od a nd o ne goes cheerfully to w o rk in the knowl edge that the challenge of a n e w idea or a n e w probl em of research is waiting. T h at is w h y w e in our wo r k s h op h a ve set ourselves acompletely 1. Karl M a r x, Teoriipribavocnoj stoimosti (Theory of Surplus Value), part I, M o s c o w, 1955, p. 142. 644 E C O N O M I C S O F E D U C A T I O N n e w a i m: every worker m u st b e c o me a rationalizer. Rationalization is an absorbing passion. O n e has to think up o ne idea for a second to present itself. W i th the third one is already a different m a n , longing to study, plunge into books, devour newspapers, living a wonderful adventure.' T h e individual is awa k e n ed to the poetry of creative effort a nd the enthusiasm for w o rk it engenders gives a n e w me a n i ng to his entire life. It wo u ld be difficult to assess the true value of every aspect of this intellectual w o r k, whi ch combines study with creative w o r k, a nd is carried on outside production or the working d a y. B ut there is no doubt that all these aspects of wo r kwh i ch are continually expanding in the U . S . S . R . are, at least indirectly, a potent factor in the growth of production and the high rate of increase in the national i n c ome, characteristic of the Soviet U n i o n. Similarly, the part played in this expansion by the most creative w o rk \ of allscientific workcannot be me a sur ed directly. T h e n u m b er of scientific workers in the U . S . S . R . is rapidly increasing. B e twe en 1940 and i960 alone, it rose from 98,000 to 3 5 4 , 0 0 0, representing an increase of 360 per cent, a nd in another twenty years, at the s a me rate of growth, it will be counted in millions! It should be r e m e m b e r ed that there are no industrial or commercial secrets in the U . S . S . R . a nd that a ny attempt to monopolize a scientific achievement for private ends is inconceivable. Perhaps it is precisely for this reason that n e w possibilities are opened up for scientific development in countries wh e re this practice no longer exists. Science is international. T h e scientists of all countries, including the U . S . S . R . , not only m a k e their contributions to the c o m m o n store of scientific knowledge, but themselves d r aw freely on the best of the achievements of world science. A n d this science, by helping m e n to master the most powerful forces wh i ch nature offers in its bounty, itself b e c omes a m o re a nd m o re powerful force of the h u m a n intellect. T h e treasures of science

accumulated by the creative labour of m a n y generations of reseachers, from every age, cost great effort only to those w h o first discover t h e m, for any discovery of genius can afterwards be easily assimilated by the people at large. O v er thousands of years, millions of great mi n d s, from Euclid to Lobachevsky, from Pythagoras a nd Ar chimedes to N e w t on a nd Einstein, from Heraclitus a nd Aristotle to such lucid, logical theorists as M a r x a nd Lenin, have carried science forward by their creative w o r k. A n d , though these m e n are n o w no m o r e, the fruits of their labours will continue to serve m a n k i nd for thousands of years to c o m e . B ut those truths the discovery of whi ch cost Euclid, for exampl e, a whole life-time of effort, are mastered by our schoolchildren, along with dozens of other subjects, in s o me three years. T h ey learn N e w t o n 's binomial theorem or Mendeleev's periodic classification of the elements in a very short time. But today's schoolchildren are tomor row's producers. A n d the comparative ease with whi ch they master all the scientific knowledge that has earlier been acquired at 645 I N T E R N A T I O N A L S O C I A L S C I E N C E J O U R N A L the cost of great effort, m a k es it possible for us to regard that knowl edge as a free gift from nature. This knowledge has special value only if it is c omb i n ed with practical w o rk for, under the present conditions of production brought about by the development of automation a nd electronics, workers have to face n e w d e m a n d s. In these circumstances, w h at really matters in production is not n o w so m u c h muscular strength a nd toil-hardened hands, but technology enriched by the knowledge and intelligence of the worker at the m a c h i ne a nd on the production line. It is for this very reason that science here b e c omes a direct productive force, along with the technology a nd p o w er necessary for production. Mo r e o v e r, it m a y be said that the 'efficiency' of science in these conditions increases in proportion to the v o l ume of k n o w ledge applied, multiplied by the degree of its assimilation and the extent to w h i ch it has b e en spread a m o n g the workers. T h e successes achieved in public education in the U . S . S . R . over recent decades have m a d e it possible, to a great extent, to apply science directly to production a nd to increase its efficiency as an element in the productive force of labour. B ut this is largely explained by the fact that Soviet workers, w h o feel themselves to be not the servants but the masters of production in the U . S . S . R . , are themselves extremely anxious to improve their qualifications through study; by tens of millions, they vie with each other in increasing output rates, and n o w a d a ys millions of talented inventors a nd rationalizers eme rge every year from the ranks of the workers. Their successes are explained also by the fact that science has greater credit in the Soviet U n i on than a n ywh e re else in the world. ' T o go forward without science is to walk like a blind m a n without a guideone is b o u nd to stumble.' This wise observation is already b e c omi ng an integral part of our folk w i s d o m. 'Science is our c omp a s s .' S u ch is our guiding principle a n d, as the whole of Soviet experience has s h o w n, it leads to brilliant results.

In the U . S . S . R . , the cultural revolution and e conomic development are closely interlinked. Investment in public education produces very g o od returns because it serves the cultural interests of the labouring masses themselves and meets the objective requirements of economic development. T h u s, in this process, the cultural factor is k n o wn m o re and m o re clearly to be economically productive a nd profitable. 646 SOCIAL R E TURNS ' TO EDUCATION M A R Y JEAN B O W M A N It is m y purpose in this paper to e x ami ne a nd untangle part of the w e b of concepts concerning interrelations b e twe en education a nd e conomic development. I shall focus on 'social' versus 'private' returns to e d ucation. This distinction is not o ne of opposites. In fact, since all returns accrue ultimately to individuals, w e could state the formal identity: aggregate social return equals the s u m of its individual c omp o n e n t s. H o w e v e r , if w e a dd w h at y ou get from your education to w h at I get from m i ne but disregard h o w m y education affects yours, or vice versa, the above identity will not exist. T h e total social return m a y be larger or smaller than the s um of individual returns viewed in isolation from each other unless a correction for these interactions is m a d e . Distinguishing between private a nd 'social' returns is not the s a me as distinguishing between ma r k et a nd n o n -ma r k e t, or real a nd psychic returns nor is it related to the distinction between education as consumption a nd as investmentthough these several distinctions often tend to be partially confused with each other. In each instance there are difficulties of me a sur em e nt a nd each has time dimensions that are often neglected. E v en in dealing with the consumption-investment distinction, classical theory tends to me a sure along time scales rather than to incorporate time into the core of the analysis. 1 Analysis of 'private' and 'social' returns to education in particular clearly necessitates exploration of processes of c h a n ge through time, for it is in these changes that the largest discrepancies arise between the s u m m a t ed individual returns a nd the social total viewed as an entity. Ne v e rtheless, such analysis can be no m o re than b e g un in these pages. T HE AGGR EGAT IVE A ND DIFFERENTIAL A P P R O A C H ES T w o , virtually polar, approaches have characterized the recent literature on the economics of education. O n e of these begins with neo-classical economics, measuring returns marginally, as differentials associated with i. Ragner Frisch 's exposition of this theme is very m u ch to the point. 647 I N T E R N A T I O N A L S O C I A L S C I E N C E J O U R N A L education. Either tacitly or explicitly it is a s s umed that all other factors are constant. T h o u g h adjustments are introduced in applying this approach to the assessment of effects of education u p on growth, the basic framework is initially a static one. This starting-point has sharply illuminated the i mportance of distinguishing between private a nd social returns in the Pigou tradition a nd as a necessary part of the attempt to identify a Pareto

Optimum. T h e opposite starting-point views total or social returns as an entity. Ma r x i an theories a nd the recently popular aggregative input-output analysis are alike in taking the State rather than the individual as the m a i n concern. This approach, confronting traditional marginal analysis, has pointed u p the fallacy of aggregation. It has the advantage that all returns are accounted for, but the disadvantage that, except by introducing a s s umed constancy in functional relationships, it is impossible to distinguish the role of education from that of other factors, or private from social education. H e re it m a y be advisable to insert a brief c o m m e n t on the importance of identifying private returns even in a socialistic e c o n o m y. ( No o ne questions the importance of identifying social returns in a predominantly private enterprise e c o n o m y .) Allowing the individual to fade from view does not eliminate the fact that production depends on the efforts of individuals a nd that incentives depend u p o n returns w h i ch the individual regards as his. Socialist planners k n o w this well e n o u g h. A n d so a reality akin to marginal analysis slips in through the w i n d o w. W h a t e v er the characteristics of a n e c o n o m y, the individual still perceives something that is a 'private' return, a differential return seeing 'other things as equal'. T h e se private returns are empirically observable. T h ey m a y be either better or poorer approximations of social returns in a socialistic than in a predominantly free enterprise e c o n o m y, despite the deliberate effort to gear the former m o re directly to social goals. T h e growing interest in using ' s h a d ow prices' in socialist planning is evidence of the recognition of this problem. T HE QUANTITY OF EDUCATION Since w e are concerned here with gross returns to education but not with rates of return on the costs incurred in providing education, it is not necessary to me a sure the quantity of education. Nevertheless, it is necessary to identify just wh a t, conceptually, the education under consideration m a y b e. T o do this requires brief attention to two critical aspects of the me a sur em e nt of quantity of education: (a) the view of the cumul a t ed education emb o d i ed in the adult population as a flow of inputs versus a store of value, a nd (b) the inclusion or exclusion of on-the-job training. I h a ve argued elsewhere that the relevant concept of quantity of education (or ' h u m a n capital' in its educational components) for analysis of education's contribution to national product is a flow concept. Focusing o n production activities only, I suggested such a me a sur e, w h i ch I labelled 648 E C O N O M I C S OF E D U C A T I O N 'effective current stock'; 1 it should be clearly distinguished from gross or net h u m a n capital concepts, w h i ch refer to stores of value, not current inputs into e c o n omic activity. N o discounting of returns is ne eded (in fact discounting w o u ld be inappropriate) in arriving at an 'effective current stock' estimate. Underlying the analysis presented in this paper is such an effective current stock concept of quantity of education, but with o ne

important modification. W e are not interested here in distinguishing b e twe en uses of education in production activities a nd in m o r e direct c o n s u m er enjoyments except as it m a y be possible to me a s u re the former but not the latter. T h e 'effective current stock' concept of education will therefore be modified w h e n c o n s u m er returns are considered, to include education e m b o d i ed not only in the m e m b e r s of the labour force but in the entire adult c o n s u m i ng population. Until recently it w as usual to refer to the impor t ance of apprenticeship a nd similar forms of training b ut without including t h em in the f r amewo rk of education. G a r y Becker has skilfully brought this training within the c o m p a ss of a theory of investment in h u m a n resources, a nd J a c ob M i n c er has applied this theory in a first estimate of the a m o u n t of on-the-job training in the Uni t ed States. 2 Ideally, lifetime i n c o me streams for individuals with the s a me a m o u n t of formal schooling are c o m p a r ed by occupational a nd other classifications, using the time pattern of average incomes in each category. A m o n g college graduates, for e x a m p l e, those w h o receive training on the job begin with lower incomes but e nd u p with higher ones than their fellows. A t first they forgo s o me i n c o me (the opportunity cost of their on-the-job training) ; later their higher i n c omes include a differential that is the return to this supplementary education. B y a s s u ming the 'internal rate of return' to investment in on-the-job training to be the s a me as that to the appropriate prior level of formal schooling, the a m o u n t of the investment in the former is estimated. This analysis a s sumes that markets operate competitively a nd that private decisions are e c o n o m ically rational. It is a s sumed also that cross-section a g e - i n c ome data are g o od indexes to lifetime i n c o me sequences. Fortunately, the a b o ve a s s u m ptions c an be eliminated without destroying the conclusion that is m o st important for the problems considered here: differentials in i n c o me associated with education, as m e a s u r ed by census data, include mo n e t a ry returns to associated on-the-job training. i. ' H u m a n Capital: Concepts and Measures.' Originally printed in: H u g o Hegeland (ed.), Money, Growth and Methodology (Festskrift in honour of Prof. Johan Akerman), L u nd (Sweden), C . W . K . Gleerup, 1961. This article will be reprinted, with minor changes and an added introductory section emphasizing the importance of the input versus the store of value concepts, in: Selma Mushkin (ed.), The Economics of Higher Education, U . S . Department of Health, Education and Welfare, Office of Education, 1962. 2. In unpublished papers prepared for the Exploratory Conference on Capital Investment in H u m a n Beings, sponsored by the Universities-National Bureau Committee for Economic Research, 1 and 2 De c emb er 1961. Gary S. Becker, 'Investment in H u m a n Capital: a Theoretical Analysis', Jacob Mincer, ' O n the-job Training: Costs, Returns, and some Implications'. 649 I N T E R N A T I O N A L S O C I A L S C I E N C E J O U R N A L

T HE RELATIONS A M O N G C O M P O N E N TS OF SOCIAL A N D PRIVATE R E TURNS Because of the frequent confusion be twe en the private-versus-social a nd other a m b i g u o us distinctions in assessing returns to education, I state a simple equation that is applicable to m a n y types of e conomic structure, though the relative values of the variables will differ considerably. T h e variables apply to a short period of time, but the relationship defined c an be examined further in a d y n a m ic context. Let tu represent returns to education as these c an be directly taken by individual i (including n o n -mo n e t a ry returns), defined as the differential over w h at the individual w o u ld c o m m a n d if he h ad no education. ( O r, if o ne looks only at o ne segment of education, it is w h at he could c o m m a n d if he h ad completed only the next lower level of education.) T h e aggregate of these individual returns is A = 02na . Starting in the opposite direction, the aggregate of social returns is designated S (again including n o n mone t a ry real a nd psychic i n c o m e ). W e m a y then set up the identity: S = A + D , D being the discrepancy be twe en S a nd A . Distinguishing the mo n e t a ry (') a nd n o n -mo n e t a ry (") c o m p o n e n t s: S = S' + S" = A ' + A " + D ' + D " For simplicity, it is a s sumed that all returns realized through exchange find expression in mo n e t a ry returns a nd that these are converted into units of constant purchasing p o w e r . 1 T h e returns actually me a s u r ed in m o st empirical w o rk are sets of average i n c ome differentials 'j, a'*, . . . for each educational group, j, k, . . . We ight ing by the n u m b e rs in the respective educational categories a nd s u m m i n g gives A ' , w h i ch is the aggregation of mone t a ry returns to education as seen by individuals, each individual taking other things to be equal. 2 A " is the aggregated contributions of education to private n o n -mo n e t a ry i n c o m e. For an individual a" depends u p on (i) the additional things he c an produce for himself (e.g., repairs to his h o m e ) because of increased skill or m o re leisure; (2) psyshic returns from the sort of job he is able to hold (job preference) ; (3) n o n -ma r k et production that his education prevents or that is incompatible with the time ne eded to realize A ' (the latter implying a negative effect o n leisure) ; (4) a d d ed enjoyments in a ny given leisure time attributable to his education (reading, music appreciation, etc.) ; a nd (5) enjoyments for w h i ch his education disqualifies h i m. Items (1), (2), a nd (4) 1. Thus, for example, housing provided by an employer would be included in 'monetary returns'. T he nature of empirical adjustments to bring actual exchangedetermined returns in line with this specification has been widely discussed and can therefore be by-passed here, though with the stipulation that corrections of this kind would be required at several points in the ensuing discussion. 2. T he need for adjustments to correct for effects of ability, effort, family contacts, etc., that are associated with differentials in educational attainments

has been frequently noted. Recently rough attempts to m a ke such adjustments have been appearing, notably by Gary Becker and Lee Hans en; but none to m y knowledge has yet been published. Distortions due to these factors enter into most of the empirically-oriented concepts discussed in this paper: warnings on this point will not, however, be repeated. 650 E C O N O M I C S OF E D U C A T I O N w o u ld be positive, (3) a nd (5) negative. H e n c e A " c an be positive or negative, t h o u gh educated people normally a s s u me that a? is typically positive a nd h e n ce so is A " . If A " is positive, then it follows that A > A ' a nd m a r k et me a sur es of i n c o me differentials understate aggregate private returns ; if A " is negative, m e a s u r ed m o n e t a ry i n c o me differentials overstate aggregate private returns. 1 T h e re is o ne o v e r w h e l m i ng reason for concluding that in W e s t e rn societies A * is a large positive quantity. This relates not to the a b o ve factors for m e n in the labour force but lies in the contributions of educated w o m e n w h o withhold their services f rom the m a r k e t. T h u s far I h a ve focused u p o n returns within a given short-term period. B ut it should be noted that w h e n looking at long-term developments the informal intergeneration effects of parental education m a y be substantial. T h e se effects e m e r ge in all of the elements of a future S, m o n e t a ry a nd n o n m o n e t a r y, private a nd non-private. D ' a nd D " n e ed not be defined merely as residuals. T h e y are the c o m ponents of social returns that c annot (even wi th full individual k n o w l e d ge a nd rationality) enter into individual motivations a nd decisions. 2 E x c e pt by a process of imputation b a ck to population subgroups, they cannot b e localized in individuals. Since D ' is the simpler, it is considered first. T h e difference D ' b e twe en A ' a nd S ' is mathematically determinate, since A ' is derived b y aggregating i n c o me differentials a b o ve s o me baseline, whe r e as S ' takes into account effects of education u p o n that baseline itself. W h e n returns to the g r a nd total of all education are being evaluated, the baseline from w h i ch to c o m p u te the differentials that m a k e u p A ' will b e the i n c o me of the totally uneduc a t ed. If the diffusion of education through a society raises the earned incomes of the uneduc a t ed (disregarding transfer i n c ome s ), the baseline is raised. It does not matter w h e t h er these i n c omes of the uneduc a t ed are increased at a greater or lesser rate than incomes of the educated ; if they are at all increased by others' education, D ' is positive. T a k i ng the baseline average i n c o me as b', it has t wo c o m p o n e n t s, b0 a nd be, the latter being the part that is determined by the education of other m e m b e r s of the population. W e then h a ve a theoretical m e a s u re of D ' as N ' , w h e re N is the n u m b e r of adults in the labour force. If D ' is positive (probably the usual case), S ' will of course exceed A ' . If D ' is negative because the spread of education reduces the baseline i n c o m e, A ' w o u ld exceed S ' .

If o ne is interested in social versus private returns to s o me part of education only, let us say post-elementary education a nd associated on-the1. T h e distinction between a% and a% would be of no importance for individual decision-making provided error in forecasting returns were the same for both components, since it is their s um that matters. H o w e v e r, there m a y well be greater error or systematic bias in one of these elements than in the other. 2. This does not exclude the possibility of 'altruism' as a factor in individual motivations. Such altruism would enter as a psychic component of A " while additionally tending to raise D . 651 I N T E R N A T I O N A L S O C I A L S C I E N C E J O U R N A L job training, the baseline i n c ome to use is the average for elementary school graduates (g'). Letting S ', represent social returns to post-elementary schooling, A ' the aggregated private returns to that schooling, a nd D ' s the discrepancy, D ' 4 will then depend u p on the interactive effects of the presence of a given distribution of post-secondary schooling a m o n g the p o p ulation u p on the incomes of those with varying lesser amo u n ts of schooling. E a ch of the latter groups will have its i n c ome raised or lowered by s o me aggregate a m o u n t. Letting the suffix e stand in this case for interactive effects of post-elementary education on average incomes of others, w e have: D ' , = rija'j, + m a'kt + . . . rigg'e + ntg'e whe re g' t is the c omp o n e nt of income of elementary school completers that is attributable to post-elementary education of others, ng is the n u m b er of elementary school completers in the labour force, a nd n, is the n u m b er in the labour force w h o have h ad post-elementary schooling, j, k, etc., specify education groups below elementary school completion. T h e s um of (rij + nk . . . + ng + n) is equal to N . Specifying effects in this w a y is of course a variant of marginal analysis closely akin to the 'other things equal' technique, but with the difference that the components are imputed from totals a nd mu st necessarily a dd up to the totals. T h us the aggregative a nd marginal approaches are m e r g e d, or, to put it m o re harshly, forced into mutual consistency. I have ignored changes in the distribution of the components of A ' , assuming t h em to cancel out w h e n aggregating for this part of S '. This is easily justified. For exampl e, suppose that because of m y education I a m able to enhance m y income by swindling s ome o ne else. I have m a d e no productive contribution, a nd m y income wo u ld overstate m y contribution to 'social' returns; m y contribution m a y actually be negative. B ut this is taken into account through the other components of A ' a nd /or b' (or a'j, a'ic . . . g') in that it lowers the monetary incomes of others. A similar formal treatment could be given to D " , though the formalization is m u c h less useful w h e n dealing with non-mone t a ry returns. T h e major

elements in A " have already been suggested. Like D ' , D " (social non-ma rket returns not included in private returns) will include both positive a nd negative components. A m o n g the most obvious positive items are the voluntary services to the c o m m u n i ty by educated m e n , w h o d r aw u p on skills acquired at school or w h o enjoy jobs allowing t h em m o re leisure. If education reduces delinquency rates, this also is a positive non-private a nd largely non-mone t a ry return. D " would include also psychic income attributable to having educated neighbours or colleagues with w h o m one can interact pleasurably. Probably the most important negative c o m p o n e nt in D " is the psychic dissatisfaction flowing from one's awareness of 'negative' status (a negative b"t). It has proved impossible to assess social non-mone t a ry real returns to education (let alone psychic returns). This is d ue primarily not to the difficulty of identifying non-private as against private returns but to the non-measurability of most non-mone t a ry returns. The se s a me dif652 E C O N O M I C S OF E D U C A T I O N Acuities explain omission of mo st non-mone t a ry elements in income from national-income accounting. A DIGRESSION ON T HE A GE INCIDENCE OF SOCIAL VERSUS PRIVATE R E TURNS I have argued elsewhere 1 that there are systematic biases that shift with age in using personal earned-income data to me a sure the productive contributions of the recipients. There are biases also in using cross-section age-income data in a given year as proxies for historical private lifetime incomes. O n ly the former, howeve r, are of direct interest in the present context; they are relevant primarily as they affect differentials of i n c ome for one or m o re educated groups above a given educational baseline. T w o such biases are especially likely to exist. First, better-educated m e n in middle age are mo st likely to m a k e substantial contributions to the social product through voluntary c o m m u n i ty services, involving social returns to education that are not reflected in either private returns or monetary social returns. Second, both formalized seniority rights a nd informal status advantages undoubtedly result in continued p a y m e nt of c o m paratively high incomes to older educated m e n even w h e n their productivity has declined; for this age category aggregate private i n c ome differentials m a y overstate social contributions. A third age distortion involves m o re a failure to realize potential contributions than erroneous measures of social contributions; this is the bias against hiring older m e n . This handicap affects especially m e n with least education ; the effect of this bias is to overstate still further the differential incomes a nd hence actual contributions of the better-educated older m e n . It leads also to understating the potential, though not the actual, contribution of the least-educated older men. Cyclical u n e m p l o y m e nt aside, the principal distortions in projected lifetime incomes arising from the cross-section approach stem primarily

from the effects of obsolescence, or, putting this the other w a y , from progressive improvements in the quality of education received by successive age cohorts. T h e resulting understatement of prospective returns, whether private or social, is not of major concern here. U se of the 'effective current stock' concept of the quantity of education automatically allows for on-the-job training. T h e re is then n o occasion for the estimating procedures (involving discounting a nd internal rates of return) that w o u ld be needed if quantity of education we re viewed as a store of value, or if our m a i n concern we re to identify criteria for the allocation of investments. Contributions of y o u ng people w h o are receiving training on the job are lowered, but at a later date w h e n these s a me y o u ng people have completed their training the returns to it eme r ge in the higher i. In ' H u m a n Capital: Concepts and Measures', op. cit. 653 I N T E R N A T I O N A L S O C I A L S C I E N C E J O U R N A L i n c o m es of m a t u re m e n . T h e se m e n are n o w providing educational inputs into the e c o n o m y that e n c o m p a ss b o th their formal schooling a nd the training they previously received o n the j o b. EMPIRICAL M E A S U R E M E NT OF CHANGES IN M O N E T A RY PRIVATE A N D SOCIAL R E TURNS O V ER TIME T h e simple equations given earlier c an serve as a starting-point for analysing the effects of e x p a n d i ng education u p o n national p r o d u c t, a nd for distinguishing the private a n d non-private c o m p o n e n ts of s u ch c h a n g e s. B u t before presenting this possibility, it is well to e x a m i ne briefly the na ture of the previous w o r k that is m o st pertinent. W i t h apologies to economists w h o h a ve p r o d u c ed detailed a n d profound w o r k , I shall grossly oversimplify he r e, b ut wi th the h o p e that a ny readers w h o m a y b e interested b ut unfamiliar wi th the sources will e x a m i ne the w o r ks cited for themselves. T h e m o s t direct a t t empt to m e a s u re contributions of education to national p r o d u ct over t i me is that of Prof. T . W . Schultz. 1 In essence his m e t h o d w a s as follows. First, quantity of education e m b o d i e d ('carried') in the labour force w a s estimated in cost t e rms, using constant dollars a n d including in costs the opportunity costs of students' t i m e. T h e n estimates of internal rates of return to increments of schooling (his o w n estimates a n d others) 2 w e re applied to his t ime series of education carried in the l abour force to derive the contribution of educ a t ion to national produc t. Several sets of estimates resulted, o ne set for e a ch a s s u m ed c o m b i n a t i on of rates of return to the various levels of education. E a c h of these could be v i e w ed as a n a p p r o x ima t i on to A ' as I h a ve defined it, t h o u gh n o n e is the s a m e as A ' except in the ye ar us ed for deriving the rates of return. Deviations of Schultz's estimates f r om A ' arise f r om several causes. O n e of these is c h a n g es over t ime in rates of return, 3

another is shifts in the a ge compos i t ion of the various educational subgroups of the adult population, 4 a n d a third is c h a n g es in the relative imp o r t a n ce of on-the-job training. W h e t h e r or i. T . W . Schultz, 'Education and E c o n omic G r o w t h ', in: Nelson B . H e n ry (ed.), Social Forces influencing American Education, Chicago, University of Chicago Press, ig6i. (Sixtieth Yearbook of the National Society for the Study of Education, Part II.) 2. Schultz m a d e rough estimates for each educational level a nd compa r ed these with Ga ry Becker's estimates for college education. (See G a ry Becker, 'Unde r investment in College Education]', American Economic Review, V o l. L , N o . 2, M a y i960, p . 346.) 'Internal rates' as c omp u t ed by Becker are the discount rates that wo u ld equate the (differential) income stream with the estimated costs of the relevant education. 3. Schultz himself is fully awa re of this, of course. H o w e v e r, direct estimates of A ' are available for selected years only, beginning with 1939. By presenting estimates based on several alternative rates of return, Schultz comes up with a range of results that should cover error arising from changes in rates of return. 4 . This m a y be very important. Schultz disregards the age-education composition of the population, treating it as though there were no age shift within any educational category. O n this point see m y ' H u m a n Capital: Concepts and Me a sur e s ', o p. cit., a nd the papers by Becker and Minc er referred to in footnote 2, p . 6 4 9. 654 E C O N O M I C S O F E D U C A T I O N not these deviations are of quantitative importance over a ny given interval of time, it is clear that Schultz m a d e no attempt to include the effects of interactive influence; D ' is omitted. 1 Starting from the opposite position, several economists 2 h a ve c o m p a r ed estimates of national product over time wi th changes in capital inputs 3 a nd in quantity of labour, seeking to determine h o w m u c h of the increase in national product c an be explained thereby. T h e remaining, unexplained increases in product are then attributed to unidentified factors, a m o n g w h i ch are education, advances in technology, a nd changes in e c o n omic organization. O n e of these studies 4 attempted to identify a functional form for this 'third' composite variable, but without identifying its inputs except by imputation from its results. Like virtually all attempts at m e a s u r e m e n t, this necessarily assumes the persistence of s o me given basic pattern in e c o n omic structure within w h i ch growth is taking place. Attention is not directed specifically to education. T h e approa ch suggested in the following pages has less ma thema t i c al elegance than the aggregative input-output technique, but it is focused

u p o n educa'tion (including on-the-job training). A s in all attempts at m e a s u r e m e nt of returns to given kinds of inputs, s o me arbitrary assumptions are unavoidable ; ultimately the test of such assumptions m u st be in their reasonableness a nd their internal consistency. I h a ve tried to fuse the t wo approaches just s u m m a r i z e d, starting with the equations presented earlier. S ' w as defined as that part of total i n c ome attributable to education, m o re strictly as that part of earned i n c o me so attributable. It w as a s s umed tacitly that n o ne of the returns reported in empirical data as property incomes w as attributable to education, t h o u gh s o me part (presumably mos t) of the incomes from proprietorships w o u ld fall in the e a rned- income category. It is of course arbitrary, a nd even false, to a s s ume that earned incomes are independent of the quantity of capital; it is n o less false, h o w e v e r, to a s s u me that une a rned incomes are independent of the quantity a nd quality of labour. As a first approximation, the m o st reasonable a nd pragmatically feasible solution is to a s s u me that these interactions are mutually cancelling; the p r o b l em is then to identify A ' , S ', a nd D ' as c o m p o n e n ts of total earned i n c o m e, Y ' . i. This omission was deliberate, and not due merely to measurement difficulties. Schultz was presenting estimates that should stand up as minimal, avoiding possible upwa rd biases. 2. A n u m b er of studies of this kind have been sponsored by the National Bureau of Economic Research; prominent a m o ng the pioneers have been Kuznets, Goldsmith, Abramovitz, Kendrik, and Fabricant. See also O . Aukrust and Juul Bjerke, Real Capital in Norway, International Association for Research in Income and Wealth. Bowes and Bowe s, L o n d o n, 1959 (Income and Wealth series, VIII), a nd Robert M . Solow, 'Technical Change and the Aggregate Production Function', Review of Economics and Statistics, August 1957. 3. T h e capital input measures used are inappropriate, problems of change in the quality of physical capital aside, because they are in fact measures of stores of values. This fact is clearly seen by Solow, op. cit. 4. Solow, op. cit. 655 I N T E R N A T I O N A L S O C I A L S C I E N C E J O U R N A L O f necessity, Y ' = A ' + N 6 ' ; it is the s um of earned incomes above the educational baseline and at that line (simplifying here to take the baseline average income as that accruing to people with ' n o' education, a nd including in A ' the total of all income differentials above that level). E a ch of these values, Y ' , A ' , a nd N>' is empirically observable, at least in principle. C o m p a r i ng these values for period t i with those for period t, w e derive their change between the two periods: A Y ' = A A ' + A N o ' . Simplifying by assuming N to be constant, w e automatically correct for changes in the quantity of labour. Empirically this correction can easily be m a d e . Unfortunately, this equation does not give us w h at w e w a n t, h owe v e r. W e need to k n ow the b r e a k d own of N e ' into Nb'0 a nd Nb'e, the latter being the measure of D ' whi ch in combination with A ' would give us S '. T h at is, w e need to fill in the values in the equation A Y ' = A A ' + A Nb'0

+ A Nb'e, or A Y ' = A S' + A N 6 ' 0 . It will be recalled that b'0 is that part of baseline average income not attributable to the indirect effects of activities of the educated population u p on incomes of the uneducated. I suggest that while it is difficult to supply a reasoned analysis of N 6 ' 0 , the most plausible value for A N 6 ' 0 is zero; this amo u n ts to assuming that all changes in b' over the time interval under examination stem indirectly from education (including that on the job), or that any changes in b' that are not attributable to education are neutralized in the total by net contributions of education to unearned incomes that have already be en excluded. T h e assumption that A N ' 0 is zero gives a m a x i m al estimate of S' only if net contributions of education to unearned incomes are zero or negative. CHARACTERISTICS A N D DETERMINANTS OF NON- P R IVATE M O N E T A RY RETURNS It w as suggested earlier that, viewing education as a whole, D ' is likely to be positive; it would be better to say that A D ' is positive. This conclusion follows from the fact that b' has typically increased over time a nd from the narrowing of the ratio of incomes of the educated to the uneducated even though capital labour ratios associated with jobs held by the uneducated have almost certainly changed less than in the e c o n o my as a whol e. E v en if A N i 'o is positive a nd significantly above zero, a sizeable positive A D ' is indicated. If this is indeed the case, it follows that A A ' significantly understates A S '. T h o u g h Schultz's measure is not quite our A ' , this conclusion wo u ld apply equally to his estimates. T h at is, even excluding non-monetary private a nd non-private returns, his me t h od wo u ld then yield estimates that are inherently conservative unless he used grossly exaggerated internal rates of return. I suspect o ne could identify m a n y situations in whi ch A D ' has be en both positive a nd large, a m o n g t h em the northern a nd m o re recently the southern sections of the United States. 656 E C O N O M I C S OF E D U C A T I O N H o w e v e r , there is currently a widespread debate in this country, o ne party to whi ch wo u ld in effect maintain that there is a tendency to a negative A D ' , that is, a negative social-private discrepancy, with respect to recent expansions in post-elementary education. T h e debate has not dealt directly with the question of social versus private returns to education, but focuses at the m o m e n t on the question as to whether there is a secular tendency to growing u n e m p l o y m e nt a m o n g n e w entrants to the labour ma rket w h o have not completed secondary schooling. If the answer is yes, it is further arguable that this could be at least partially explained by expansion in the proportion of youth completing secondary school. First, employers seem to have developed a strong preference for secondary school graduates over those with less schooling, whether the former actually are m o re productive or not. Second, it m a y be that the approach to universal secondary schooling is stimulating a restructuring of the e c o n o my through general adoption of m o re refined technologies. Taking averages per m e m b e r of the labour force, not per employed worker, empirical measures w o u ld

s h ow a drop in average baseline incomes, V, under existing circumstances; if this is not merely temporary, the higher u n e m p l o y m e nt rate should of course be allowed for in measuring b'. 1 If this decline in b' is legitimately imput ed to the spread in secondary schooling, then A D ' , is evidently negative and A A ' s is an overstatement of A S'. N o one working on the economics of education has as yet attempted to consider the implications of such values of A D ' for generalizations about social over- or under-investment in schooling. Attempts have indeed b e en m a d e to c omp a re private a nd 'social' internal rates of return, a nd ,to c o m p a re each of these with rates of return on corporate investment. 2 H o w e v e r , 'social' rates as estimated in these studies do not capture the c omp o n e n ts of D ' . T h e rate are 'social' only in that the returns analysed are pre-tax and the costs with whi ch they are c omp a r ed to get internal rates of return are inclusive of public as well as private expenditures on education. Failure to take D ' into account is hardly surprising in view of the difficulty of isolating it empirically, but I w o n d er whether the assumptions that m u st be m a d e to do this are any less arbitrary than the tacit assumptions that a'j, a'k . . are attributable solely to education, or so attributable after m a k i ng s o me adjustment for ability, effort, etc., associated with education. Conc eding that distortions or at least artificialities are introduced in a ny available set of assumptions by whi ch elements in an e c o n o my are segregated from the general outline, it is time that a careful look be taken at the range within wh i ch A D ' m a y fall under various conditions. T h e re is a second a nd even m o re important proposition that is sharply pointed up by recent developments. This is the whole problem of h o w to treat structural change. Putting this another w a y , returns to education i. For a discussion of the treatment of unemployment in related connections see m y ' H u m a n Capital: Concepts and Measures'. 2. Primarily by Gary Becker and Lee Hansen, in unpublished work. However, see Gary Becker, 'Underinvestment in College Education', op. cit. 657 I N T E R N A T I O N A L S O C I A L S C I E N C E J O U R N A L depend in large degree u p on h o w the educated people are used, a nd h o w they are used is only in part a function of the distribution of education within the population. T h e problem of assessing returns to education (either social or private) is of course aggravated. But this is not all. Potential social returns from investment in education m a y be quite different from realized returns. Examination of the divergence between these a nd the causes of these deviations m a y have important policy implications having little connexion with the specific problem of allocating investments to education. A M O RE D Y N A M IC VI EW

T h e static assumptions of the m o d el were relaxed only partially in the last section. 'Organization' factors were brushed aside, yet m a n y writers have attached great significance to various external economies or economies of scale in bringing about e conomic growth. Neither have I given m o re than passing attention to intergeneration effects of parental education. Finally, technological advance has crept into the analysis without explicit recognition. These elements b e c o me of increasing importance for total i n c ome growth w h e n comparisons are m a d e over cumulative time intervals. In the foregoing section s o me part of these w as absorbed into unearned incomes, the rest into earned incomes (by assumption into S ' only, not Nb'). It is often a nd justifiably argued that it is a gross exaggeration to attribute to education all the increase in national income not explained by increase in the size of the labour force a nd the quantity of physical capital. This objection is well founded, especially w h e n one considers h o w m a n y improvements in the quality of physical capital escape me a s u r eme n t. It does not follow, howeve r, that the opposite position is tenable, n ame l y, that only w h at is included in s o me current estimate of our A ' should be attributed to education. Quite clearly the contribution of parental education to the future efficiency of children is a deferred return to education that m u st be accounted for in taking a long-term view. Identifying the part of technological advance or of economies of scale to be credited to education is m o re difficult. Indeed, w e lack even approximate measures of ' a m o u n t s' of technical innovation, a nd w e have almost no insight into the relative part of sheer ability versus systematic effort in invention or productive adaptation of inventions. A considerable part of the effect of education u p on economic growth lies in this sphere of innovation, a nd these contributions, like the noneducational elements in technical advanc e, eme r ge eventually in both earned a nd property incomes. Technological imp r o v eme nt a nd economics of scale tend to go hand-inh a nd and both involve structural transformations. Taking a long view (which m a y not be m a n y years if educational advance is rapid), the interplay between education a nd economic growth is likely to involve economies 658 E C O N O M I C S O F E D U C A T I O N of scale and major structural changes. 1 I suggest that m a n y of these transformations depend on the spread of one or another level of education as a necessary, but emphatically not in most cases as a sufficient, condition. 2 S o m e of the observed and confusing leads a nd lags in the relations between educational a nd economic advance m a y well turn on this point. Major steps in economic growth m a y be blocked until both the me d i an level of schooling and the distribution of schooling a m o n g the parts of the p o p ulation have passed certain thresholds. Perhaps less attention should be paid to the relation of education to individual earning powe r, a nd m o re to its

role in facilitating the particular processes that go together under the umbrella phrase 'external economies'. T he fact that economic growth, especially in the early stages, normally means a considerable shift from subsistence to market activities, introduces a major bias into estimates of advance based on monetary measures alone. Just as in attempting to compare national incomes or per capita incomes among countries, it is obviously necessary to adjust for this, so that the real coverage of monetary measures will remain as nearly comparable over time as possible. See M a ry Jean B o w m an and C . Arnold Anderson, 'Concerning the Role of Education in Development', to be published in a forthcoming symposium sponsored by the Committee for the Comparative Study of N e w Nations, University of Chicago. Also, by the same authors, ' T he Role of Education in Development', in: Development of the Emerging Countries, T he Brookings Institution, 1962. 659 THE CONCEPT OF H U M AN CAPITAL M . D E B E A U V A IS T h e term ' h u m a n capital' is of comparatively recent origin a nd relates ta a concept whi ch is still difficult to define scientifically or to analyse from the quantitative standpoint, for it associates two ideas wh i ch had hitherto been regarded as different in kind a nd h ad been studied under different disciplines. As used by the economist, the w o rd 'capital' covers a category of goods wh i ch are called 'productive' because they have the property of producing other goods; m a n , as the producer of these goods, plays his part as the ' m a n p o w e r ' factor w h i ch m u st be associated with the 'capital' factor in order to initiate the process whe r eby wealth is produced. Regarded from this angle, the t wo are independent variables. F r om the standpoint of the other branches of humanistic studies, m a n m a y be a subject for research a nd me a s u r eme n t, but analysis in these disciplines is focused on his individual or social behaviour, and not on his wealth-producing function, whi ch is seen as an aspect of homo economicus to be studied by the political economist. T h e education experts consider the essential a im of education to be the highest development of the intellectual a nd moral aptitudes of the individual, rather than the training of producers. As in m a n y other branches of social science, the division into these separate categories is at present being called in question on various sides. In economics, research on the long-range growth of the A m e r i c an national product recently carried out by the National Bureau of E c o n o m ic Research has s h o wn that only a small proportion'of economic expansion in the United States (from 30 to 50 per cent, according to the calculation me thods adopted) c an be explained in terms of the classical factors of production (capital, m a n p o w er a nd natural resources). This has led to a search for other factors of growth hitherto left out of account by the analysts. It is generally agreed that one of the most important of these factors appears to be the rise in the level of skill of the labour force resulting from the rise in general level of training a nd education. O f course, other factors

of growth have been suggested, including technical progress and improved organization in individual firms and at the national level. All these elements are held to promote growth, but the difficulty is to determine the part 660 E C O N O M I C S O F E D U C A T I O N played by each of t h em separately, a nd to isolate t h em from o ne another; this is one of the newest fields of investigation for economic research. If w e consider only the 'education factor', it seems obvious that highly skilled labour will be m o re productive than unskilled labour, a nd that an hour's w o rk by a trained worker w h o has h ad six years of primary education followed by three years of technical education will produce m o re goods, for instance, than an hour's w o rk by an illiterate worker in the early nineteenth century. It might s e em s o m e w h at surprising that such a c o m m o n - s e n se h y p othesis should not h a ve be en accepted by economists until quite recently. W e m a y account for this pa r adox by giving a highly simplified outline of the m a n n er in w h i ch e conomic theory has vi ewed m a n as a producer. In considering the process of the production of wealth, the economists h a ve concentrated on the factor of physical capital, on the implicit a s s u m ption that the m e n ne eded to exploit that capital w o u ld be available in the labour market at a ny m o m e n t , equipped with the necessary skills. In the event of a scarcity of m a n p o w e r in a particular sector, w a g e variations w o u ld suffice to restore the balance between supply a nd d e m a n d . T h e p h e n o m e n a whi ch mitigate the effects of competition a nd w h i ch are the result of the organization of the workers in trade unionsthus increasing their bargaining powerthe incidence of u n e m p l o y m e nt a nd the existence of monopolies are all viewed as modifying this theoretical situation to s o me extent, but without radically invalidating it. In planning, the desired increase in the product (over-all or sector) is analysed solely in terms of physical capitalthat is, in the light of the a m o u n t of additional investment required to obtain an additional unit product. This ratio, k n o w n as the capital-output ratio, w as adopted, for instance, in assessing the capital needs of the underdeveloped countries (in the light of population growth a nd the increase d e e m ed desirable in per capita income) a nd the a m o u n t of foreign aid wh i ch w o u ld be necessary, the latter being equal to the difference be twe en the total investment figure thus estimated a nd the v o l u me of domestic savings regarded as possible. 1 In attempting to define the role attributed to m a n in these theories of growth, w e find that, as in so-called 'classical' economics, calculations concerning producers are m a d e in elementary form, i.e., in terms of n u m b e rs of workers or in m a n - h o u r s. T h e concept of productivity is used in their e conomic forecasting for the purpose of apportioning the general labour force between the different sectors of activity, a nd productivity is me a sur ed by compa r ing the output of each sector (expressed in a d d ed value or in physical units) with the n u m b e r of workers (expressed

in units) in the branch of activity concerned or with the n u m b er of m a n i. See in particular: Lon T a b a h, in: Population, N o . 27, I N E D , Paris, 1956; Millikan and Rostow, A Proposal, N e w York, 1957; Paul Hoffman, 100 Countries, N e w York, 1961. 661 I N T E R N A T I O N A L S O C I A L S C I E N C E J O U R N A L hours. T h e variations in this ratio whi ch are noted over a given lapse o time express the increase in productivity a nd are extrapolated (linearly) for the period of the plan. It thus be comes possible, on the basis of the anticipated increase in sector output and of a hypothesis for an increase in productivity, to estimate the n u m b er of workers required for the sector concerned. It will be seen that this m e t h o d, wh i ch in a ny case is very imperfect, considers the individual as an interchangeable unit, without regard to his qualifications, a nd as the end-product of a calculation whi ch w o u ld appear to measure the degree of mechanization of the production process, or the technical progress achieved, rather than the productivity of the individual worker. Marxist economic theory puts forward a different concept of the relations between capital and the worker. M a r x , in developing Ricardo's theory that labour is the source of value, touched u p on the probl em of skilled labour, explaining that this w as distinguished from 'average social labour' by higher training costs. Since the 'production' of a skilled worker costs m o re in terms of working time, the value of his labour p ower is greater than that of the unskilled w o r k m a n , a nd m a y be expressed as a multiple of this labour p o w e r. H o w e v e r , M a r x carried his analysis no further in Das Kapital, merely pointing out that this distinction between skilled and unskilled labour is of little practical interest, since the effect of capitalist industrial techniques, according to h i m, is to reduce the importance of skilled labour by spreading assembly-line techniques in whi ch the worker is subordinated to machines. This historical development of the workers' status during the first phase of the Industrial Revolution helps to explain w h y economic theory, whether classical or Marxist, considered the labour force from a purely quantitative angle, without taking professional skill into account. Mo r e o v e r, as compulsory primary education b e c a me m o re widespread, the provision of the elementary education required of the worker w as taken over by the c o m m u n i t y, whereas in earlier times vocational training h ad been the responsibility of the guilds. T h e result w as that the nineteenth-century industrialists a nd economists we re apt to ignore expenditure on training, wh i ch w as not included in the calculation of a firm's production costs. T h u s, neither training costs nor the effect on economic growth of a rise in training standards were taken into consideration, so that the term ' h u m a n capital' w as meaningless to the economist. It w as in a completely different context that the importance of the skilled worker in the production process recently c a me to oc cupy the forefront of attentionwhen an acute shortage of scientists, engineers

a nd technicians b e c a me evident. It w as n o w realized that the labour market w as in no position to provide the highly skilled m a n p o w er needed for accelerated industrial expansion at a m o m e n t ' s notice, a nd that w a ge increases no longer sufficed as a m e a ns of adjusting the growing d e m a n d for highly skilled personnel to the supply, wh i ch w as limited in the last 662 E C O N O M I C S OF E D U C A T I O N resort b y the educational system. Industrial expansion, accelerated technological progress a n d the d e v e l o pme nt of n e w industries are all contributing factors of the increased d e m a n d for engineers a nd technicians, not only in the production process itself but also, a nd especially, in offices a n d laboratories. M o r e o v e r, scientific a n d technical personnel are seen n o longer as merely a short-supply area, b ut as a determining factor for the technical progress necessary to g r o w th (pure a nd applied research, a nd the perfecting of n e w products a nd n e w manufacturing processes). A t the industrial stage, at a ny rate, innovation has ceased to be regarded as the result of unforeseeable c h a n c e; it d e p e n ds u p o n the considered policy of the undertakings conc e rned; training for personnel a nd the expansion of research services are a m o n g their constant concerns a nd are reflected in their costs. O n the national scale, the supply of scientists, engineers a n d technicians is recognized as an essential element of e c o n omic p o w e r. T h e competition b e twe en East a nd W e s t, especially since the conquest of space, has underscored the decisive part played by the production of such highly skilled personnel, a nd the 'challenge' is n o longer confined to the v o l u me of output or to the rate of investment. N u m e r o u s studies h a ve accordingly b e en m a d e , particularly in the Uni t ed States a nd by O E C D , with a v i ew to estimating national requirements in scientific a nd technical personnel. T h e y serve the useful purpose of perfecting the existing m e t h o ds of forecasting m a n p o w e r requirements a nd s h owi ng the necessity of adapting educational systems to those requirements, but suffer from the d r a w b a c k, as far as the present p r o b l em is concerned, of covering only a very limited s e gme nt of the total labour force. Before w e c an legitimately speak of h u m a n capital, w e should be able to a n swer t wo preliminary questions: Is a higher level of skill on the part of the labour force as a w h o le (and not me r e ly a privileged section of it) to be regarded as a condition of g r o w th ? Is this higher level of skill merely a necessary condition for e c o n omic g r owt h, or is it a positive factor of deve lopment? T h e experience of history suggests that the first of these questions c an be answe r ed affirmatively. It is found that the general level of training rises simultaneously with the e c o n omic level, a nd comparisons b e twe en different countries indicate a certain correlation b e twe en the share of the national i n c o me devoted to education a nd per capita i n c o m e . 1 It w o u ld be possible in this w a y to c o m p a re the increase in national expenditure on education over a given period with the increase in the national product, either during the s a me period or with a time-lag to

allow for the effects to m a t u r e. B ut such comparisons are still unreliable, for the a m o u n t spent on education is still not accurately k n o w n , because of the difficulty of estimating the cost of private education. Exc ept in a few countries w h e re these data h a ve b e en assembled in the past few years i. F . Edding, Internationale Tendenzen in der Entwicklung der Ausgaben fr Schulen und Hochschulen, Kiel, 1958. 663 I N T E R N A T I O N A L S O C I A L S C I E N C E J O U R N A L (United States, United K i n g d o m, France, Italy 1 ), the figures on whi ch international comparisons are based are d r a wn from Une s co's statistics, wh i ch cover only public expenditure on education a n d, in mo st cases, only expenditure by ministries of education. Analysis of the development of the vertical structure of e m p l o y m e nt points to the s a me conclusion: in industry, technological progress increases the d e m a n d for skilled a nd highly skilled workers, even in regressive sectors, w h e re the total n u m b er of workers tends to diminish. T h e s a me tendency is to be observed in the public services, as is s h o wn by the higher educational level d e m a n d ed of senior, intermediate or junior personnel. H o w e v e r , studies conducted on the ma c r o - e c o n omic scale with a view to analysing historical changes in the general level of training of the labour force have still to be m a d e . T h e second question goes right to the heart of the probl em, for it is no longer concerned merely with establishing a correlation between the economic system a nd educationa correlation wh i ch could, in fact, be interpreted in two w a ys from the standpoint of causality: either education is one of the causes of e conomic growth, or else the increase in the national product m a k es it possible to increase the expenditure on education. N o r is it sufficient to s h ow that m a n p o w e r is indispensable to the development of physical capital, a nd that to obtain a desired degree of growth, additional m a n p o w er m u st be trained u p to the various levels of skill a nd in the n u m b e rs required by the technical composition of the n e w capital. For by doing this w e should merely be showing that the h u m a n factor is a negative condition of development, in the sense that a shortage of m a n p ower at certain levels of skill constitutes a bottle-neck impeding growth. It might then be argued that once the balance be twe en the supply a nd d e m a n d for skilled workers h ad been restored, the o p t i m um educational level determined by the general technical level of a given country would be exceeded ; a ny subsequent rise in the level of training w o u ld then cease to contribute to further e conomic development, a nd w o u ld merely be economically unproductive expenditure. Another limiting factor might also be invokedthat of ability, the effect of whi ch wo u ld be to apply to education a law of diminishing returns under wh i ch the rise in the level of training for an ever-increasing proportion of the active population wo u ld involve ever higher expenditure. F or the m o m e n t these are purely theoretical problems, for even the most a d v a n c ed industrial countries s e em

to be a long w a y off these limits. Nevertheless, they have a certain interest w h e n it c o m es to an attempt to define the concept of h u m a n capital a nd study the relationship be twe en economic growth a nd m a n p o w e r , taking into account the level of skill. Until such time as the probl em can be posited in measurable quantitative terms, all that can be stated is that the i. See, in particular: (for United States) T . W . Schultz, in: Journal of Political Economy, December ig6o; (for France) I N S E E, Cot et Dveloppement de l'Enseignement en France, Paris, I N S E E, 1958; (for the United Ki n g d om) John Vaizey, The Costs of Education, London, Allen and U n w i n, 1958. 664 E C O N O M I C S O F E D U C A T I O N role of the m o re highly skilled personnel is not confined to exploiting physical capital; they are also responsible for conceiving a nd perfecting the inventions or innovations w h i ch push b a ck the frontiers of present-day production techniques a nd thus m a k e further g r owth possible. N o n e of the hypotheses a d v a n c ed here c an at present be confirmed by quantitative analysis, but they all help to guide research in n e w directions. In the field of education, the concept of h u m a n capital originated with the first attempt to establish a connexion b e twe en educational a nd econ o m ic development. In the underdeveloped countries, a nd in certain industrialized countries, the spontaneous or deliberate increase in school enrolm e nt has resulted in the application to education of the techniques of e conomic forecasting a nd planning. W o r k along these lines has so far be en directed mainly to calculating the cost of education in order to w o rk out operating or investment costs over periods of several years, in the light of the anticipated increase in enrolment. Since this objective has m u c h in c o m m o n with those of the research already described, various economists we re p r omp t ed to start considering the subject of education from the standpoint of financing. O n e practical result of this w o rk is already a p p arent: in countries w h i ch have d r a wn up an educational plan a nd put it into application, the establishment of targets expressed in precise figures has led to a m o re rapid increase in educational expenditure than in other countries; this is the case, for e x amp l e, in Franc e, Tunisia, M e x i co a nd India. Requests for funds have b e en based on precise calculations of the current a nd non-recurring expenditure needed in order to me et the selected targets, a nd this has given the ministries of education a basis for discussion with their respective treasuries w h i ch afforded t h em better prospects of success than w as formerly the case with the traditional type of requests for increased appropriations submitted annually by each recipient of funds from the State budget. But once educational targets b e g an to be fixed in the light of educational requirements a nd no longer in order to produce an automatic increase in the previous figure for expenditure, the p r o b l em of finance took on quite different dimensions; indeed, the estimates w h i ch h a ve been tentatively m a d e at the regional level are so large that they far exceed the possibilities of the countries concerned c o m b i n ed with those of outside aid. T h e Ka r a c hi Conference organized by U n e s co (De c emb er i960) m a d e an estimate of the cost of enrolling the entire school-age population of Asia (exclusive

of Ch i n a) by 1980; this s h o w ed that it w o u ld take $36,000 million to achieve satisfactory enrolment in primary education alone up to that date. For the p r o g r a m me adopted at the Addis A b a b a Conference ( M a y 1961) for the development of primary, secondary a nd higher education in tropical Africa, the present expenditure on education, estimated at $500 million in 1961 (including $ 150 million in the form of outside aid), w o u ld h a ve to be increased to $ 1 , 1 54 million in 1965 and $2,600 million in 1980. T h e Santiago Conference ( M a r ch 1962) yielded information on the basis of w h i ch a similar estimate can be m a d e of Latin A m e r i c an needs in the 665 I N T E R N A T I O N A L S O C I A L S C I E N C E J O U R N A L matter of educational finance; it indicates an analogous increase. It is possible, of course, to question the calculation me thods by wh i ch these figures were arrived at, just as the use of the capital-output ratio in estimating the total v o l ume of investments required to ensure growth in the underdeveloped countries has also been questioned. But the m a in value of these regional estimates is that they specify the order of magni tude of the expenditure w h i ch would be necessary to m e et m i n i m u m educational requirements if the me thods hitherto followed are maintained. T h e impossibility of meeting such d e m a n ds leads to a search for criteria enabling an order of priority to be established in the light of available resources. W h a t proportion of the national i n c ome should be devoted to education? W h a t should be the choice of objectives within the educational system itself? The se problems bring us back to the questions already raised at the economic level. It is no longer merely a question of estimating the cost of education in terms of desirable a ims; w e have to decide whether education is to be regarded, in national accounting, as an item of c o n s umer expenditure, or as a social investment, or as a productive investment. O n ly in this last hypothesis w o u ld it be proper to speak of h u m a n capital, for educational expenditure wo u ld in that case b e c o me the cost of training the individual in his capacity as a factor of economic growth. If this concept is adopted, it will be possible to apply a rational criterion in m a k i ng appropriations for education, d ue account being taken of the country's financial resources, the stage of development it has reached a nd the econ o m ic targets it has set itself. W e wo u ld thus have a m e t h od of assessing the social return of the various investment projects, including those concerned with education, whi ch w o u ld be infinitely m o re satisfactory than the m e t h od of 'expediency costs', wh e r e by the value of a social investment (education, housing, public health, etc.) is assessed by comparing it with the productive investments wh i ch must be given up in order to m a k e w a y for it. T h e problems involved in research into e conomic growth, the dearth of highly skilled personnel a nd the development of education all help to give definition to the concept of h u m a n capital. H o w e v e r , n o ne of these me thods of approach at present enables that concept to be satisfactorily analysed. In consequence, research along these various lines employs the concept as a working hypothesis a nd not as a scientifically established

economic category. W e shall n o w describe briefly s o me of the attempts n o w being m a d e to confirm this hypothesis. f T h e over-all m e t h o d, referred to earlier, seeks to discover a correlation \ between total expenditure on training a nd the gross national product, or between an increase in educational expenditure a nd an increase in that product; a nd it presents both practical a nd theoretical difficulties. If series are established covering a long period, account will have to be taken of structural changes: the quality of the education provided, the extent and nature of the knowledge acquired, the combination of the factors involved a nd their respective costs (buildings, equipment, salaries) m a y have 666 E C O N O M I C S O F E D U C A T I O N developed along different lines; in considering the gross national product, account will have to be taken of changes in the e conomic situationmedium a nd long-term cycleswhich affect the relationship with educational expenditure w h i ch is less sensitive to such fluctuations. 1 Pending such time as these obstacles c an be o v e r c ome by constructing h o m o g e n e o us series (by weighting the educational quantities a nd calculating trends so as to leave fluctuations out of account), it will nevertheless be possible to relate educational expenditure to the gross national product, probably in the form of an elasticity ratio, although this will not answer the question whether education is to be regarded as consumption expenditure or as investment expenditure. If this kind of research is to confirm the hypothesis of ' h u m a n capital', the necessary time w o u ld have to be allowed for the investments to m a t u re (the formation of individuals takes longer than that of physical capital) a nd a criterion established for the b r e a k - d o wn of educational expenditure into the portion w h i ch meets c o n s umer needs a nd that wh i ch covers vocational training. H o w e v e r , w e still h a ve a long w a y to go before w e c an calculate a capital-output ratio applicable to individual m a n p o w e r for the purpose of determining the v o l u me of educational investment necessary to obtain a given increase in the national product. Compa r i sons be twe en different countries are also of great interest, but they, too, provide no scientific answer to the probl em raised. Compa r a t ive tables h a ve b e en d r a wn u p by U n e s co a nd O E C D 2 to s h ow the relationship be twe en educational expenditure (or school enrolment) a nd the gross national product, or be twe en educational expenditure a nd the national budget, total public expenditure or total gross or net investment. W h a t they m a k e it possible to d o, chiefly, is to c o m p a re the efforts m a d e by individual countries in the field of education, for it is only if other factors are included to account for the distribution of the figures that correlation b e c o m es significant. 3 T h e comparisons are useful for assessing the possibilities of increasing

educational expenditure, or for setting national or regional targets, but not as a m e a ns of establishing that the gross national product, or its annual rate of increase, is in direct ratio to the share of the national i n c o me devoted to education. T h e se investigations undoubtedly a dd to our knowl edge of educational expenditure in different countries, a nd are encour agement to greater national efforts, but they cannot be used as an a r g ume nt to s h ow that e conomic growth depends o n the formation of h u m a n capital. T h at proposition w o u ld imply, in fact, that all educational expenditure produces the s a me effect: in other wo r d s, that educational systems are incapable of i. See John Vaizey, op. cit. 2. O E C D , Policies of Economic Growth and Investment in Education, Paris, 1962. 3. Reference should also be m a de to the work of M r . M . Ramirez, which establishes, for nine countries, a close correlation between the average n umb er of years of school attendance completed by the active population and the level of productivity. 667 I N T E R N A T I O N A L S O C I A L S C I E N C E J O U R N A L i m p r o v e m e n t. O n the contrary, it w o u ld a p p e ar that o ne of the essential purposes of educational d e v e l o p m e nt is to m a k e the m o st effective use of the resources devot ed to it. T h e o p t i m u m in this respect s e ems b y n o m e a n s to h a ve b e en attained, w h e t h er in pedagogical efficiency (internal criteria) or e c o n omic efficiency (external criteria). It is in this direction that e c o n o m ic research should m a k e its m a i n contribution. Efforts h a ve b e en m a d e , taking another line, to express h u m a n capital in quantitative terms, o n the assumption that the over-all level of vocational skill in a country c an be m e a s u r ed by the n u m b e r of years of study compl e t ed b y the w h o le of the active population. Professor Schultz 1 has established, for e x a m p l e, that the 'stock of education' of the labour force in the U n i t ed States rose f r om 2 16 million years of study in 1 9 00 (an ave r age of 7.7 years per wo r k e r) to 7 76 million years in 1 9 57 (an average of 11 years). W e i g h t i ng this with a coefficient for 'equivalent years of study', totalling 152 d a ys of school attendance a year, 2 w e obtain a ratio of 116 million years in 1 9 00 to 7 40 million years in 1 9 5 7, or an a n n u al increase of 3 .3 per cent; b ut according to the classical analysis of the factors of e c o n o m ic g r o w t h, only the increase in the labour force (me a sur ed b y the adjusted n u m b e r of m a n - h o u r s w o r k e d) is taken into a c countan increase w h i ch a m o u n ts to only 0 .8 per cent a year for the s a me period. This 'stock of education' c an also be expressed in m o n e t a ry t e rms, if w e calculate the total operating a nd investment expenditure for all types of education (primary, secondary a nd higher) in the public a nd private sectors, following the classification a d o p t ed b y U n e s c o . 3 T o t al expenditure

o n education is then divided b e t w e en the inactive a nd active sections of the population, a nd the figure thus obtained expresses the cost of production of the stock of education of the labour force, w h i ch Professor Schultz estimates for the U n i t ed States of A m e r i ca at $ 2 4 8 , 0 00 million for 1 9 40 a nd $ 5 3 5 , 0 00 million for 1 9 4 7, on the basis of constant 1956 prices. T h is gives a series w h i ch c an be included in the e c o n o m ic calculation together wi th the series for investments in physical capital. B u t several questions n e ed to be c o nsidered. 1. Is it possible to regard all expenditure o n the education of future wo r k e rs as investment expenditure? In v i ew of the impossibility of distinguishing the proportion of education w h i ch satisfies cultural ne eds from that which confers vocational skill, it s e ems preferable to consider that all education given to the labour force helps to determine the technical skill of the producers, just as the w h o le of that s a me education helps to 1. T . W . Schultz, 'Education and Economic G r o w t h ', in: Nelson B . Henry (ed.), Social Forces influencing American Education, Chicago, University of Chicago Press, 1961. (Sixtieth Yearbook of the National Society for the Study of Education, Part II.) 2. According to a me thod described by C . D . L o ng in The Labor Force under Changing Income and Employment, Princeton University Press, 1958. 3. Manual of Educational Statistics, Paris, Une s co, 1961, p . 2 4 0. A m o re precise me thod, enabling an analysis to be m a d e of the economic components of a given system of education, has been proposed by I. Svennilson ( O E C D , o p. cit., p . G 9 ). 668 E C O N O M I C S O F E D U C A T I O N raise their cultural level. Education performs this dual function simultaneously for the individual a nd for the e c o n o m y. 2. Should account be taken of ' i n c ome foregone', that is, of the wages a student mi g ht have received h ad he entered the labour ma rket i m m e d i ately instead of continuing his studies? T h e importance of this question lies in the fact that this ' i n c ome foregone', according to Professor Schultz's estimates, appears to exceed the direct cost of education. Objections have been m a d e to this estimate of indirect (or rather fictitious) costs: on the scale of the national e c o n o m y, it w o u ld not s e em justified to set aggregates such as the gross national product or the real expenditure on education against a s u m representing w o rk not d o n e; a nd even on the individual scale, decisions concerning the extension of studies involve m a n y other factors of a psychological a nd sociological nature wh i ch are at least as powerful as e conomic motives. Mo r e o v e r, the possibility of choice is restricted, chiefly by the supply of education (places available at schools), selection me t h o d s, family living standard, etc. It is true, w h e re planning is concerned, that State decisions on matters such as the extension of compulsory schooling or the increase in the n u m b e r of places available at schools h a ve the effect of diminishing the labour force ; but all that is necessary in that case is to allow for these external factors in forecasts of the active population, without bringing in the concept o f ' i n c ome foregone' or having to face the difficulties involved in estimating it.

3. A fundamental objection to the attempt to express the stock of h u m a n capital in terms of quantity m a y also be m a d e by citing individual differences, such as unequal abilities for w h i ch the n u m b er of years of education provided for each individual are no yardstick; by ascribing the major part of vocational training to the experience acquired in active life; or by adducing a tendency for the general level of studies to decline, or, alternatively, to rise. W e shall not go into the details of these argument s, but merely point out that e conomic calculations relate only to w h at can be measured in mone t a ry terms, a nd that on the m a c r oeconomic scale it is averages that matter. O n c e w e begin to study m a n as a form of capital, w e mu st consider not only the cost of training h i m, but also his e conomic profitability. T h o u g h it is true that m e n cannot be bought, like ma chine s, an attempt c an be m a d e to me a sure their profitability in the light of the wages they receive. For the United States, a comparison between the cost of educating a m a n a n d the total a m o u n t he will receive in wages during his active life has been m a d e by G . S. Becker 1 for the three levels of primary, secondary a nd higher education. Becker estimates that the 'rate of interest' is 14.3 per cent for a person w h o has continued his studies up to the e nd of secondary school, a nd 9 per cent for o ne w h o has continued t h em u p to the fourth university year. 1. G . S. Becker, 'Underinvestment in College Education', American Economic Review, Vol. L, N o . 2, M a y i960. 669 I N T E R N A T I O N A L S O C I A L S C I E N C E J O U R N A L S . Strumilin 1 m a d e use of a similar m e t h od as far b a ck as 1924 in calculating the profitability of education o n the basis of the differential w a g es of Soviet workers. This m e t h o d of evaluation is also described by B . Hoselitz 2 a n d F . Perroux. 3 Us i ng these m e t h o d s, it is possible to define the elements of the individual profitability of educational expenditure, provided that only that b o r ne by the individual is taken into a c count; but, in fact, the bulk of educational expenditure is borne by the c o m m u n i t y. F u r t h e rmo r e, these calculations of profitability are governed by a particular w a g e structure, unless it be a s s u m ed that w a g es always provide an accurate m e a s u re of the e c o n o m ic profitability of the labour of e a ch individual. F or these reasons it is difficult to transpose such studies of individual profitability to the m a c r o - e c o n o m ic scale, although this is the m a i n issue. At t emp ts h a ve also b e en m a d e to analyse the profitability of h u m a n capital, o n the national scale, by seeking to define the h u m a n factors of increased productivity. B . H o r v a t 4

distinguishes four c o m p o n e nt factorsindividual consumpt ion (inasmuch as it satisfies m i n i m u m h u m a n needs) ; health; social organization; k n o w l e d ge (education). In countries in process of deve lopment, the situation with regard to the first three of these factors constitutes an obstacle to d e v e l o pme n t, a nd a ny i m p r o v e m e nt in c o nsumpt ion (food, housing, etc.), health or social organization produces a rise in the level of productivity. B ut o n ce the threshold of satisfaction is reached, additional expenditure will no longer be reflected in increased productivity. In the case of education, o n the other h a n d, there w o u ld appear to be n o u p p er limit, a nd investment in that sector is a decisive factor of g r o w th in every society. V i e w ed thus, a ny increase in the national product d e p e n ds simultaneously on material investment a nd o n investment in the h u m a n factor, the latter representing, in the final analysis, the limiting factor for productive investment. T h u s the concept of capital absorption capacity s e ems to be linked to the level of d e v e l o pme nt of the h u m a n factor, in that the v o l u me of investment is determined by the stock of education of the country concerned. Before they c an b e c o me the subject of quantitative calculations these propositions call for further investigation of the conditions governing the satisfaction of m i n i m u m h u m a n ne eds. T h e y o p en the w a y to the inclusion of h u m a n costs, in Professor Perrouxs sense of the t e rm, in considering e c o n omic d e v e l o pme nt policies. T h e analysis of m a n p o w e r requirements is another field in w h i ch e conomi cs a nd education m e e t. W h i le it is not yet possible to calculate a ' h u m a n capital output ratio' w h i ch w o u ld m a k e it possible to estimate 1. S. Strumilin, La Planification en U.R.S.S., Paris, 1947. 2. B . Hoselitz, in: Tiers-Monde, N o . 1-2, Paris, i960. 3. F . Perroux, 'Note sur le Niveau d'Aspiration et le Niveau d'Attente', Revue des Sciences conomiques, September 1961, p. 9. 4 . Branco Horvat, ' T he O p t i m um Rate of Investment', Economie Journal, De c emb er I958670 E C O N O M I C S OF E D U C A T I O N the v o l u me of investment in training required to reach a given target of g r o w t h, attempts h a ve at least b e en m a d e to ascertain shortages of skilled personnel a nd to define short-term a n d long-term requirements so that a n order of priority c an be established in the field of education. T h e International L a b o ur Organisation has for several years b e en e n g a g ed in standardizing the classification of occupations, 1 using this m e t h o d of analysis for estimating the needs of the e c o n o my as far as vocational training is concerned. F or the purpose of m e d i u m - t e rm forecasting, countries h a v i ng centralized e conomi es use the m e t h od of balances in d r a w i ng u p m a n p o w e r plans in accordance with the balances established for supplies a nd finance. In F r a n c e, the Planning C o m m i s s i on h as a M a n p o w e r C o m m i t t ee w h i ch is responsible for translating the pl anned targets in terms of e m p l o y m e n t,

a nd in the light of the g r o w th forecast for e a ch sector of the e c o n o m y , a n d of the hypotheses for increased productivity in e a ch sector. T h e object of its w o r k is not me r e ly to define the conditions required for a balanced supply a nd d e m a n d in e m p l o y m e n t, but also to anticipate bottle-necks in the supply of skilled personnel a nd suggest remedies. H o w e v e r , these m e d i u m - t e rm forecasts h a ve not succeeded in establishing a link b e t w e en the e c o n o my a nd education except in the limited sphere of vocational training. T h is is because other types of training take m u c h longer (from five to fifteen years) to b e c o me adjusted to the d e v e l o p m e nt of e c o n o m ic n e e d s: the periods required for school construction a nd for teacher training, for e x a m p l e, h a ve to be a d d ed to the time taken for training pupils. T h is explains w h y m o st educational plans h a ve b e en d r a w n u p without regard to m a n p o w e r requirements. Latterly w o r k has b e en d o ne to formulatethough still empirically m e t h o ds of establishing long-term forecasts of these requirements. It is not possible, in fact, to use the s a me m e t h o ds here as for m e d i u m - t e rm forecasting: the techniques of linear p r o g r a m m i ng are not e m p l o y ed w h e n it c o m es to m a k i n g forecasts covering a long period, n or c an hypotheses for productivity be based on the extrapolation of past trends, o w i ng to such still unforseeable factors as the d e v e l o p m e nt of technological progress. F . H a r b ison has estimated Nigeria's requirements in intermediate a nd senior supervisory staff in 1 9 7 0 2 by formulating a series of 'reasonable hypotheses' concerning g r o w th in e a ch sector of e c o n o m ic activity a nd a s suming that productivity w o u ld r e m a in u n c h a n g e d. T h e Puerto R i co Planning B u r e au adopt ed the hypothesis of a 6 per cent annual increase in the national i n c o me from 1955 to 1 9 7 5, dividing this increase b e t w e en eight sectors of activity; a nd it also a s s u m ed that b y 1 9 75 the productivity level in Pue r to R i co w o u ld be the s a me as that of the Uni t ed States in 1 9 5 0. In Italy, S V I M E Z used a similar m e t h o d to estimate m a n p o w e r needs for 1 9 7 0, a s suming that by that date the productivity level w o u ld be c omp a r a b le to 1. International Standard Classification of Occupations, Geneva, 1958. 2. F. Harbison, in: Investment in Education, Lagos, Federal Ministry of Education, i960. 671 I N T E R N A T I O N A L S O C I A L S C I E N C E J O U R N A L that of France in 1950. In France, the M a n p o w e r C o m m i t t ee recently m a d e o ne estimate of requirements in 1970 by employing a m e t h od resembling that of S V I M E Z , a nd another based on the employers' forecasts ; but the t wo results are not compa r abl e. Although these first attempts to forecast training requirements are of dubious accuracy it is already possible to d r aw from t h em certain m e t h o d ological indications. At the outset a system of classification for the analysis of h u m a n capital m u st be selected, capable of application to both economic categories a nd those of the educational system.

T h e active population should therefore be divided into sectors of activity corresponding to those used in the national accounting system, thus m a k i ng it possible to use tentative long-term accounts. It should also be classified horizontally by educational levels, a nd not according to categories of occupation, w h i ch are too n u m e r o us a nd too ill defined from the standpoint of training. This m e a ns that for each country, equivalences wo u ld be established be twe en the levels of vocational training corresponding to the principal stages in the educational systemprimary, short secondary (including technical education), extended secondary, a nd higher. T h e workers c an thus be differentiated according to the n u m b e r of years of their training (actual n u m b e r , a nd that d e e m ed desirable at the end of the period forecast). Incidentally, there is nothing to prevent the use of separate sub-categories for types of training characteristic of certain special types of education, such as vocational training, technical education, higher scientific education, a nd so o n. Clearly m a n p o w e r statistics will have to be reinterpreted if such a classification is to be m a d e , for they have so far be en based on very different occupational criteria. For e x amp l e, the labour force for the French m a n u facturing industry wo u ld need to be classified in the following categories: 1. Unskilled or semi-skilled workers: seven-years' study with, in s o me cases, an additional few m o n t hs of vocational training at the compulsory schooling level. 2. Skilled workers, foremen a nd clerical staff: ten to eleven years' study (technical a nd commercial education, secondary vocational training, etc. ) . 3. Technicians a nd skilled clerical staff; twelve years' study (level of extended secondary education, either technical or general). 4 . Senior technicians a nd highly qualified clerical staff; fourteen years' study. 5. Engineers a nd senior administrative staff: sixteen years' study (level of the licence or of the engineering colleges). 6. Engineers a nd highly qualified administrative staff: eighteen years' study (level of the coles d'application, or the doctor's degree). It m a y be preferable, in countries in process of development, to m a k e a distinction be twe en illiterate workers (category 1), semi-skilled workers (complete or partial primary education), skilled workers (short secondary education or technical education), intermediate technical a nd administrative staff (extended secondary education) a nd senior technical a nd 672 E C O N O M I C S O F E D U C A T I O N administrative staff" (higher education). T h e se levels of study m a y be raised, incidentally, by an educational policy adapted to the needs of the e c o n o m y w h i ch m a y be reflected in increased productivity, with or without an increase in capital. ' H u m a n capital' can in these circumstances be represented in the form of a graph consisting of a pyr amid with five or six levels for each sector of activity, a nd expressed in mone t a ry terms according to the cost of training. T h e pyr amid resulting from totalling the data for each sector will give the general profile of the active population of the country concerned at the

beginning a nd e nd of the period covered by the forecast, thus m a k i ng it possible to determine the changes to be m a d e in the pyr amid of the educational system in order to modify the structure of the labour force. T h e application of these me t h o ds is fraught with m a n y difficulties, of course: long-term productivity hypotheses are still highly unreliable; tentative long-term accounts h a ve been attempted only in a few countries ; a census of the population rarely includes (as it does in the United States) an indication of the n u m b er of years of study completed ; the criteria to be used for specifying desirable changes in the distribution of training levels r ema in vague ; a nd little is k n o w n about the rates of internal promotion, often very high, wh i ch modify the equivalence of levels of vocational skill in relation to levels of training. Mor eove r, it is difficult to determine the o p t i m um level of expansion of a ny educational system in terms of a given target for economic growth, since both require the largest possible additional n u m b er of qualified workers. A m o d el with two sectorsteachers a nd non-teaching workerswould need to be prepared in order to m a k e clear w h at distribution of h u m a n resources be twe en education a nd the e c o n o my is desirable. T h e present uncertainties arise from the fact that only recently have training problems been studied from the e conomic angle, a nd that only n o w is the field of long-term forecasts beginning to be explored. It is already possible, howeve r, to form s o me idea of the progress that might be m a d e in the analysis of the h u m a n factors of growthprogress whi ch will enable us, in the near future, to determine m o re rationally the desirable relationships between the different levels of the educational system, the priorities to be established a nd the financial a nd h u m a n resources to be devoted to training people for economic development. This m e t h od of studying education as a m e a ns of preparing people to play an active part in the e c o n o my will also m a k e it possible to direct educational reforms in accordance with society's n e w needs. T h e yield from educational systems, considered as industrial undertakings, appears low partly owi ng to losses represented by failure at school a nd repetition of classes, a nd partly owing to the very slow progress in me thods of training. Efficiency has not be en such an overriding consideration in education as in the other sectors of the e c o n o m y, a nd the 'propensity to innovate' 1 is i. F. Perroux, L'conomie du XX e Sicle, Presses Universitaires de France, 1961. 673 I N T E R N A T I O N A L S O C I A L S C I E N C E J O U R N A L w e a k er thereas is only natural for an institution w h o se mission is to preserve a nd pass o n knowl edge a nd values. T h e re is no doubt that great progress can be m a d e , as concerns the efficiency of teaching, the organization of a sector w h i ch employs an ever larger proportion of the labour force, a n d the m o re judicious use of the time of the pupils, w h o , in the mo st a d v a n c ed

industrial countries, already m a k e u p one-quarter of the population. C o n c e rn wi th the deve lopment of h u m a n resources also leads to a broadening of the concept of education ; teaching adults to read a nd write in countries w h e re a considerable proportion of the population is still illiterate, a nd continuing education in all communi t i es n o w developing at an accelerated p a c e, mi g ht well be considered as conditions of e c o n o m ic g r owt h, a nd w o u ld then m o re easily secure recognition as priority objectives of deve lopment. T h e mobilization of the u n e m p l o y ed labour force (disguised u n e m p l o y m e n t ) , w h i ch has sometimes b e en referred to as a ' h u m a n investment', m a y also be considered in relation to training: it has b e en found that the profitability of w o rk thus carried out without capital (or with very small investment) d e p e n ds largely o n the n u m b e r a nd quality of the personnel supervising this unskilled labour force. W h e r e there is o ne foreman for every h u n d r ed workers, only l ow productivity tasks c an be undertaken (earthworks, road repairs) : but if this force includes a higher proportion of skilled a nd semi-skilled workers, m o re productive w o r k (building, etc.) c an be accomplished. It c an be argued, in this ext r eme case, that education multiplies e m p l o y m e n t, in the sense that the e x p e nditure on training skilled workers m a k es it possible to e m p l oy a m u c h larger n u m b e r of workers o n productive tasks. In the foregoing observations, the h u m a n factors of increased p r o d u ctivity have b e en considered only in terms of education. W e h a ve thus inevitably neglected the other elements, in particular pure a nd applied research, w h i ch has be en under study for s o me time from this s a me point of vi ew. It should merely be noted that research is o ne of the c o m p o n e n ts of the level of higher education, w h i ch produces both highly skilled senior staff a nd research workers. F r o m this point of vi ew, this category is seen to be a strategic sector of educational systems, since u p o n it depends the capacity to innovate a nd to apply innovations a nd the ability to extend the present frontiers of technological progress, social progress a nd e c o n o m ic g r owt h. A s w e h a ve seen, the w o rk d o ne in different disciplines appears to converge towards a c o m m o n conceptthat of the cultivation of h u m a n resources as a factor a nd objective of deve lopment. C a n w e conclude from this that the concept of h u m a n capital is n o w scientifically established? M a r y J. B o w m a n 1 a nd R . S. E c k a u s 2 have listed the analogies that m a y b e i. M a ry Jean B o w m a n , H u m a n Capital: Concepts and Measures, in: H u go H e g eland (ed.), Money, Growth and Methodology, L u nd (Sweden), C . W . K . Gleerup, 1961. 2. R . S. Eckaus, On the Comparison of Human Capital and Physical Capital, M I T , Center for International Studies, M a r ch 1961. 674 E C O N O M I C S O F E D U C A T I O N

d r a wn between h u m a n capital and physical capital. As w e have seen, the resemblances between them are n u m e r o u s; the stock of h u m a n capital c an be measured by years of study, the cost of its formation can be calculated and the increase in productivity resulting from its growth can be measured ; and its amortization can be represented by the a m o u nt of expenditure on training required to renew the stock of m a n p o w e r ; investment in training can also be likened to physical investment, and its v o l ume and nature determined in respect of an economic development target. But these resemblances have their limits; while h u m a n capital and physical capital are both required for the production of goods, and can therefore be regarded as factors of growth, m a n obviously presents irreducibly specific characteristics which it would be superfluous to seek to demonstrate. M e n are not bought and sold, m a n remains the master of his o w n labour p ower (at any rate at the individual or micro-economic level) ; and last but not least, m a n is not only a producer, but the endobject of the e c o n omy, the very goal of development. Seen against this wider background, the term ' h u m a n capital' resumes its place as a useful and to s o me extent operative concept which has m a d e it possible to focus economic analysis on a hitherto over-neglected factor of growth and to underline the need to maintain a balance, to determine an o p t i m u m, between physical investment and the training of h u m a n beings. W e can restore m a n to his due place in the economic system and in the c ommu n i ty without going to the length of reducing h im to a ma chine. 675 TEACHING METHODS AND THEIR COSTS Productivity of present educational systems CHARLES BENSON In this paper I shall m a k e s ome observations (a) on factor costs in the schools and (b) on efforts that have been m a d e to measure the returns from education. T h e discussion will be limited to public elementary and secondary education in the Un i t ed States of A m e r i c a. T o restrict the discussion, to o ne country is admittedly to give it a s o m e w h at 'provincial' scope. H o w e v e r, that w e deal only with public elementary a nd secondary schoolsis less restrictive, since these institutions represent the largest single e d u cational activity in the Un i t ed States a n d, I believe, display ma j or problemsof resource utilization in clear a nd extreme forms. COSTS OF INSTRUCTION In 1957-58 total expenditure on public elementary a nd secondary schools w as 813,500 million. Approximat e ly three-quarters of the s u m w a s applied to current operations; one-quarter w as spent on capital outlay (land, buildings, n e w e q u i pme n t) a nd interest. O f current expenditures ($10,300million), the largest share, i.e., $6,900 million or 67 per cent, fell u n d er the functional classification of instruction. T h e remaining categories in the current budgets covered such operations as transportation services, operation a nd maint enance of physical plant, health services, administration, etc. 1

In the c o m m e n t s to follow I shall deal mainly with instructional expenditures, for t wo reasons. First, a certain a m o u nt of the effort to provide shelter (and, say, health services) for pupils a nd teachers w o u ld in a ny case have to be m a d e in a m o d e r n society, regardless of whe ther the persons so sheltered (or nurtured) w e re engaged full-time in educational activities; second, in so far as instructional a nd non-instructional expenditures are related, it is m o r e likely to be changes in the former than in the latter, that lead to changes in the v o l u me of auxiliary services. 1. Biennial Survey of Education in the United States, 1956-58, Office of Education, U . S . Department of Health, Education, and Welfare, 1961. 676 E C O N O M I C S O F E D U C A T I O N Instructional expenditures in terms of real costs Let us consider growth in instructional expenditures during the period 1 9 2 9 / 30 to 1 9 5 7 / 5 8. In current dollars, the size of the instructional budget changed by a factor of 5 . 2, i.e., from $1,318 million to $6,901 million. This rise reflects inflation, g r owth in the student population to be served, changes in standards of attendance, a n d, presumably, s o me gain in quality of inputs. Certain peculiarities of the educational budget should be noted before attempting to me a sure m o v e m e n ts in real costs. Salaries of instructional personnel are the largest single item: $ 1 , 2 50 million in 1 9 2 9 / 30 a nd $6,368 million in 1 9 5 7 / 5 8. (Instructional salaries represented 95 per cent of total instructional expenditures in 1 9 2 9 / 30 a nd 92 per cent in 1957/58.) T h e ratio of students to instructional staff is relatively constant. For e x amp l e, in 1 9 2 9 / 3 0, there we re 24.9 students in average daily attendance per staff m e m b e r ; in 1 9 5 7 / 5 8, the corresponding figure w as 2 2 . 3 . 1 This is a difference of only 10 per cent over a twenty-eight-year period. It follows that the process of deflating the salary portion of the instructional budget by the average of teachers' salaries will s h ow an almost constant flow of real imputs per pupil into the educational activity. 2 I a m willing to a s s ume that a rise in teachers' pay has s o me effect in raising the calibre of staff. T h at is, I do not believe that rises in average teachers' salaries should be regarded wholly as inflation. Accordingly, I h a ve prepared estimates (Table 1) of the real inputs in the instructional p r o g r a m m e, per pupil per d a y, using the average of all earned i n c o me in the United States as the deflator for the salary portion of the budget. 3 ( T he remainder of the instructional budgetfor books a nd a wi de variety of suppliesis deflated by the U . S . D e p a r t m e nt of Labor 's wholesale price index; this w h o le non-salary c o m p o n e n t, as ment ioned, is a m i n or item as c o m p a r ed with the salary expenditures.) Instructional expenditures in constant (1949/50) dollars are s h o wn to h a ve risen from $.748 per pupil per d ay in 1 9 2 9 / 30 to $.915 in 1 9 5 7 / 5 8, an increase of 22.3 per cent in

real inputs. Advocates of higher spending o n education are not likely to be impressed by such a gain, h owe v e r, especially w h e n per capita gross national product in constant (1954) dollars w e nt u p by 60 per cent during the s a me years. T h e gain of 22.3 per cent w as not spread evenly over the grades. T h e period 1 9 2 9 / 30 to 1 9 5 7 / 58 w as m a r k ed by a m u c h m o re notable increase 1. These and all similar data in this article were taken from the following issues of the Biennial Survey: 1928-30; 1938-43; 1948-50; and 1956-58. 2. For example, see the estimates of We rner L . Hirsch in: 'Analysis of the Rising Costs of Public Education', Study Paper N o . 4 in: Material Prepared in Connection With the Study of Employment, Growth, and Price Levels, Joint Economic Committee, U . S . Congress, 1959. Use of own-salary deflators (or, what is the s ame thing, measurement of real inputs of public employees by man-hours) is the standard practice, of course, of the national income accountants of the U . S . Department of C o m m e r c e. 3. These salary data are prepared by the U . S . Department of C o m m e r ce and are listed as average annual earnings per full-time employee. 677 I N T E R N A T I O N A L S O C I A L S C I E N C E J O U R N A L T A B L E I. Instructional expenditures per pupil per day in current and constant dollars, selected years, 1929/30 to 1957/58 Current dollars 1 Constant dollars' 1949/50 = 100 Other _ , . Other _ Ye ar Salaries e x p e n d i t u r e Total Salaries expenditure Total 1929/30 O.34O O.O17 O.357 O.719 O.029 O.748 '939/4 0.360 0.013 -373 0.822 0.026 0.848 1949/50 0.730 0.054 0 - 7 84 0.730 0.054 0.784 x 957/58 1.210 0.101 1.311 0.829 0.086 0-915 x. Deflated by index of average annual earnings of full-time employees in the United States, 1949/50 = 100. 2. Deflated by U . S . De p a r tme nt of Labor, Wholesale Price Index, with base shifted to 1949/50. Sources .- Office of Education, U . S . Depar tment of Health, Education, and Welfare; Office of Business E c onomi c s, U . S . Depar tment of C o m m e r c e; Bur eau of L a b or Statistics, U . S . Depar tment of L a b o r. n secondary enrolments (up 79 per cent) than in pr ima ry (up 21 per cent). F r o m first-hand observation, I w o u ld judge that expenditures per pupil are usually s o me 30 per cent higher in the secondary grades than in the pr ima ry; it is, h o w e v e r, difficult to obtain national data on this point. In a ny case, it is highly likely that the increase in real inputs per pupil per d ay in the primary grades w as considerably less than 20 per cent over the last

twenty-five years. W h e r e as the average annual rate of increase in real inputs for elementary a nd secondary instruction together w as 0 .7 per cent, that for elementary alone m a y well be close to 0.5 per cent. Differences among the states T h e figures I h a ve noted above are national averages. Actually, there are fifty school systems in the country, o ne per state. Expenditures of the systems are highly correlated with individual state i n c o m e. T h e differences in instructional expenditures b e twe en the rich a nd poor areas are wo r th investigating. Table 2 lists instructional salary expenditures per pupil per d ay in current dollars for five high-income a nd five l ow- i n c ome states. T h e differences are quite large, larger probably than c an be accounted for b y differences in cost of living a nd by differences in the n o n - m o n e t a ry advantages of location of w o r k. H e n c e, while it appears true that the real inputs in education h a ve increased during the last twenty-five years, it also appears true that the geographical distribution of educational opportunities is quite u n e v e n. In this connexion it is wo r th noting the substantial differences in expenditure levels within the states as well, i.e., from o ne school district to another. T HE PRESSURE T O W A R D HIGHER STANDARDS OF INSTRUCTION Citizens in the Uni t ed States h a ve not be en extremely cost-conscious with respect to public services. A s R o b e rt G . W o o d has stated, '. . . few inhabitants . . . of the nation . . . h a ve ever looked to their local gove rnments to 678 E C O N O M I C S O F E D U C A T I O N T A B L E 2. Instructional salary expenditures per pupil per day in five rich and five poor states, current dollars, 1929/30 and 1957/58 1 Salary expenditures States Absolute change 1929/30 1957/58 Rich California Connecticut Delaware Nevada N e w York National Average Poor A l a b a ma Arkansas Kentucky Mississippi South Carolina 0.52 0-39 0-35 45

0.53 0-34 0.19 0.18 0.21 0.18 0.21 1.44 1.42 1.56 1.41 1.74 LSI O.78 O. 74 O.78 O.63 O.76 + O.92 + 1-03 + 1.21 + O.96 + i.ai + 0.87 + 0.59 + 0.56 + 0.57 + 045 + 0-55 1. Rich states are the five highest in terms of per capita personal income in i960 : the poor states were similarly chosen. Ranking by income is quite stable from one point of time to the next. Sources: Office of Education, U . S . Department of Health, Education, and Welfare; Office of Business Economics, U . S . Department of Comme r c e. "optimize" or " m a x i m i z e" anything'. 1 Taxation is oppos ed, but the resistance to it is focused on the degree of alleged public 'inefficiency', not o n the question of whether too large a v o l u me of services has b e en provided. O n the other h a n d, voters occasionally d e m a n d a larger v o l u me of services but, w h e n they do so, they usually fail to m a k e a connexion b e twe en such expansions a nd the increase in taxation required to finance t h e m. In the post-war period w e h a ve seen a considerable increase in d e m a n d for higher standards of instruction in the local school systems. In part, this is a matter of public acceptance of the ever-higher goals set by professional

educators. 2 S u ch acceptance c o m es naturally to the people of the United States of A m e r i c a, because of their long-standing belief that educational opportunities should be widely distributed. O f late, the vi ew that education of high quality has an important bearing o n e c o n omic strength a n d, even, on national defence has also gained acceptance. Lastly, m a n y parents feel acutely the ne ed for i m p r o v e m e nt in schools so that their children m a y have a better chance to gain admittance to a satisfactory college, such a dmi ttance having b e c o me quite highly competitive. 1. Robert C . W o o d , 1400 Governments, Cambridge (Mass.), Harvard University Press, ig6i, p . 198. 2. See Charles R . Adrian, 'Metropology: Folklore and Field Research', Public Administration Review, S u m m er 1961, p. 156. H e states, '. . . "shortages" exist in considerable part because they are defined in terms of the standards of the professionnals in the various fields and w e m a y be sure these are standards we can never catch, for they disappear before us like gossamer on an autumn wind'. 679 I N T E R N A T I O N A L S O C I A L S C I E N C E J O U R N A L T h e response of local school b o a r ds to the d e m a n d s for higher standards of pe r formance has not b e en to hire m o r e teachers per ioo pupils, i.e., there has b e en n o serious effort to r educe class size, except in cases w h e re the n u m b e r of pupils per teacher far exceeded n o r m al standards. S u c h a n attack w o u ld h a ve required increases in tax levies far a b o ve the traditional pattern of a d v a n c e . 1 G i v en that citizens of the local district benefit in u n e v en m e a s u re from increases in educational output (e.g., parents receive m o r e benefit than non-parents) a n d given that, o w i ng to the differences in level of support in a decentralized structure, the aggrieved taxpayer c an a lways point to s o me other district w h i ch 'has just as g o od schools at half the cost of ours', the traditional pattern of a d v a n ce in tax rate c a n n ot be seriously violated. In m o st districts the response has b e en rather to offer s o m e w h at higher teachers' salaries (nationally, the a d v a n ce in the average of teachers' p ay has b e en about 5 per cent a year for the last ten years) a nd to provide i mprovements in physical plant, albeit in the conventional structural forms. M o r e recently, the progressive districts appe ar to h a ve c o m e to the following conclusions: (a) that the actions of the individual district wi th regard to salary h a ve very little effect o n the calibre of the candidates attracted to teaching, i.e., o n the character of basic recruitment to the field; (b) that there are limits to the quality of staff that c an be d r a w n f rom the existing supply of teachers' services; a nd (c) that the conventional type of schoolhouse m a y not serve to a c c o m m o d a t e n e w instructional practices that m a y he brought about b y advanc es in educational technology. T h u s , w e find s o me school districts responding to the rising d e m a n d s in w a ys that are

highly eclectic a nd frankly experimental. 2 T h e changes in instruction m i g ht be divided into three categories: (a) revision of curriculum, (b) r e d e p l o yme nt of staff, a nd (c) use of n e w physical devices. A s a n e x a m p le of revision of the curriculum, I w o u l d cite the proposal of the Illinois C o m m i t t ee o n School M a t h e m a t i cs (the c ommi t t ee d r a ws its m e m b e r s h ip f rom the Colleges of Educ a t ion, Engine e r1. For example, suppose a district has 1,000 pupils, a pupil-teacher ratio of 25 to 1, and an average teachers' salary of $5,000. It will have an annual teachers' salary budget of $200,000. N o w , suppose the district considers whether to cut the pupil-teacher ratio to 20 to 1. This must be done district-wide, not schoolby-school, since gross inequalities within the schools of a district are intolerable. T h e salary budget would increase in the order of $50,000. In addition, it would be necessary to build n ew classrooms, unless the district h ad idle ones, which is unlikely. A n y lesser adjustment of the pupil-teacher ratio, however, would probably be regarded as trivial by both staff and voters. It remains true, nonetheless, that teachers appear to wo rk less than a full year in their regular jobs. T h e U . S . Department of Labor (Monthly Labor Review, October 1961) has reported that 23.5 per cent of married males in teaching hold two jobs during the regular school year. A large proportion probably take on extra s u m m er work. T h e marginal cost to a school district of garnering s ome of their extra hours for teaching might well be less than average cost. 2. T h e changes are eclectic in the sense (a) that they are not promoted by class interests, as w as the earlier vocational school m o v e m e n t, a nd (b) that they are not embraced within any single 'philosophy of education', as were the changes that occurred under the 'progressive m o v e m e n t '. 680 E C O N O M I C S OF E D U C A T I O N ing, a nd Liberal Arts a nd Sciences of the University), with respect to content a nd m e t h o ds of instruction in the subject. 1 No rma l l y, if a school board institutes such revisions of the curriculum, a certain a m o u nt of retraining of staff in the subject is required. By redeployment of staff I refer to such practices as the use of the u n g r a d ed primary, contract correcting, 2 t e am teaching, a nd a hierarchy of subject specialists. A s for the third category, s o me of the n ewer physical devices being put in use are television, language laboratories, a nd teaching ma c h i n e s. It is too early to me a sure the returns of these various changes. O n e difficulty, of course, apart from the general o ne of estimating the yield of a service activity, is that a ny c h a n ge in process in the schools is likely to yield a s o m e w h at different set of products. T h at is, not only m u st o ne try to me a s u re whether certain skills, capacities,.etc., we re developed, to a greater or a less degree, but also o ne needs to identify, if possible, w h i ch of the old products disappeared altogether a nd w h at n e w ones w e re provided. Indeed, it is difficult to take w h at should be the easier first step, n a m e l y,

to find w h at the n e w processes cost as c o m p a r ed with the old ones, for school accounting, traditionally, is rather primitive. This failure of the schools to apply the techniques of cost accounting m a y well result from the voters' lack of interest in public costs (provided the annual increase in the tax bill is no greater than usual). Further, the innovations are often priced to school districts at less than full cost. F or e x amp l e, a person already teaching in a school district m a y volunteer as a t e am leader; he or she receives, let us say, a mode st extra stipend. H a d the district prepared a job description of the t e am leader's role a nd sought a suitable candidate in the ma r k e t, the cost mi g ht well h a ve be en several times greater. In the s a me context, the cost of innovation is often underwritten in part by foundations a nd universities in connexion with their o w n educational research p r o g r a m m e s. Lacking a ny me a sure of the costs a nd benefits of alternative schemes of instruction, h o w does a district decide to e m b a rk o n a venture? First, s o m e o ne in the district m u st b e c o me a w a re of the n e w practices, thus it m u st be o ne w h i ch is included in s o me kind of c ommu n i c a t i on circuit. Secondly, with or without benefit of subsidy, the innovation m u st be absorbed within the traditional pattern of cost advanc e. Thirdly, the present staff of the district m u st see that the innovation offers s o me possibility for i m p r o v e m e nt in instruction. Fourthly, the present staff m u st feel that chances of serious h a rm being d o ne to students are slight. i. See M a x Bebe rman, An Emerging Program of Secondary School Mathematics, C a m bridge (Mass.), Harvard University Press, 1958. A similar group is the Physical Sciences Study Committee, established initially at the Massachusetts Institute of Technology. a. 'Contract correcting' is simply the practice of engaging the services of housewives in a community on a part-time basis to correct the English themes of students. This practice allows the regular teacher to assign a larger volume of writing to his students. 681 I N T E R N A T I O N A L S O C I A L S C I E N C E J O U R N A L M E A S U R E M E NT OF RETURNS E v e n as rich a country as the Un i t ed States faces difficult questions wi th regard to the allocation of resources to education. H o w m u c h of the various kinds of educational services should be provided? T h e professional educator often appears to urge ' m o r e a nd ever m o r e '. T h e economist r ema i ns sceptical a nd argues that there are strong c omp e t i ng d e m a n d s within the public sector itself a nd even within the sub-category of public investment in h u m a n capital. It is not clear, for e x a m p l e, w h e t h er the long-run return f r om education is greater than that f rom health services or internal migration wi th respect to such investments in certain groups of the population, at least. Fortunately, w o r k is well u n d er w a y o n assessment of returns o n investment in h u m a n capital. I shall call this the m a c r o - e c o n o m i c, or aggregative, a p p r o a ch to m e a s u r e m e n t. T h e re is, h o w e v e r, another m a j or task of m e a s u r e m e nt w h i ch I shall call mi c r o - e c o n omi c. A s w e h a ve not ed, school boa rds are n o w in the position

w h e re they m u s t m a k e explicit choices of the m e a n s to accomplish the given ends of the schools. It is n o longer u p to the teacher, enclosed in the four walls of the classroom, to decide h o w best to carry on his or her function. Choices affecting the use of specialized h u m a n resources a nd the use of c o m p l ex types of e q u i p m e n t a nd w h i ch affect the w h o le school district in quite fundament al wa y sa re at present be ing m a d e b y the boa rds in the absence of clear me a sur es of the costs a nd benefits of alternative s c h e m e s. T h e re is a n obvious n e ed for da ta to help guide the allocation of resources within the educational activities. The macro-economic view T h e general a p p r o a ch is to relate 'additional lifetime earnings to the costs of the education associated wi th such earning. . . '1 T h at is, if high school graduates h a ve lifetime earnings w h i ch exceed those of persons wi th, say, only eight years of elementary schooling b y x dollars, a n d if the total cost of the (four) years of high school isj dollars, the ratio o f* toy is an estimate of the return from secondary school education. In the Un i t ed States this w o r k c an be said to h a ve b e en b e g un by H a r o ld F . Clark, of C o l u m b ia University, in his Life Earnings in Selected Occupations (1937). In the last few years it has received great impetus f rom T h e o d o re W . ~ S c h u l t z, of the University of C h i c a g o, a nd f rom G a ry S . Be cker a nd J a c ob M i n c e r, of C o l u m b ia University. 1. T . W . Schultz, 'Education and Economic G r o w t h ', in: Nelson B . He n ry (ed.), Social Forces influencing American Education, Chicago, University of Chicago Press, 1961 (Sixtieth Yearbook of the National Society for the Study of Education, Part II). See also, G a ry S. Becker, 'Underinvestment in College Education?', American Economic Review, Vo l. L , N o . 2, M a y i960, a nd Jacob Mincer, 'Investment in H u m a n Capital and Personal Income Distribution', Journal of Political Economy, August 1958. 682 E C O N O M I C S O F E D U C A T I O N T o date, the estimates of returns from education c o m p a re favourably with those from business capital. For e x a m p l e, Schultz estimates, using j 9 58 data, the return on high school education to be 11.8 per cent. T h e average of return on all business capital could be set at about 8 per cent. For college education, Schultz judged the return to be 10.96 per cent. 1 The se estimates, though tentative, suggest that it might be g o od policy to see that the terms on w h i ch education is financed be m a d e m o re nearly equal to those under wh i ch business capital is itself financed. Admittedly, the estimates of returns are focused solely on the 'producer's g o o d' c o m p o n e nt of education. T h at is, schooling is viewed as an intermediate product in the s a me w a y as a m a c h i ne tool. T h e re is also a consumption c o m p o n e n t; hence, the yields are necessarily underestimates of the returns from education. This is not a fair ground for criticism. B ut s o me other questions m a y be raised. It is perhaps doubtful whether there

is a clear e n o u gh separation of the education variable as distinct from such variables as motivation, nepotism, a nd barriers to entry into certain of the higher-paying occupations. Let us say, howeve r, that the yields mi g ht be taken as a useful guide to aggregate investment in education (provided, of course, that the marginal returns do not drop too far be low the averages). I still question whether the market c an be relied u p on as a suitable guide to w h at particular kind of education should be e x p a n d ed or contracted. For instance, will the market tell us h o w medical training should be distributed as be twe en practitioners a nd scientists? I fear not, just as I do not think that the ma rket provides reliable information on the proper allocation of resources toward investment in the physical instruments of research. H o w e v e r , mistakes in the allocation of types of physical capital c an be m o re quickly a nd easily rectified than can mistakes in the allocation of training opportunities. The micro-economic view It is easy to state the conditions of efficient allocation of resources in e d ucation in general terms. 2 It is not easy to apply the tools of economic analysis to real situations in the schools. T h e outputs are multidimensional a nd hard to me a sur e. T h e .inputs cannot be a s sumed' to be h o m o geneous, but it is difficult to evaluate the qualitative differences in these inputs. A certain a m o u nt of pioneering w o rk in relating costs a nd benefits has been d o ne by Paul M o r t a nd his associates at C o l umb ia University. 1 Mort has sought to determine h o w 'adaptable' school districts are in changing their processes of instruction a nd in offering a larger n u m b er of products. In general, he has found that adaptability a nd expenditure per pupil are 1. Schultz, op. cit. 2. G . S. Benson, The Economics of Public Education, Boston (Mass.), HoughtonMifflin C o m p a n y, 1961, Appendix G . 683 I N T E R N A T I O N A L S O C I A L S C I E N C E J O U R N A L positively related. H o w e v e r, M o r t 's a p p r o a ch includes no serious attempt to evaluate the yield to the student (or to the society) of o ne set of products as c o m p a r ed with another, nor does it offer a m e a ns of evaluating the relative efficiency of o ne set of processes as c o m p a r ed with another. M o r e recently, Jos eph A . K e r s h aw a nd R o l a nd N . M c K e a n h a ve suggested that systems analysis can be applied to the schools. In a simplified e x a m p l e, they propose that o ne mi g ht evaluate innovations of a s s u m ed equal dollar cost (change in pupil-teacher ratio, c h a n ge in salaries of teachers, expansion of the guidance p r o g r a m m e , etc.) in terms of the apparent effect on pupil achievement scores. 2 F or the given increment in the size of the school budget the type of c h a n ge that h ad the greatest

positive effect on scores of the students generally (or on that class of students in w h i ch the school authorities h ad the greatest i

Affordability & Access;Costing, Pricing&Accountability Course Goals: 1) Apply theoretical and empirical methods of economics to education policy. Economics is a powerful framework for thinking about education policy. Economics focuses us on the incentives created by a policy, allowing us to predict its intended and unintended consequences. 2) Distinguish good empirical research from bad. There are thousands of empirical education studies. This course helps you separate the wheat from the chaff by teaching you to read closely and critically. We will discuss the key challenges in education research and learn the best-practice methods that overcome them.

Economics of Education
Education economists analyze both what determines or creates education and what impact education has on individuals and the societies and economies in which they live. Historically at the World Bank a great deal of emphasis has been placed on determining outcomes to educational investment and the creation of human capital. The primary mission of the economics of education group is to identify opportunities for improved efficiency, equity, and quality of education and promote effective education reform processes; to help improve, among both World Bank staff and clients, knowledge of what drives education outcomes and results; to better understanding how to strengthen the links of education systems with the labor market; and to build and support a network of education economists and build bridges to all those who are interested in their work. The World Bank addresses key themes of Economics of Education through work in these five topic areas:

Why is economic analysis important? All countries face the problem of allocating limited resources such as capital, skilled and unskilled labor, land and other natural resources to a variety of different uses such as production of consumer goods, investment in industry, infrastructure, education, health etc. with an aim to reach a more fundemantal goal of reducing poverty, accelerating economic growth and/or reducing income inequalities. Given limited resources, choices have to be made between alternative uses of these resources such that the benefit to the economy and society is as large as possible. Therefore, all World Bank projects must meet certain selection criteria, including economic justification. in this regard the education sector specifically seeks to answer the following questions in a country's public expenditure analysis:

How much is spent on education and what is the share of the government's expenditure? How do governments finance the education sector and what do they finance? Is there equitable distribution of the public resources? Is the public getting its money's worth?

Is the spending adequate and sustainable?

Public-Private Partnerships in Education

What is school-based economic management?\ Goals Investment in Education COST ANALYSIS IN EDUCATION DemandandSupplyofSchooling-AccesstoEducation Economic and Social Externalities in Education The course aims to provide students with a perspective on the pressing issues concerning the education sector in countries -theoretical, methodological and applied- and to bring them towards the research frontiers in certain selected topics. It will also try to shed light on the evolution of economics of education as a sub eld and its contribution particularly in developing economies. Education is central to development and the course will focus on emphasizing this relationship. Broadly, it will assert the importance of education as an investment, adding to the stock of human capital which becomes an essential primoever for economic growth. It will then introduce students to the various methods used in analyzing both the inputs and outputs that go into the provision of good quality education services

Economics of Education MA
The MA will develop your understanding of the economics of education, which assesses the benefits and costs of education to individuals and society, and uses this framework to evaluate policies in developing and developed countries. It will:

equip you with knowledge and understanding of the economics of education and its research methods enable you to apply the economic concepts and methods for analysing and researching education policy to your own field of work and a national context provide you with the capacity to evaluate and critique the ways in which economists have applied economic principles to education and education policy-making.

Features
The programme:

examines relationships between education and its benefits to individuals and society explains and applies key economic concepts, methods and ideas to education in both developing and developed countries provides the tools for the critical assessment of education policies.

Who is this programme for?

People who would like to work in educational planning or policy, to engage in research in educational finance or economics, or to extend their understanding of the economic and financial aspects of education.

Attendance
Face-to-face evening sessions.

How are you assessed?

By coursework assignments and a 20,000-word dissertation or 10,000-word report.

Entry requirements
A good first degree or equivalent comprising at least 50 per cent economics. If your qualification is in another subject, you should also have appropriate experience of working in education.

Economic Perspectives Of Education Policy

Course Code:MMAECO_02

This module is designed to give an overview of the entire field of economics of education to students who are new to economics, as well as to those who have previously studied economics. The main criteria for evaluating policy efficiency and equity are applied throughout the module. Using these two criteria, markets, governments and networks are analysed as coordination mechanisms for deciding the allocation of resources to education. The module develops arguments and evidence about the appropriate roles of the state and the private sector in the provision of education by combining economic theories with research evidence on education systems across the world. Topics studied will vary each year and have included these policy questions:

Educational Planning, Economics and International Development MA

We will provide you with the relevant knowledge, understanding and skills to work in educational planning, policy, management and administration in the context of international development. We will enable you to:

Do smaller class sizes improve students' learning, or could the schools budget be better spent elsewhere? Will higher pay improve the recruitment and retention of teachers? Does education cause economic growth? Are higher education tuition fees equitable? Would performance related pay improve teaching quality? Will parental choice and school competition improve educational standards or simply increase inequalities? Should we pay poor families in developing countries to send their children to school?

appreciate the interconnections between society and the economy and the implications for educational planning understand and apply economic principles to educational planning issues critique the ways in which economic analysis and evidence are applied to education policy issues understand the strengths and weaknesses of various approaches to educational planning analyse an educational planning problem in the context of the national economy and its development.

Features
The programme will provide an opportunity for you to:

interact with international guest speakers and internationally renowned researchers share your thoughts and experiences with a diverse group of colleagues make a study visit to Paris (not included in the course fee) incorporating UNESCO and the International Institute for Educational Planning (IIEP).

Policy Studies in Education MA

We will:

introduce you to ways of critically analysing education policy within a broad social, economic and political context familiarise you with contemporary developments in education policy in institutional, local, national and global contexts explore existing policy issues and practices enable you to apply the insights provided to your own experiences.

Features
The programme offers:

a broad range of perspectives which allow you to develop a comprehensive theoretical understanding of education policy the chance to study major and topical themes within social science (such as globalisation, markets and managerialism) with a particular emphasis on their effects on and relevance to education the opportunity to work with tutors who are nationally and internationally recognised individuals access to recent and current research.

The Economic Value of Education

Course Code:MMAECO_05

The module aims to enable you to understand how education policies can be rigorously evaluated using empirical evidence. It develops and applies the tools of economic analysis, supported by empirical evidence, to the assessment of educational policies. The estimation of private and social rates of return to education is studied in greater depth than in Economic Perspectives on Education Policy, including the problem of identifying education as a causal factor in earnings differentials. Cost benefit and cost-effectiveness analysis are applied to some specific policy examples. Policy issues examined include how to Increase access to education and enhance the quality of schools and whether performance related pay improves teacher performance. Applications are chosen to reflect policy Issues in both developing and developed countries as well as general, vocational and higher education. Some seminars include hands-on experience in using spreadsheets for analysing economic and educational data.

Policy Studies in Education MA

We will:

introduce you to ways of critically analysing education policy within a broad social, economic and political context familiarise you with contemporary developments in education policy in institutional, local, national and global contexts explore existing policy issues and practices enable you to apply the insights provided to your own experiences.

Features
The programme offers:

a broad range of perspectives which allow you to develop a comprehensive theoretical understanding of education policy the chance to study major and topical themes within social science (such as globalisation, markets and managerialism) with a particular emphasis on their effects on and relevance to education

the opportunity to work with tutors who are nationally and internationally recognised individuals access to recent and current research.

Who is this programme for?

People involved in or interested in contemporary education policy, administration and governance.

Economics of Education MSc

During the course, you will:

develop your ability to apply economics and quantitative research methods to the analysis of educational issues assess the benefits and costs of education to individuals and to society, and use this framework to evaluate education policies in both developing and developed economies gain a critical understanding of the use of quantitative research methods and mixed methods in evaluating specific education policies acquire the skills you need to appraise critically research in the economics of education and to work on research projects using quantitative and mixed methods prepare for starting PhD-level research in the economics of education.

Features
The MSc is unique in the UK and is recognised by the Economic and Social Research Council (ESRC) for '1+3' funding.

Who is this programme for?

Anyone planning to undertake doctoral-level research in economics of education, or anyone who is interested in a career in educational research using quantitative and qualitative techniques. If you have studied economics in your first degree and need a better understanding of policy evaluation and the economics of education for your work in educational policy, this will give you the skills you need.

The CEE is a multidisciplinary centre with three partners: The Centre for Economic Performance at LSE; the Institute for Fiscal Studies; and Education. All three partners have a proven and distinguished track record in the economics of education and training, and a great breadth relevant issues relating to education, training and the labour market.

The CEE seeks to undertake systematic and innovative research in the field of the economics of education by applying the latest technique We look at a broad range of questions including why people invest in education and training, the way education systems are organized, and and skill acquisition on economic and social outcomes. Our programme covers education at all stages of peoples lives.

The research is heavily orientated towards doing empirical work which can inform policy by taking questions to the data. We apply (and dev methodological approaches for this purpose, from evaluation methods, statistical analysis of large-scale micro-data sources, through to mo work.

In the time since our establishment we have made a substantial contribution across many areas. These include: returns to education; adult between family background and educational attainment; the effects of school choice and competition; the value of primary education as re

market; the labour market for teachers; intergenerational mobility and educational inequality; evaluation of many government policies. Rese downloaded from the CEE Discussion Paper series. A good overview of much of our work will be found in our book Whats the Good of Ed from Amazon.co.uk)

Work is now being undertaken on areas including the Every Child Matters, andChildrens Workforce strategies the Government programm framework to support the "joining up" of childrens services; on the impact of education on Social Mobility; Further and Higher Education; th Programme. In addition to projects undertaken as part of the CEE agenda, researchers in CEP are involved in many other research project and skills.

The CEE is headed by Professor Stephen Machin, an expert in the economics of education and currently Professor of Economics at Univer is the Research Director at the Centre for Economic Performance and is affiliated with the Institute for Fiscal Studies. Sandra McNally is the Education and Skills programme at CEP.

Details of CEE projects from 2005 onwards and of our plans for 2008 onwards are available on the CEE research projects pages. In additio some examples of other projects which are currently being investigated by CEP researchers.