The Diseconomics of Growth

H.V. Hodson


Chapter IV.

How Growth Is Measured

The statistical measurement of economic growth derives, as we have seen, from that of national income (in the sense of the money income of the whole nation, not the earlier sense of the national public revenue). This is ordinarily expressed, in this context, as Gross National Product, or GNP, that is, the nation’s aggregate money-measured output of goods and services, including government services and not only consumption goods but also capital goods and increases of stocks (inventories in American usage) and that part of exports which is not counterbalanced by imports. Thus economic growth is, by common custom, measured by growth of GNP. Most economists would probably agree that a better basis, if it could be determined and relied on, would be Net National Product, or NNP, IE GNP less deductions for waste, deterioration and replacement of capital and physical resources: but the right scale of such deductions is held to be so arguable, and the information for assessing them so patchy, that in practice the gross figures are almost invariably used by economists and statisticians as the basis for percentage growth, and it is with these that official publications and journalistic or political comment have made us all familiar.

We must accordingly always bear in mind that such figures have the inherent defect, even within the limitations implied in a compilation of money income, that they are not diminished by any allowance for deterioration, decay or exhaustion of resources whether natural or man-created. What this means may be seen from a homely analogy. Three men enter into possession of similar houses, in the same state of structural and decorative repair. One uses his spare time to keep up the condition of his house, painting and making good and clearing drains and gutters. The second man does nothing, and lets his house run down. The third man works overtime or takes a second job, and spends his extra earnings on getting a builder to do the house repairs and decorations. The gross money income of the first two men is the same; that of the third is larger by the amount of his additional earnings. But if allowance is made for the condition of the houses, which affects both the amenity of those who live in them and their market value when they come to be sold, the net incomes of the first and third men are the same, while that of the second man is lower by the extent of the deterioration. There is no doubt which of these two comparisons gives the truer picture of the economic position of the three householders.

When the measurements are applied to overall national income, such personal variations are absorbed and averaged out in the total. But the truth remains, that gross national product, while including repairs or replacements that are actually paid for, ignores and obscures the extent of deterioration of resources; and that changes in GNP over a period of time, or comparisons of GNP between one country or another, take no account of simultaneous changes or differences in the extent to which resources are wasted, used up or allowed to decay. For example, in times of major war all postponable repairs are abandoned; gross national product, GNP, keeps up, because everyone is occupied, though on inherently less valuable objects; but the net national product, NNP, goes down and the deterioration must eventually be made good; the making-good again enters into GNP, but at the expense of the consumable income of the nation.

There is another form of decline of capital which even the figures of Net National Product do not usually or wholly allow for. That is the running-down of natural resources. These may be land, or minerals, or water, or natural forests, or valuable wild-life (game, fish, vegetation). A country could have a high and rising GNP yet end up poor because its natural resources were becoming exhausted.

A farmer reckoning the gross and net return from his acres must do his accounts on the basis that his land, including the hedges and ditches and watercourses, is in as good heart at the end of the year as at the beginning. To keep it sois a first principle of good husbandry, and a condition of most farming leases. Output reckoned or income gained at the expense of a deterioration of the land is an illusion. Every farmer or landowner knows that. Yet this is what we do when we tot up GNP and derive from it figures of economic growth. We may be impoverishing the land and forests, exhausting the mines and quarries and strata of oil or natural gas, lowering the water-table, killing the fish in rivers and coastal waters, desecrating such wilderness as remains, yet still showing apparently healthy economic growth. For the only yardstick by which we measure it is the value of the products we get out, plus paid-for services, or the incomes of those who produce or render them.

If a trustee puts money into a mining company or a leasehold investment he must set aside enough from the yield to balance the depletion of the mine or the running-down of the lease before he can count the return on his fund or its capital growth. As nations we are trustees for the natural resources we inherit; as generations we are but leaseholders of all our assets. Our economic growth figures are like a false prospectus; they conceal unpleasant facts which might disclose what improvident managers, bad husbandmen and negligent trustees we have been.

To return to the more banal critique of growth statistics, for some purposes a better measure of economic growth than that of GNP itself is growth of GNP per head, or average individual share in national income. Clearly there would be little sense in talking of a country’s satisfactory economic growth if its economy were growing more slowly than its population, so that on average its people were becoming less well off—something which can certainly happen in less developed countries and could happen in the more advanced. However, crude division of GNP by population numbers shows up a complexity in this matter, such as we shall encounter in other respects as we consider the basis of growth statistics further.

How population grows is highly relevant to the question how far or fast the people’s average economic take-out is growing. To the extent that population is growing by excess of births over deaths, we can think of the extra people as being babies. These contribute nothing to national economic output; on the other hand, they need, to begin with, less of it than adults. A married man’s income does not have to double to maintain the same standard of life when he and his wife have two small children. Not only does their food and clothing probably cost less than their parents’, but, more important, they use to a large extent the same basic life-equipment—the family house and its furnishings, the family car and so on. Until they get to school age, furthermore, they need little or no addition to public capital or expenditure—EG schools and their equipment, roads and public transport; then they need a lot. When they go to work they need further additions of productive capital—factories, offices, commercial transport, and the rest. Contrast this with increase of population by net immigration of adult workers, who may be assumed forthwith to contribute the average share to output, to consume the average share, but to need more than an average share of provision of public and private capital (for their homes, their public facilities and utilities, the places and means of their work). Evidently, increase of GNP divided by population, as a formula, is only a first approximation to growth of average economic income as people and their governments know it.

Nevertheless, we may use the formula as the statisticians do, provided we realise its shortcomings. Much the greater error arises when it is treated as a measure of average well-being. To understand this, we must go back to the structure of statistics of GNP.

Before, however, we examine the make-up of national income figures themselves we need to beware of a problem affecting any comparison of them from year to year, or over longer periods, and therefore vital to the computation of economic growth. This is the problem of changing price-levels. If all prices moved upwards or downwards at the same rate, there would be no difficulty in correcting national-income figures (expressed in current money) for price changes, beyond a little simple arithmetic. But obviously this is not what happens in real life. In a given year, prices of some things may go up by 25 per cent, some by 5 per cent, some not at all; others may fall. Economists wanting to convert money-expressed income figures into real terms therefore have to apply averages, which can be arguably weighted in different ways; the resultant exercise has been candidly described as a makeshift. [See note 1.] An error of, say, and a half per cent in determining the average price-change factor to be applied would make a and a half per cent difference in the apparent growth rate—a very significant amount. The matter is further complicated by the fact that the goods entering into the price-table themselves change: new products are made and sold, some existing products disappear, others change their character. How to allow for such changes in applying average price movements is clearly a very complex business. Over a longish period, the cumulative effect of variables of this nature may seriously affect the accuracy of national income comparisons, and still more the value of the economic growth figures derived from them, as indices of real material welfare.

The next, essential thing to underline about national income figures, which enter into GNP, is that they are confined—to borrow a phrase quoted earlier—to “those goods and services which are customarily exchanged for money”. They exclude all benefits which are not paid for in money, such as light and air and the beauties of nature. These can deteriorate miserably without any effect on GNP or any discount on growth. “It is a paradox,” wrote Professor Pigou in his Economics of Welfare (1929), “that the frequent desecration of natural beauty through the hunt for coal or gold, or through the more blatant forms of commercial advertisement, must, on our definition, leave the national dividend intact.”

National income figures exclude also all unpaid services. The anomaly so often quoted is that if a man marries his housekeeper and ceases to pay her a salary the national income goes down; if he hires a cook to let his wife go out to paid work the national income goes up, even though the wife displaces somebody who then takes the job which the cook used to do, so that the net total of work is the same. These paradoxical cases are not of any great significance in themselves. The propensity of men to marry their housekeepers is probably fairly constant. It is more important if there is a trend, over a period of time, for more or fewer married women to take on salaried work and pay others to do their housewifely or motherly duties—at home, in restaurants, in schools or play centres. Such a trend could easily distort changes in national income figures as measures of change in real economic output. And if the trends differ between countries, then comparisons of their so-called economic growth-rates are further vitiated.

Men are concerned here as well as women. Do-it-yourself jobs have no place in national income statistics. There has manifestly been an enormous expansion of do-it-yourself in American and British homes, and presumably in other Western countries, because the main reasons are common to them all. The first is that, as real incomes rise and working hours are reduced, one of the things that people most want and are able to do is to improve their home surroundings and comforts. The second is that all services, not least those of plumbers, electricians, carpenters and decorators, grow more and more costly, even after allowance is made for general inflation of prices and wages. It becomes worth while even for quite well-paid people to save money that would be spent on such services by doing the jobs themselves. This behaviour brings a hidden increment of real income-benefit to the family which they would otherwise have paid for—without any addition to calculated national income, because no money passes.

Besides all such unpaid work in and around the home, there is another body of effort, directly adding to general welfare, which goes unrecorded in the statistics. This is voluntary service and public work of all kinds. A favourite illustration for early British income-statisticians, comparable to the man marrying his housekeeper, was the fact that when Members of Parliament were first paid salaries, modest as they were, the national income went up by a quarter of a million pounds a year. Pursuing that example might lead us into amusing but barren argument as to whether politicians contributed to national output or well-being. The fact remains that government is an essential service, and that in Britain by far the greater part of elective government and administration of petty justice is unpaid service.

Politics and local government and the magistrates’ bench, however, are only the fringe of voluntary unpaid work in the public cause. In all countries, rich and poor, volunteers work in welfare and health services, hospitals, schools, churches, organisations for care of the sick, the disabled and the needy, youth centres, charities of all sorts, and countless other ways. All this amounts to a very big slice of real national income. The services are no less substantial to the beneficiaries than they would be if rendered by paid officials and professional or other employees, in which case they would enter into national-income statistics. The size and nature of that slice varies from country to country and from time to time. In most Western countries, certainly in Britain and the United States, it seems to be steadily growing in volume and in many respects becoming better organised and more effective, and therefore adding more to real welfare. In America there are many hospitals which simply could not run without their volunteer labour. But the compilers of GNP and economic growth statistics have to be blind to all this because it is not measured in money.

On the other hand, all services and goods that are paid for are added in, regardless of their inherent value. When things are actually bought and sold we can assume that they are worth their price to the purchaser: otherwise the whole exercise, indeed most economics, would fall to the ground. But when things are paid for, not by a purchaser who sees what he is buying and decides for himself whether to pay the price, but out of public revenue drawn by compulsion from taxpayers who have no direct voice in the spending decisions, we have no assurance as to their real worth. In the formative days of national-income theory and measurement, many economists thought that “unproductive” government expenditure, like the cost of military forces, ought to be left out. (Some still do.) But it was seen that this distinction involved value-judgments alien to scientific economics, that defence against war or invasion was as much an economic object as defence against disease, and that the work and pay of a soldier ought to rank on the same footing as the work and pay of a public employee of any other kind.

Consequently most of the cost of the Vietnam war, for instance, has been incorporated in the national income of the United States, as has the whole cost of journeying to the moon and exploring space. They rank on all fours with American government expenditure on education or police or hospitals. Have they added to the welfare of the American nation or the American people? It is for politicians and moralists to answer such questions, not for economists. All that is exchanged for money is grist to their mill: it includes food for the greedy as much as food for the starving, the labour of prostitutes but not of wives, activities which foul the air or water as much as those which clean them up, fighting wars as well as sending aid to poor countries, ploughshares and swords alike. To expect economists to assess the real worth of things is like asking a man with a tape-measure and a weighing machine to measure someone’s intelligence. So long as we realise that this is the limitation as well as the merit of economic calculations, well and good, but too often when people talk about growth the limitation is forgotten.

It is another native defect of national-product figures that being based on cost or price they are unaffected by changes in quality of goods or services. A poor service, if it costs as much, is entered at the same figure as a good one. Take postal services as an example. If, for the same internal first-class-mail charge, you get one delivery a day and collections no later than, say, 4 PM, you are clearly buying a worse service than when you got two or three daily deliveries and collections up to late evening; likewise if the mail takes longer on its way. If the postman will not go up your garden path or beyond the lobby of your apartment building you clearly have to do for yourself, unpaid, what the postman might do and may have done in the past in return for his wage, which is met from the postal dues. There is a loss of real welfare, but no change in the figures of economic income.

Consider another case, of goods rather than services. An automobile is an automobile: such and such horsepower, capacity and other specifications, and costing so much. A million automobiles sold are a million automobiles added to the consumer-durable benefits of their purchasers, and the price is so many billion dollars added to the earnings of all contributors to their manufacture. But what the purchasers enjoy depends on the quality, comfort, safety, convenience and workmanship of those cars. If, for the same price, quality falls, if more minor adjustments and repairs are needed, more little defects appear, the car’s economic lifetime is shorter, clearly the welfare represented by those same billions of dollars is less. If the quality improves, the welfare expands. But the National Product is unaffected.

It can even happen that greater efficiency directly brings a loss of welfare. Thus, for example, when the London Underground introduced new rolling stock which held more passengers, without any loss of comfort, it not unnaturally made the trains less frequent: same speed of travel, same passenger comfort, same fare, less cost for power and drivers’ pay. So “productivity” undoubtedly went up—but the passenger, on average, had to wait longer for his train. Whatever the gain, he had lost.

Another instance of difference between real income and national-income statistics may be seen in agriculture and fishery. Demand for foodstuffs tends to be “inelastic” in advanced societies: that is to say, price-falls tend not to excite proportionately increased demand. Thus a good crop may sometimes fetch less in total than a bad crop, especially where the food is perishable. To avoid complications, suppose that the grower of main-crop strawberries gets from the fresh-fruit market the same total return when his crop is doubled in a bumper season as he does for a normal crop. No change in his income, no change in the consumers’ outlay, no change in gross national product—but the people have enjoyed twice as many fresh strawberries, which for the purpose of this argument we assume to be good for them.

It may be argued that this is fiddling with detail, that good seasons and bad seasons average out, that quality of some goods may go up while that of others goes down, and that “it all comes out in the wash.” So it may well be, provided that there is no long-term trend in one direction or the other. But in fact there are such trends. Here are some of them.

In foodstuffs there is a trend towards better quality (in the sense of standards and grading) off the farm, offset by a fall in quality (in the sense of inherent nutritional value) in the product as delivered to the consumer. Thus, while milling-wheat improves, the nutrient quality of bread has fallen; while milk is purer and more plentiful, more wrapped and processed cheese is sold in proportion to fresh and varied cheeses; fresh vegetables are displaced by frozen, dehydrated and other treated products; supermarkets often do not sell the most economical cuts of meat or most nutritious offal (liver, hearts, brains, sweetbreads, tripe, black puddings and so on).

This substitution of selected, processed and packaged articles may be regarded, no doubt, as a roundabout way of paying for the services of the housewife. Instead of shopping every day or two at a variety of scattered shops or market stalls, and taking the food home to clean, prepare and cook, she can go once a week to the supermarket and get all she wants, most of it prepared for the minimum of culinary effort—meat cuts for the grill, uniform broiler-house chicken which tastes of nothing, instant tea and coffee, instant soup, breakfast cereals, pastry mixes, ready-made puddings, canned or frozen fruit and vegetables, bread and cakes and processed cheese which will keep indefinitely in the freezer, long-life milk... All this means a great saving of time and trouble for one of the most hard-working groups of people in our society. It has to be paid for, of course, but who would grudge the wives and mothers the price of such relief? The point is that the price of food at the point of retail sale, which goes into the GNP figures, is higher, while it covers a decline in the nutritional quality, taste and variety of the food the family eats, and this diseconomy is wholly unaccounted for in the national-income statistics.

A second long-term trend, with far less counterbalance, is a decline in the quality of services, especially government services. Again, this is not uniform: if postal service gets worse, telephone service may get better; if air travel improves, train travel may deteriorate (though here Britain is far better off than the United States). But generally those activities which depend most on direct individual service—like street cleaning, garbage collection, postal deliveries, family doctoring, gardening, hotel and domestic service, home deliveries from shops, plumbing and decorating, or car repairs—tend over the years to decline in quality or, to put it another way, in value for money.

This is economically inevitable. We need not blame the spirit of the times, democracy, indolence, government inefficiency, trade-union or professional greed, the welfare state, or other popular scapegoats. The fact is that as the general standard of living rises, in money-measurable terms, the performers of services demand their share of the rise, even though the nature of their functions largely prevents them from contributing to higher productivity (as those engaged in manufacturing industries can) by means of factory efficiency, mechanisation, automation, mass production or technical innovation. To put it a little over-simply, growth takes place in goods rather than services, the services get their share of growth, so the cost of services in terms of goods goes up, and their value for money (at a constant price-level) goes down. We could change this process only by arresting the economic growth in manufacture and like industry, but this would hardly please either the workers in manufacture or those in services. There may indeed be ways of increasing the efficiency of service trades, but they are usually difficult or slow to apply, because services tend either to come in small packets, like automobile repairs, or to be government monopolies, like the posts and the refuse collection.

Yet another general trend is towards a decline in consumer options. The supermarket is a useful invention, but the woman who shops there can buy only what is on the shelves. The essence of supermarket business is rapid turnover and highly efficient use of shelf-space, which impels a strict limitation on brands and varieties on sale. Many supermarket chains make or package their own brands of food and other domestic articles, and can hardly be expected to offer their competitors’ goods, unless there is a big difference in price or kind. Even the smaller and more specialised shops are driven by the same sort of competitive pressure (often making itself felt in high rents) to reduce their range of offerings. In a degree a similar process operates in other fields of retail trade—American delicatessens, British pubs, ironmongery and hardware shops, and so on. Restriction of consumer choice means reduction of consumer satisfaction.

Choice is itself an ingredient in personal welfare—a marginal one, no doubt, but important, like salt or herbs in cooking. To be deprived of choice is a loss of welfare. Ask the Russians, who under a centralised state-controlled economy have been offered the one sort of article that the Communist authorities decided upon, and nothing else: one kind of stockings, or lavatory fixtures, or whatever, take it or leave it. Modern industry and distribution have limited the consumer’s choice in many fields. The supermarket stocks only two or three brands of particular groceries, sometimes only one; in public transport we rarely now have choice or competition between trains and buses; most major American cities have only one morning or evening newspaper. In other spheres choice has enlarged: in popular cars since the days of the Model T Ford or even the snub-nosed Morris, in domestic appliances, in cosmetics and drug-store articles, in television and radio (though here again a commercial tendency to concentration is operating), and in many other items. The point to be made is that none of this, upward or downward, is reflected in the measurement of national income or economic growth.

This line of argument has to defend itself against ideological criticism. The accusation is that this whole approach is typically middle- or upper-class. To comment adversely on economic changes which have accompanied an undoubted and progressive rise in average standards of consumption is, according to these critics, to idealise a bygone world of privilege in which those who had the means could buy all the varied food and household goods they wanted, and be waited upon by servants and attended by private professional skills, while the poor were lucky enough to have enough to eat. It is the working-class family, so the argument continues, that has benefited most from the modern ways of transporting, preparing and distributing food and other necessities. The rich may grumble that they can no longer buy good Stilton or Caerphilly at the little grocer’s which has been forced out of business, but the poor get good cheese, wholesome, standardised and cheap, and former luxuries are brought within their daily grasp.

Even if the premise about working-class needs and habits were true, which it is not, this is to misunderstand the argument. It is a purely factual one. Certain trends exist, whomever they may injure or benefit. They affect the general standard of real welfare, but they make no difference to the common statistics of national income or economic growth. There is no imputation of moral or class values. Nor is there any imputation that economic advance and technical improvement are bad. Of course they have brought an immense access of well-being, and relief from hardship and anxiety, to the people as a whole, though to some more than others. It cannot but be right to observe that there have been unmeasured side-effects, which may also be unevenly distributed, though not in the way or to the extent that the above criticism implies.

Figures of GNP and economic growth in fact take no account of changes in the distribution of national income between different classes and groups. This may be, and in some countries is, a very important element in the welfare of the people. As a first example, if national income in, say, a Latin American country is very ill distributed between a few rich and many poor, the sum total of human welfare is plainly enhanced by a fairer sharing of the nation’s current wealth, without any increase in its total. As a second example, inflation in any country effects a re-distribution of national income between those advantageously placed to keep up with it and those in weak positions who are left behind. This does not increase the national welfare. But the GNP may be temporarily increased, even when discounted for changes in price-level.

Political rights and personal freedoms are also a very important element in human welfare in society. Their absence or restraint vitiates any deduction as to genuine well-being or progress from figures of economic output or growth. It is possible that a phenomenal growth-rate could be achieved by a slave society, and that the slaves themselves could be shown to enjoy a higher standard of consumption than poor but free men. Yet nothing would thereby be demonstrated either about real welfare or about economic growth-rates as its index.

Still another social trend, ignored by GNP and growth-rates, is of even greater significance. Almost all the technical changes that have brought about the great economic advances of the twentieth century involve sophisticated capital equipment and an economy of direct labour. This is as true of agriculture and rural industry as it is of urban factories. Higher standards of agriculture mean fewer workers per acre. The result, all over the world, is inevitably a highly disproportionate growth of cities and large towns as against the small towns and countryside. At the same time, rising standards of life lead to a disproportionate growth of the work-force of service industries as against manufacture and primary production. This compounds the same effect; for services have to be rendered where the people who use them are, which means increasingly in the cities—apart, of course, from those services rendered to tourists and others travelling through rural areas. Therefore, alongside the rise in economic output there have occurred and will continue to occur all those unmeasured consequences of massive urban growth which so greatly trouble our generation. The urban problem will be examined more fully in a later chapter. Here it is enough to say that increased ugliness, strain and sense of alienation are part of the unmeasured debit that offsets the gain of economic growth as measured in the familiar statistics.

So, to sum up, the statistics of economic growth, calculated from GNP, measure something undoubtedly, but they do not measure growth in national welfare. Statistically, economic growth can go on while national welfare declines. Or the opposite can happen. Economic growth has been relatively slow in Britain, and almost at a stop in the last couple of years, yet Britain has not thereby become a relatively or absolutely less comfortable and pleasant place to live in, nor has the welfare of its people deteriorated, except for higher unemployment, which can co-exist with high as well as with low rates of economic growth.

On the other hand, all the sustained and measured growth of the economy of the United States since World War II has not spared her from afflictions which take a lot of the real value out of higher income-levels: rising crime and violence, drug problems, city deterioration, smog, slums, minority unrest, fears and anxieties on all bands. One could make such observations, on the good side or the bad, about any country in the world, with high growth or low.

It is when our income falls (for instance through retirement, change of job, slackening of business, or some misfortune) that we personally learn the true relative values of the things we have enjoyed, and which of them can be sacrificed with the least loss of real welfare. Traditions, family relationships, security of home, natural surroundings, physical health, all rank high; many trappings and trimmings of modern life rank low. Perhaps the most important element in true national welfare, when basic needs are met, is a sense of stability, of “belonging” humanly and physically in a place, a community, a society, a country. This has little to do with growth: indeed the reverence for growth, the envious striving for more income, more consumption, has done much to undermine it. Some of the richest countries with the highest past and present growth-rates have much less of such values than others economically behindhand but socially more stable and secure.

Statisticians and economists are of course aware of the shortcomings of their figures as measures of well-being. “Economists,” writes Professor Walter Heller of the University of Minnesota, “labour under no illusion that GNP is a satisfactory measure of welfare or that it can be turned into one.” A big quest is on for an alternative, in what are professionally called “social indicators”. A British financial commentator recently wrote:

One of the most fundamental goals of traditional economic argumentation—a high rate of growth of GNP—seems to be so far out of our reach, at least in this country, that we might as well stop wasting our energies on a further search for the economists’ means to the questionable end and turn our attention instead to the best policies for growth and social content as measured by quite different criteria. [See note 2]

The deduction appeals more than the somewhat sour-grapes premise.

The interest in social indicators in the United States, the world leader in this field, seems to have originated in the assignment given by the National Aeronautical and Space Administration to the American Academy of Arts and Sciences in 1962 to study the possible effects of the space programme on American society. [See note 3.] One of those responsible for this research, Raymond A. Bauer of the Harvard Business School, had posed this question: “If we have highly organised economic indicators, why can’t we set up a system of social indicators?” The report of the Presidential Commission on technology, automation and economic progress, published in February 1966, recommended the creation of a system of “social accounts” covering four fields: (i) the measurement of social costs and net returns of economic innovations, (ii) the measurement of social ills (EG crime, family disruption); (iii) the creation of “performance budgets” in areas of defined social needs (EG housing, education); and (iv) indicators of economic opportunity and social mobility. (The ambitiousness of this programme almost takes one’s breath away.) It was the appointment of an expert panel on social indicators, following a message from President Johnson to Congress, that led to the publication by the Department of Health, Education and Welfare in January 1969 of the much-quoted “Towards a Social Report.”

President Nixon did not follow quite this line. He assigned to the Bureau of the Budget the task of setting up a consistent system of social indicators and publishing the results; and in July 1970 he announced the creation in the White House of a National Goals Research Staff; responsible, among other things, for preparing, over a period of time, social indicators that would reflect the present and future quality of American life, and the direction and speed of its changes.

In Britain the lead has been taken by the Central Statistical Office, which in December 1970 published the first in a new series of collected statistical studies related to social policy. In France it has been taken by the Commissariat du Plan, which ever since 1967 has been compiling and analysing statistics having social significance. Private economists have not lagged behind these public initiatives. Some, like William Nordhaus and James Tobin (in “Is Growth Obsolete?”, National Bureau of Economic Research Colloquium, December 1970), have sought to quantify the deficiencies of GNP and NNP as indicators of social welfare, though this pioneer work is not yet scientifically convincing.

Thus great efforts are being made to forge the tools for the job of finding some measurement of national well-being closer to the realities of the matter than Gross National Product and its derivative, economic growth. The task is hard. In many relevant fields, though we already have some figures, they tend to be negative, so that the picture we might draw from them would be rather like a description of business activity based only on figures of unemployment, or of the state of religion based on particulars of people who do not go to church. For example, we have many statistics which tell us how ill people are (hospital admissions, causes of death, incidence of diseases, and so on) but few to tell us how well they are: statistics of illiteracy but not many of what people read. Positive measurements are much harder to come by than negative: people do not go to the clinic to say how well they feel, nor to the labour exchange to say they are happy in their work.

But there are worse difficulties than that. Some matters of social welfare are highly personal and subjective. How can you measure the value to a family of having a grandmother living in the same neighbourhood? All you can measure is the rise in parental stress and juvenile delinquency when families are transplanted away from a three-generation background. Some people like cities, others abominate them. How do you balance the scarring of a lovely countryside with caravan parks against the pleasure of the people in the caravans?

Still harder than the job of measuring such intangible things as go into social welfare, and of collecting mass figures in a reliable form, is that of combining such social indicators as we have or may invent into a composite picture of the welfare of a nation or community. How do you add up, in an index of national welfare growth, better education, purer water or air, warmer houses, healthier teeth, deeper traditions, longer vacations...? How do you subtract anxiety about crime, inter-racial conflict, the uglification of towns and countryside...?

One of the foremost American experts in this field, Edward F. Denison of the Brookings Institution, has stated bluntly: “A single generally acceptable index of welfare cannot be constructed.” [See note 4.] He goes on to cite some of the reasons. Certain of them have already been mentioned here. Others that he names include the difficulty of establishing the real costs of labour, having regard to working hours and all the conditions of work; inability to measure the real cost of “abstinence” or saving; changes in people’s needs over time; the appearance of new products or benefits (EG television, modern medicines); lack of agreement on better or worse distribution of national income; changing tastes; fears and anxieties not supported by reality; the impossibility of assigning weights to the different aspects of welfare even if we could measure them. And he concludes:

We need, can obtain, and should obtain additional information, including statistics, on many aspects of American life that affect welfare. We can and should explore ways of presenting and analysing such information in a comprehensible form... However, we cannot obtain a comprehensive index of welfare. There are likely to be pressures to make ad hoc changes in the existing national product, measures that, it is supposed, will move the national product series closer to a complete welfare measure in one way or another. Such suggestions should be welcomed if they improve the measurement of the nation’s output. I would myself urge regular publication of series for NNP (Net National Product) and national income, as well as GNP, in constant prices. But some suggestions to change the measurement of national product will derive from confusion between output measure and a comprehensive welfare measure. Such proposals must be rejected. GNP and NNP cannot be transformed into a comprehensive welfare measure. Efforts to do so can only impair their usefulness for the very important purposes of both long-term and short-term analysis that they now serve well.

One can only agree with Professor Denison. It has not been the purpose of this chapter to prove that statistics of National Product and of changes in it described as economic growth are bad or useless; only that they should not be taken as indices of national welfare or of its growth or decline. Nor has its purpose been to claim that we have, or could readily create, some substitute which in a single set of figures would measure those things; but only to demonstrate that the problem is extremely complex, and to beg that no judgment of value be imported into the economic-growth figures, which they themselves make no pretence of possessing. The application of values must be piecemeal, it may involve conflicts of values (which are, after all, a commonplace of experience), and it cannot be definitive. Nevertheless it must be attempted if national and international policies are to be rightly directed. We cannot get everything into the picture, but we can get in some of the most important things, besides material income, that make up the quality of life, and that play no part in the accepted measurement of national product and economic growth.


Economy in the University of Oxford, in The Social Framework (4th edition, 1971), which includes as excellent an analysis as can be found of national capital, income and growth for beginners in the study of economics.


Note 2. Joe Rogaly in the Financial Times, 30th March 1971.


Note 3. I gratefully owe this historical account to Bernard Cazes, of the Commissariat du Plan de l’Équipement et de la Productivité, Government of France, in a paper prepared for a Ditchley conference on social indicators in April 1971.


Note 4. Survey of Current Business, January 1971.


Created from a nineteenth or early twentieth century text by Athelstane E-Texts