Introduction

Canadians enjoy one of the highest living standards in the world. In strictly economic terms, Canada is one of only about a dozen countries with a national income per capita over $20,000 (unless otherwise stated, all values in this report are in Canadian dollars). More broadly, because of the quality of life resulting from benefits such as educational opportunities and the health of our citizens as well as from material well-being, Canada has, since 1994, been ranked first among all nations by the United Nations in its human development index. Many international experts regard Canada as the best place to live.

It was not always this way. At the time of Confederation, Canada was a relatively underdeveloped and sparsely populated country with questionable prospects. In the 130 years following, however, there has been an extraordinary growth in living standards. In the past hundred years alone, real incomes have increased tenfold (Lipsey 1996: 5). All of the quantitative measures presented below attest to the strong economic performance of the Canadian economy over the decades and the resulting benefits to the vast majority of Canadians. But economic indicators alone cannot tell the full story of the improvement in living standards.

In 1996, the typical Canadian could afford not only many more goods and services than his counterpart two generations ago but--and this is a key point--a whole array of different and much better commodities. Let us start with the basics. In 1951, there were 3.4 million households in Canada, of which 74 percent had indoor running water, 64 percent had exclusive use of a flush toilet and 57 percent had exclusive use of a bath or shower. By 1996, Canada had 11.4 million households in which 99.8 percent contained full bathroom facilities. In 1951, only 50 percent of Canadian kitchens had an electric or gas range and only 47 percent had a mechanical refrigerator. Many households cooked food on a wood stove and stored food in an ice box. By 1996, 99.6 percent of households had an electric or gas oven and 99.6 percent had a refrigerator. But the modern kitchen typically has much more than this. By 1996, 57.1 percent had freezers, 85.3 percent had microwave ovens and 47.7 percent had dishwashers. These items have greatly reduced the time and trouble of meal preparation and cleanup, and thereby increased the potential leisure time for most families (Statscan 1951: tables 98, 104; 1996a; 1997b).

To take advantage of the increased time for leisure, the typical household now has a wide variety of facilities that were not even in the imagination of folks in the 1950s. In 1996, 99 percent of households have radios (compared to 92 percent in 1951) but, in addition, 99 percent had at least one colour television, 74 percent had cable television service, 84 percent had a video cassette recorder, 53 percent had compact-disc players and 32 percent had home computers. None of these latter facilities were even available in 1951 (Statscan 1951: tables 98, 104; 1996a; 1997b). I hasten to add that the fruits of modernity have not been universally welcomed; more gadgets and facilities do not guarantee happiness to the purchaser. Nevertheless, Canadians have enthusiastically embraced these items, regard them as beneficial and enjoyable and continue to look forward to new home-entertainment technologies on the horizon.

Households were more crowded 40 years ago. In 1951, the average household contained 5.3 rooms and 4.0 persons. In 1996, there were 5.9 rooms per household and only 2.6 persons, on average. Only 59 percent of households had a telephone and only 42 percent an automobile in 1951 compared with 99 percent and 74 percent respectively in 1996 (Statscan 1951: tables 98, 104; 1996a; 1997b). People not only worked harder and longer to cook and clean in the 1950s but, for those in the labour force the standard work week is now about 13 percent shorter than in 1951 (Gunderson and Riddell 1993). Workers not only work fewer hours than their counterparts 4 decades ago but have more benefits, more holiday time, safer working conditions and are much more likely to have a pension plan funded by the employer. People now live healthier and longer than 40 years ago. Life expectancy has increased by more than 10 years since the 1950s (Statscan n.d.1 [Cat No. 82-221-XDE]). And, the average wage earner now earns almost three times, in real dollars, what the average worker earned in 1951.1

What has caused this dramatic improvement in living standards? In a recent essay, Richard Lipsey (1996), one of Canada's foremost economic thinkers, argues that technological improvements essentially explain the impressive economic growth in Canada and other countries. The process of saving, investment, and capital accumulation inherent in the neoclassical model does not begin to explain the sort of living standard that Canadians enjoy today. It has been technological innovation, in Lipsey's view, that is primarily responsible. New technologies--for example the use of electricity as a power source replacing steam and, most recently, computerization of almost every aspect of modern life--fundamentally change the nature of the production process, use fewer resources than before, and permit the fairly inexpensive mass production of new commodities. New technologies, along with training and education, make workers more productive and allow real wages to rise.

Fundamental technological change has a pervasive impact on the economy and on society. Computers, for example, have not only changed the method of production and the range of useful commodities available but have already transformed the way we communicate with each other, the way we are being entertained and the way we learn. But adaptation to new technologies is not frictionless. The more radical the technological change, the greater will be the adjustment to it. Change upsets normal habits and patterns. More ominously, it can affect workers and their families adversely by making certain skills obsolete. Some of the long-term rise in unemployment since the mid-1970s can be attributed to the deep structural adjustment following upon the adoption of the new information and communication technologies.

Lipsey argues persuasively that technological improvements and rising living standards are good things but do not happen automatically. There are a number of necessary conditions that need to be in place for economic growth to occur. He mentions as critical ingredients the freedom to innovate, decision-making about the private use of resources through markets rather than through political channels, a system of rewards and penalties, the rule of law, and free scientific enquiry as critical ingredients. In summary, and by way of advice to lesser developed countries, he emphasizes the importance of ``an open society with freedom of the market to operate in a relatively unrestricted way ...'' (Lipsey 1996: 42). He warns us against neo-Luddites who, under the banner of environmentalism, religion, or political correctness, are resolutely anti-technology and blindly opposed to market-based solutions. The simple fact is that, wittingly or not, they would condemn us to lower standards of living. At the same time, Lipsey points out the need to design schemes that would take the sting out of poverty and displacement, whether brought on by technological change or not, without destroying self reliance.

There are many enduring false ideas about technological change. Let us consider a few.

False idea 1: new technologies destroy jobs. While new technologies do replace old, less efficient ways of doing things, they also tend to create far more jobs than are lost. This happens partly because of the ``spin-off'' opportunities that inevitably accompany new technology but mainly because new technology, by lowering production costs, lowers prices and offers consumers more purchasing power. This extra demand creates jobs.

False idea 2: new technology results in more poverty. History suggests just the opposite. Revolutionary technological change in this century has improved the living standards of all citizens, including the poor. There is now in Canada much less poverty (in the sense of deprivation of basic needs) than there was even 50 years ago. And, the fastest improvement in the material well-being of those who were poorest occurred prior to the big expansion in social programs in the 1970s.

This is not surprising. Without denying the inevitable displacement that often accompanies rapid change, technological innovation invigorates an economy, increases economic growth and, ultimately provides more jobs than are lost. Earnings from employment offer best way for the poor to improve their lot. The impressive record of impoverished immigrants to Canada in this century is a prime example of this process.

False idea 3: new technology means greater inequalities. Technological innovation is the engine of economic growth. The benefits of economic growth are rarely evenly distributed. In the first instance, free markets tend to reward intelligence, risk-taking, and hard work. But those qualities are not equally distributed in the population. So, initially the able and ambitious benefit more from rapid economic growth. Over time, however, the benefits of job creation, better infrastructure, and more comprehensive insurance schemes become widely dispersed. This diffusion has the potential to benefit the poor and the less able differentially. The evidence presented below supports this conjecture. Thus, once all things are considered, economic growth driven by technology does not increase inequality. All of us, poor and well-off alike, stand on the shoulders of the creative geniuses and entrepreneurs of the past.

Productivity, often measured as output per worker, plays an important role in the determination of living standards. Improved technologies permit workers to be more productive and greater productivity allows real wages to rise which, in turn, permits higher levels of consumption. Performance and Potential, a recent report from the Conference Board of Canada (1996), reveals that productivity in Canada rose strongly over the post-war period to about the mid-1970s and has grown very slowly since then. Thus, despite massive structural change and the widespread adoption of new technologies throughout the economy, there has been little evident payoff in terms of productivity. This so-called ``productivity paradox'' could be due to the inability of current statistical measures to capture the fundamental changes, especially in the area of services or it could be that the payoff is still to come. Lipsey favours the latter explanation. In any case, the apparent slowdown in productivity over the past 20 years or so may account for some of the recent slowdown in key indicators of the standard of living.

Purpose and plan of the report

This report attempts to answer the questions: What has happened to Canadian living standards in recent decades? Are we better off than we used to be? Are the poor getting poorer? What has happened to the middle class?

A fairly large body of statistical data bearing on these questions is available but statistics alone never tell the full story. I will provide some interpretation, commentary, and critical analysis to assist the reader in making informed judgments. However, since the intent is to relay this important information to a broad, lay audience, the report is largely free of technical jargon. This is, mainly, a descriptive report. It pulls together information about living standards from a wide variety of sources and presents it in a clear, concise fashion. The tables and graphs show key relationships over a longer period of time than is usually presented in reports of this type. As well, many of the graphs are new, showing relationships and trends that have not been presented elsewhere. For the interested reader, the data appendix at the back of this report has tables with complete and up-to-date values for many important variables. It is my hope that this report will serve as a useful resource to anyone working in the field of social and economic policy.

The report begins with some important definitions and a brief discussion of measures and indicators. Second, in the section The Big Picture the key measures relating to Canadian living standards are presented and discussed. Third, much more detail is provided on a variety of aspects of how we live. Finally, an attempt is made to construct an index intended to capture the essential components of the standard of living.

I wish to acknowledge the cooperation and assistance of a variety of officials and researchers at Statistics Canada. In many cases, people went out of their way to provide me with data on a timely basis. As well, I am grateful to the ``movers'' behind the Data Liberation Initiative. This program has made Statistics Canada databases more easily available to researchers and this report has greatly benefited from that initiative. I would urge Statistics Canada to extend the program to older databases as soon as possible.

I am particularly indebted to Andre Roy who developed a computer program to generate income distributions and Gini coefficients especially for this project. As well, I wish to thank Aaron Bertrand and Krista Lariviere for diligent research assistance.

Defining ``standard of living''

The term ``standard of living'' used in this report refers to the material well-being of individuals and families, to the extent that that can be determined. We use quantifiable measures of economic well-being such as income, consumption, assets, and facilities to gauge the standard of living. While this approach is common in economics, it nevertheless limits the range of factors under consideration. It ignores, quite deliberately, qualitative aspects of the standard of living.

This approach has come under increasing fire in recent years; geneticist and environmental activist David Suzuki, for example, has been very critical of quantitative economic measures and the economics profession for their apparent ignorance of anything beyond gross domestic product (GDP).2 He has argued that so-called economic growth is not progress at all if it comes at the expense of environmental deterioration. Suzuki is not alone in calling for more broadly based measures of economic progress.

Over the years, a variety of attempts have been made to capture societal well-being with a broader measure than GDP per capita. The latest effort comes from Redefining Progress. This organization, based in San Francisco, has developed a Genuine Progress Indicator (GPI) that incorporates numerous factors such as crime and divorce rates, pollution and other environmental damage, leisure time, the value of household and volunteer services, life span of consumer durables, and income inequality as well as the standard economic measures. Tracking the GPI for the United States since 1950, the group found that after the mid-1970s, GPI per capita declined, while GDP per capita continued to increase. The major problem with the GPI is that all factors, including crime, divorce, the value of leisure and volunteer work are combined (somehow) into a single monetary aggregate. Nevertheless, the concept is an interesting one.

The Report on Canadian Living Standards is far less ambitious but the problem of finding an adequate indicator of the standard of living persists. Typically, measures of improvements in the living standard use the ``per-capita'' approach. For a whole society, we use GDP per capita and, at the micro level, we use income per household member. As Ringen (1991) has pointed out, this approach makes sense intuitively: a given income can buy a given consumption and this consumption must be divided between the individuals it supports. However, this is clearly not sufficient. What it ignores chiefly is economies of scale: larger families involve some efficiencies in the production and consumption of well-being within the household. It is clear, for example, that a single person living on $10,000 has a far lower living standard than each member of a 5-person family with a total family income of $50,000.

This issue becomes important if, over time, there are demographic changes in fertility and marital stability that affect the average size of the household. Ringen (1996), for example, using equivalent incomes after accounting for household economies, found that per-capita measures have exaggerated the rate of growth in well-being in Britain between 1976 and 1986 because of the trend to smaller households. The degree of overstatement was found to be about 10 percent or about 1 percent per year. Ringen suggests that the differential would be much greater in the comparison between societies with much different household structures--between industrial and developing nations, for example.

Another problem with per-capita measures is that they implicitly assume that everyone in a given household enjoys an equal standard of living. This is not likely to be the case and, indeed, the differential contributions and consumptions within the family may be quite large. Researchers have found no easy way to adjust for this particular inadequacy. There are additional problems with income and consumption data (per capita or not), which we discuss below. Given the data currently available, it is impossible to correct for most of the difficulties. A partial adjustment, correcting for one problem while leaving others uncorrected would not be a desirable solution, especially since some considerations tend to counteract others. The preferred option, in my view, is to broaden the perspective by using a wider variety of variables relating to well-being than is usually presented in examinations of living standards. Included in the basket of indicators will be the familiar per-capita measures, unadjusted.

Quality of life versus quantity of life

Economists are often criticized for their apparent love affair with facts and data. It seems, to many non-economists, that we shine our floodlights on considerations that can be carefully quantified to the exclusion of everything else. To some, economists are those who ``know the price of everything and the value of nothing.''3 I disagree. In my experience, economists typically develop quite elaborate and comprehensive theories. Those engaged in either empirical or applied researchers usually take great pains to point out any omitted but important considerations that could not be quantified or had no valid proxy.

Life and the standard of living are, of course, far more than material ``things.'' Much of what truly enriches our lives is not measurable and has no monetary value. Friendship, political freedom, appreciation of the beauty of nature, the quality of one's imagination, the quality of one's environment, the delight of discovery and spirituality--in the broadest sense of that term--are examples of important yet elusive considerations affecting well-being. And these considerations may be more important (far more, indeed) than purely material circumstances for all save the impoverished. It is well recognized by economists that our overall well-being is affected by both economic and non-economic considerations. The problem is that many of these non-economic factors are very difficult or impossible to measure. In most cases, we simply have no reliable data to work with.

In this report, I have deliberately narrowed the focus of discussion to material living standards. The one exception is life expectancy. This variable captures, I would argue, in one number all of the various influences on health and environment that ought to be part of any discussion of living standards.

Statistics Canada

Since World War II, Statistics Canada has increasingly been providing Canadians with more and better information about how we live. For example, in 1948, Statistics Canada conducted the first national survey of family expenditures for use in the construction of the consumer price index. It has continued its family-expenditure survey on an occasional basis to the present. In the 1951 census, Statistics Canada began collecting fairly detailed data on earnings and subsequently expanded their questionnaire to cover all forms of income. The census has also provided interesting information about the facilities, the number of rooms, and the condition of Canadian households since 1951. Since the late 1960s at the latest, Statistics Canada has conducted regular (and now annual) surveys of household facilities. And, beginning in 1971, Statistics Canada has published the results of their annual income survey (Survey of Consumer Finances) conducted in conjunction with the broader labour-force survey. Of course, there are a wide variety of other surveys, statistics, and special reports relevant to non-economic aspects of the standard of living--everything from single-parent families to the use of leisure time. Data drawn from the wealth of statistical sources of Statistics Canada form the basis of the evidence used in this report.

Choosing the right indicator

Finding a reliable indicator of material well-being is not as easy as we might think. Some measure of assets or wealth would clearly be desirable. However, there is little systematic data on the wealth of Canadians and what little there is from reputable sources such as Statistics Canada is dated. As a substitute, indicators of the trends in ownership of particular assets and facilities, information that is more readily available, are clearly useful. Measures of income (real and nominal, pre-tax and after-tax, and so on) have tended to be the choice of economists examining trends in living standards. However, income as an indicator of material well-being is badly flawed and, while it cannot be discarded and is used extensively in this report, it is important for readers to be clear on the problems with income.

Income

The income data used in this report comes either from the census or from Statistics Canada's annual income survey (Survey of Consumer Finances). In each case, respondents are contacted by phone or in person and asked to provide a detailed breakdown of their income from various sources during the previous year.

Ideally, a person's income should indicate command over goods and services in the marketplace since income represents portable purchasing power that can be used to buy the commodities that generate ``utility'' or well-being for the household. Thus, the higher one's income, the greater the implied material well-being of the members of the household.

But the income reported to Statistics Canada is different, sometimes substantially different, from the household's ``true'' command over goods and services. This discrepancy is due to a number of factors:

Sloppiness or laziness on the part of the respondent

Some respondents regard the Statistics Canada survey as an imposition and treat it in a cavalier fashion. As a consequence, they may not examine their personal and income tax records to ensure accuracy but may instead give rough estimates or respond casually. Interviewers are not supposed to question responses even if they suspect the values are inaccurate. It is the case that all responses are edited and some adjustments are made where other information makes it clear that the original responses were not correct.

Non-response

The annual income survey is a carefully constructed random sample of Canadian households designed to represent all Canadians. However, about 20 percent of those chosen to be in the sample do not respond fully or do not respond at all. While Statistics Canada uses sophisticated techniques to impute information to these households so that they can be included, this is a large gap to fill. The accuracy of the imputation is considerably reduced if the households not responding are not themselves randomly distributed.

Unreported and under-reported incomes

Certain types of income are routinely under-reported. Statistics Canada reconciliation checks reveal, for example, that unemployment insurance payments and welfare benefits are about 20 percent and 40 percent under-reported, respectively (Wolfson and Evans 1990: 26). This is a particularly significant omission for the poor since about half of low-income households have one of these two as their major source of income. Wages and salaries from employment, on the other hand, are almost fully reported.

Smith (1996) of Statistics Canada notes that a growing underground economy, driven largely by tax evasion, is also responsible for some of the unreported income. Substantial under-reporting in the construction and renovation industries, among businesses selling used vehicles and appliances, and certain small business retailers (especially in the area of personal services) likely totals about $10 billion annually. As well, under-reported tips in the restaurant, alcoholic-beverage, and taxicab industries amounts to several billion dollars. Finally, the incomes that are earned from a variety of illegal activities such as prostitution, gambling, drug dealing, tobacco and alcohol smuggling and petty theft go almost entirely unreported. In all cases, of course, these incomes contribute to the well-being of their recipients even though they are not reported.

In-kind income

Many Canadians receive a variety of gifts, free benefits, and services that contribute to their material well-being but do not cost them anything. Included here would be such things as subsidized rents; monetary and non-monetary gifts from family, relatives, friends, and charitable organizations; and medical, dental, and drug benefits related to certain social programs. Because recipients of ``in-kind'' income do not have to pay for items that most others do, their reported income further understates their true living standard.

Business losses

Readers may not be surprised that many Canadians have very low incomes. What may be surprising is the large numbers of households with negative income. In 1994, for example there were 12,000 households with reported incomes less than zero. Virtually all of them had declared large business losses against other income sources. Such declarations come about for tax purposes and do not accurately reflect the true standard of living of these households.

Student loans

Loans of any kind, including student loans, are quite rightly not counted as income. However, loans are a way of transferring future income into the present and do add to the household's standard of living in the current period. For the most households, excluding loans as a source of ``income'' is a relatively insignificant omission. Not so with student loans. This is because student loans are a big part of the student's overall resources. During the 1980s, for example, student loans were about 20 percent of total student resources (Statscan n.d.1 [Cat. No. 52-179]). Due to higher real tuition fees and lower student incomes in the 1990s, student loans are now likely to be an even greater portion of resources. By not including student loans, the standard of living of independent students appears much lower than it in fact is.

Miscellaneous considerations

There are a variety of other circumstances in which reported income will belie the household's true standard of living. People in religious orders, who often live under a vow of poverty, would be one example. There are, in addition, several thousand people (Old-order Mennonites and Hutterites, for example) who, following the regime of their religion, willingly live ascetic but quite fulfilling life-styles. They eschew most aspects of modernity and have market incomes that are very low or zero but which again fail to gauge their true standard of living accurately. And, there are a small but apparently growing number of families that have decided to retreat from urban life and live a far simpler, poorer but more satisfying life-style on farms and in small communities.

A quite different circumstance in which reported incomes will understate the household's true living standard is the case of part-year families. For a variety of reasons (marital breakup, immigration, de-institutionalization, etc.), a household head, asked to report on income for the previous year, will only have income for a portion of the year. For example, suppose an immigrant family came to Canada in the fall of 1996. Their ``Canadian'' income for 1996 might only be $10,000, which reflects a very low standard of living if viewed as a total annual income. Their reported income belies their rate of income and therefore their true living standard.

Since the Statistics Canada income surveys omit certain geographical areas and some categories of persons, this may introduce additional bias into the reported incomes. For example, residents of the Yukon and Northwest Territories are not included in the survey. However, with a combined population of about 100,000 (roughly one-third of one percent of Canada's population) and average incomes at or above the Canadian average, this omission is unlikely to introduce significant bias either way. The exclusion, however, of households located on Indian reserves is likely to be important. While their is no precise determination of the number of people living on Indian reserves, largely because most bands decline to participate in the census and similar surveys, estimates by Statistics Canada and the Department of Indian Affairs suggest that there are at least 200,000 people living on those lands (Sarlo 1996: 154). More importantly, various reports inform us that people living on reserves are much less well off in material terms than Canadians in general. So, even though the reserve population is probably less than 1 percent of the overall population of Canada, living standards are somewhat overstated and poverty somewhat understated by this particular omission.

In sum, it is reasonable to assume that the combined effect of under-reported incomes and the underground economy, income-in-kind, omitted resources such as student loans, business loss tax write-offs, part-year families, voluntary austerity, and native reservations results in Canadian living standards being understated by reported incomes. Further, it is my view that reported incomes are much less reliable a gauge of living standards at low incomes than at high incomes, mainly due to the underground economy and the under-reporting of government transfers. Finally, because the underground economy appears to be growing and because social assistance and unemployment insurance caseloads are now much higher (proportionately) than in the past, it may be that the data on income is less reliable than ever.

Despite the well known limitations of reported income, it is widely used in studies examining poverty, inequality, and material well-being. And it will be used extensively in this study, with some reluctance. The reader is urged to recall the disadvantages of the income as an indicator and interpret results with appropriate caution. As will be seen, income can be a useful and interesting guide and should not be entirely discarded.

Consumption

Information about the consumption levels of households is an increasingly attractive alternative to income as an indicator of economic well-being. This information is also increasingly available to researchers as national surveys of family expenditure are now conducted more frequently and more broadly than in past decades and published as microdata files.

The main advantage of consumption information is that it is much less likely to be under-reported because the incentive of tax evasion is absent. It also avoids some of the understatement of living standards due to the exclusion of money gifts. We presumably get a more accurate picture of the material well-being of people with business losses and some students if we use consumption. On the other hand, data on family expenditure does not deal with some problems, such as in-kind gifts, bias due to inaccurate responses or non-response or voluntary austerity any better than income data. However, unlike the income survey, the family expenditure data excludes part-year families and students who live with their families for part of the year.

So, consumption data is not perfect but it will give us a better indication in many respects of the household's true standard of living. Consumption, after all, is a step closer than income to what really interests us--utility or well-being. Moreover, the whole life-cycle analysis of income and consumption persuades us that consumption is a more reliable gauge of economic well-being than income.

The life-cycle hypothesis

Economists assume that households make decisions rationally and that they operate with a view to a much longer time horizon than just the current period. According to this theory, household heads formulate some notion of their ``permanent'' income (roughly, an expected lifetime average) and gear their consumption to that permanent income rather than to their current income only. As income varies from time to time, the household's consumption will remain relatively stable because transitory income changes have only a small impact on permanent income. Households will adjust their estimate of permanent income (and thus consumption) according to new information as long as it impinges upon expected lifetime income. Thus, according to this theory, we might expect newer households to spend more than their income for a while as they gear their consumption to what they expect will be a higher average income. We would expect, as well, that families in their peak earning years--usually when the wage-earners are in their 40s and 50s--would be net savers as they will already have paid for most of their larger durable goods and are preparing for the time after retirement when their earnings will fall, probably to zero. So, in this phase, the theory would predict that income would regularly exceed consumption. Figure 1 depicts the typical lifetime pattern of income and consumption that the life-cycle hypothesis predicts.

The empirical evidence strongly supports the life-cycle hypothesis. Most people apparently do gear their consumption to a notion of permanent rather than current income. Consequently, observed consumption is much more stable than observed current income. In some years, a family may have exceptionally good income, due, say, to modest lottery winnings, an inheritance, or a special bonus at work, and, in other years, may have exceptionally poor income due to layoff from work or ill health. Throughout, however, consumption patterns remain fairly constant and are a better reflection of the material living standard of the family than the ups and downs of income.

Click Here to View Figure 1

The fiction of accurate measurement

An additional problem with the measurement of living standards and comparisons over time is that, using the current measurement methodologies, we cannot account in full for fundamental changes in the quality of existing products and the introduction of new products. Real output or real incomes are determined by ``deflating'' nominal (money) output or income by an appropriate price index. However, these price indexes or ``deflators'' track the changes in the prices of a fairly constant basket of products. If the quality of the products are improving and if new products are constantly being used in society, the price index may miss much of the action. As The Economist (1996) points out, ``The measurement problem becomes more acute as output shifts to goods and services where quality is an important factor.'' The areas of medical services, information technologies, and financial services are good examples of this. While Statistics Canada does attempt to account for quality changes and new products in their calculations, they acknowledge that the price adjustments are complex and sometimes impossible to solve in a fully satisfactory manner. Because of this, current statistical measurement may understate improvements in the standard of living.

In a recent paper, William Nordhaus (1994) uses the example of lighting to illustrate the nature and extent of the understatement of true growth in the economy. He notes that improvements in the product we might refer to as ``illumination'' happened very infrequently and were relatively minor over the course of human history until about 200 years ago. From about 2000 BC, when controlled wood fires and shallow, bowl-like, fat-burning Paleolithic lamps were replaced by candles and closed, pottery and bronze lamps burning oil from olives and other plants, there was little advance in lighting technology until the Industrial Revolution. Rapid improvement came in the early nineteenth century with the use of coal gas and kerosine (distilled from petroleum) for illumination of streets and homes and, in the early twentieth century, with the introduction of the electric light bulb using a tungsten filament. The traditional indexes of light (based on the prices of given products such as lamps and fuel) shows that the real cost of light has increased about four-fold since 1800. The true cost of lighting, however, measured by examining the labour-time needed to buy a given unit of illumination, has declined precipitously. Nordhaus concludes that significant ``jumps'' in technology are simply missed by the standard indexes with the result that ``traditional price indexes dramatically overstate the true increase in prices ...'' (Nordhaus 1994: 29). This, in turn, means that the true improvement in the standard of living is significantly underestimated.

Nordhaus is somewhat pessimistic about the possibility of ever getting truly accurate picture of real output. He says: ``a complete reckoning of the impact of new and improved consumer goods on our living standards is at best an epic task and at worst infeasible'' (Nordhaus 1994: 2). I shall return to this issue at the end of this report in a discussion of the possibility of an index of living standards.

info@fraserinstitute.ca You can contact us at the above email address for any comments or information requests. Please report any dead links or technical problems.

If you know someone who would be interested in this web page, please enter their email address below, and we will forward this URL to them: Email Address:

Last Modified: Wednesday, October 20, 1999.