FEBRUARY 18, 1971
500 Million Questions
AN ADDRESS BY Lorne Rowebottom,
ASSISTANT DOMINION STATISTICIAN, OTTAWA
CHAIRMAN The President,
Harold V. Cranfield
GRACE Lt.-Col. W. Pedlar
I find it most interesting that the man responsible for births, marriages and deaths in Canada failed to provide me with the simplest of statistics in respect to himself. I don't know where he was born, or when, nor do I find any information that confirms or denies that he is married.
This secrecy is a fundamental to census. One may write the Bureau for census information about oneself but the Dominion Bureau of Statistics will not provide such information to any one other than the person who gave this information or his legal representative. So, if our speaker doesn't have these facts about himself, and I judge he did not, since he did not provide them for me, he must write himself a letter and if he considers he is an appropriate recipient of such facts he will release this information to himself. This is how he will know when he is entitled to old-age assistance for example. It gives one great confidence in the Dominion Bureau of Statistics to realize that personal information is safeguarded so completely.
From the record, then, he came into being as far as the Great Statistical Machine is concerned in 1943, for that is when he joined the bureau. Prior to that he probably existed somewhere but to my information in hand I can say with complete positiveness he has been a life-time student of statistics and how to gather them. Of course he came qualified to this work for he had his degree in Economics from the University of British Columbia and it would seem likely that he has his origins in British Columbia for that is where he went to work, armed with his "sheep skin". He joined the Bureau's Regional Office at British Columbia and spent three years there. The Dominion was then in the post-war years and you will recall the inflationary state of Canada at that time. The country needed a specialist in price statistics and he became that person. Living costs are the direct result of price changes and the methods of measuring these changes were developed by our speaker. His skill at this was recognized by his being appointed as the Director of the "Prices Division" early in the years that followed.
The growth of this area of government necessitated that Dominion Statistics functioned best as a separate department. So by 1966 our speaker became Assistant Dominion Statistician with his responsibilities extending into the academic and political spheres as well as retaining a continuing interest in vital statistics as they affect us all--births, deaths and marriages. Education, health and welfare, agriculture and consumer prices are all subject to his direction and scrutiny. Despite this long preparation it is a tremendous task he now undertakes, for it is the biggest peacetime operation this country has known! With great rapidity he must recruit and train a census "army" larger than the first three divisions of Canada's military army in World War II. As that was a national emergency they took only six months in 1939-40 to be mobilized. His army must be ready in half that time (three months) and be fully operational. In the light of this, our guest speaker is being properly introduced as the Field Marshal of our 1971 census battalions. His subject, as you have read it on the cards circulated to you in the mails, is "Five Hundred Million Questions". Gentlemen, here now is Mr. Lome Rowebottom, Assistant Dominion Statistician, Ottawa.
Mr. Chairman, it is an honour and a privilege to address the Empire Club on the subject of the most important taking of an inventory, ever to be performed in Canada--namely, the 1971 Census.
One of the first principles which teachers of public speaking emphasize, is never ask a rhetorical question, because if you do, someone is liable to give an answer you don't want. I would like to start by breaking the rule and running the risk. Which is more important, to put a man on the moon or take a census? If human beings had to decide whether to do one or the other, I think there would be no doubt that they would conclude that it would be more important to know how many people there are on this earth, where they are located, and something about what sort of people they are and what they do; than it would be for an unknown number of people to put one of their own kind on the moon. But, we need not seriously address ourselves to the alternatives since man has decided to do both. Actually, of course, that is not quite true. A group of people living on the North American Continent, between the 49th Parallel and the Rio Grande, have decided to put a man on the moon. But, men everywhere in all countries have decided they must know those things about themselves which can be learned only from censuses. Thus, the Census Office of India will hire over one million enumerators to take the next Census of that country.
As to the second question, which is more difficult, I posed the question not because I intend to answer it, but to suggest to you that the difficulties are of the similar order of magnitude, although vastly different in kind. Let me suggest a few of the most significant differences. First, with every landing on the moon it becomes progressively easier and if they become important enough, they will become routine. Not so with censuses, partly because they are taken so infrequently--we have only had the twelve in one hundred years of census-taking in Canada--partly because census objectives keep changing, and what was a good census, adequate to the needs of an earlier decade, would, measured by today's requirements, be a poor thing indeed. But the greatest difference relates to the differences in the number of people whose co-operation is required to produce a good census, and I would argue that it is more difficult to get a total of 500,000,000 correct answers provided by 21,000,000 people, than it is to put 1, 2 or 10 men on the moon for much longer periods of time than the few hours so far achieved.
Two other significant differences relate to resources available to do the job, which are quite different order and finally the fact that a moon-launching is more postponable than a population census. Ready or not we are legally obligated to take a Census this June 1, 1971 and we will do our best to have all systems GO, but there are no possibilities of "holding".
I should like first to put the census in a context of a complex of management information systems, which is the product of the Dominion Bureau of Statistics. DBS is itself a large complex organization with an annual budget of some $40,000,000 and over five hundred programmes of statistical production and analysis. You will all be familiar with some DBS statistics, but to illustrate the range of the products we produce, I would simply enumerate the following sample of subjects for which DBS produces an array of information prices, production, employment (and unemployment), international trade, transportation, retail sales, gross national product, productivity, international balance of payments, and income and expenditure of corporations and governments. But, in addition to such other economic aspects of Canada, DBS measures critical aspects of the demographic, economic and social conditions of the people of Canada. As you might expect, we start at the beginning, and produce statistics of birth, and then education and marriage; and people at work, and their incomes and how they spend them, and the illnesses they acquire through life and how they are treated and cared for in hospitals and welfare institutions; the crimes they commit, the jails they occupy, the universities they teach and learn in, and finally of course, we observe death and its multiple causes. But, the statistician is concerned with life not with death, and with the human, social and economic facts and conditions of life.
And the reason for all this is, of course, not to satisfy the human, or even the statistician's curiosity about the human condition, but because in our complex world, information has become essential to enable us to manage our affairs as well as we do, and it constitutes an essential basis for improvement in economic and social policy and programmes in both the public and private sectors. I am sure that it is a common place to members of this audience that no large business could long survive without knowing a good deal about its internal affairs and the markets in which it buys and sells. It may be less of a common place, but it is no less true that a country--a province--a city--cannot effectively or efficiently manage its affairs without comparable information--which very frequently exists in the form of statistics--or not at all.
And, of course, central to management of our affairs is knowledge of our basic resource--ourselves--we the people of Canada. The foundation of our knowledge--and in a good many instances the extent of it--about Canada's population is derived from censuses. For a census is the only occasion on which we acquire a comprehensive body of information about all the people, which makes it possible to establish comparisons and relationships between a large number of variables such as location, age, sex, occupation, income, education, mobility, marital status.
Before leaving this subject of Why a Census, and turning to ways and means, I would like to make the point that the statistician is the agent of those who want statistics. His actions mirror the needs of our society, and while he fashions the instruments of measurement, it is the user of his product that determines what he measures. Thus, in the months ahead if (perhaps when is the better word) you are annoyed at snooping nosey statisticians who ask so many questions, remember, if you will, he is acting on behalf of others who have a real "need to know". If he is to create an output of statistics, the statistician requires an input of data--of answers to questions--which many people find it annoying--to say the least--to answer. It occurs to me that the process of "knowing oneself" may be just as difficult for us collectively as a society, as it is for us individuals to follow Socrates dictum, "Know thyself".
THE 1971 CENSUS
The increasing complexities of modern-day problems of planning, administration, and research, have led to a substantial escalation in the demand for census information. More statistics than ever before required of one census will be expected from the 1971 Census. Regional development planning, urban renewal projects, education and manpower programmes, poverty and welfare assistance measures, and marketing research analysis, are some of the fields in which increasing needs of users have been articulated and evaluated during the planning stages for the 1971 Census.
The Census will, in effect, be a stop-action snapshot of Canadians on June 1, 1971. In many ways it will tell us how far we have come. It will also help us to plan where we want to go.
The original legal purpose related to parliamentary representation remains, but the Canadian census of today has far wider uses than to apportion electoral representation. Its importance hinges on its role as an inventory of the people--their numbers and local distribution, age and sex structure, language, ethnic and religious composition, educational attainment, occupational and industrial employment, income levels, housing and agricultural conditions. Of critical significance are the uses made of the census results in the development of plans and the formulation of social and economic policy by government departments and the business community.
Governments, business, universities, individuals--all of us--need good information to make good decisions. The census will provide that kind of information. Uses of census statistics are almost endless. This explains why Canadians have always co-operated so completely in census-taking.
Canadians will find that something is different when the Census of Canada is taken on June 1, 1971.
In the 1971 Census you, or someone in your family, will be your own census-taker. You can study the questionnaire and accompanying instructions on your own time, look up records where necessary, and check with other members of the family before setting down the answers. Or, if you need further help, you can get it on the phone, at a local number listed on the form.
If you live in one of the larger urban areas, you will be asked to mail back the completed questionnaire--in a postpaid envelope provided for this purpose. That will finish your task, unless the census officials find errors or omissions that may make it necessary for them to contact you.
In other areas, the census representative will call to pick up and review the questionnaire.
Thus it will be a self-census for most of us. Only in the remote outlying areas, or where special problems exist, will the traditional door-to-door canvass be made.
Sound simple? We hope it will be--for you.
For Canada as a whole, the task is gigantic. The 1971 Census will be one of the biggest peace-time operations the country has ever seen. It will require about 50,000 workers, specially-made electronic equipment to "read" and process millions of questionnaires, computers to organize and compile the information.
The cost of taking the census will be over $35,000,000. And it's one of the best investments Canadians can make. The pay-off--for everyone--is huge.
The History of the Census
The modern census originated in Canada! And we are still regarded abroad as a world leader in many aspects of census-taking.
But the idea of census-taking goes back almost 6,000 years.
The very word "census" comes from a Latin word, "censere", which originally meant "to assess". And in those ancient days, the word was apt, because the primary purpose was to collect taxes.
But in the last few centuries, all that has changed. Censuses today have nothing to do with assessment or taxation. A modern census is the complete listing or cataloguing of a people and many of the things that affect their lives.
Under the British North America Act of 1867, a census was to be taken in' 1871 and "every tenth year thereafter". This decennial census was extended to the territories and to new provinces as they joined Confederation.
In addition to the decennial census, a special agriculture and population census was taken in Manitoba in 1896 and extended to all three Prairie Provinces in 1906 to keep up with the rapid settlement of the West. This "mini" census continued every tenth year up until 1956 when it was extended to the whole country. It has proved invaluable in updating population and agriculture statistics between major census periods and plans indicate it will be continued in 1976.
Three Censuses in One
The 1971 Census will thus mark 100 years of national census-taking in Canada. And it will really be a combination of three censuses--a combined population and housing census; an agriculture census; a merchandising census.
The population and housing census is designed to tell us how many people live in Canada, where they live and the kinds of homes they live in.
Every household will receive a questionnaire. Two out of three will get a short form, requesting basic population data. They will be asked to record the name, birth date, relationship, sex, marital status and mother tongue of each household member and to answer nine housing questions.
A longer questionnaire will go to every third household. In addition to the basic questions, it will ask for further information about housing and such matters as education, employment, migration and income.
This is known as "sampling"--a technique which enables accurate conclusions about a whole society to be drawn from the replies of only some of its members. The degree to which sampling is employed is directly related to the reduction of costs and to the production of more timely results.
The self-census and the sampling method will apply to about 97 per cent of the population. The only exceptions will be those living in remote parts of Canada, where geographical and other problems make it necessary to use traditional methods of door-to-door enumeration. All households in these localities will be asked the full range of census questions.
Census of Agriculture
The farmer has a special responsibility in the census. He is asked not only to account for all the members of his household but also for his livestock, machinery, and other aspects of his business.
It is a big and important job for the farmer. Canadian agriculture has been undergoing rapid change during the past 20 years. And this year farm problems are more urgent and complex than ever.
Census of Merchandising
Census representatives will compile a list of businesses in the retail, wholesale and service trades. Early the following year, each business will be mailed a detailed questionnaire for completion and return. The information required is the kind readily available in business records covering financial activities in 1971. When the data has been compiled and analyzed, it will provide a definite picture of the value and trends of Canada's channel of distribution, from wholesaler to the final user--the Canadian consumer.
The intent and purpose of some census questions occasionally baffle people and cause comment. But there's a good reason for every one.
A favourite eyebrow-raiser is bound to be Question H9 on the 1971 questionnaire. It asks: "At what telephone number can this household be reached?"
Obviously, this is not a statistical question. But, with self-census procedures, where householders complete their own questionnaires and, in urban areas, mail them back to the census office, it sometimes becomes necessary during processing to contact the householder to obtain missing information or clarify answers on the census form. A follow-up telephone call is quick, economical and more convenient to the householder than a personal visit. The telephone number will be retained as part of census information.
There are also questions as to whether plumbing facilities are used by one or more households. This information is important to housing authorities. It helps to indicate the standard of housing and thus assist in identifying areas that require urban renewal or redevelopment.
Census information is so useful that interested individuals and organizations constantly suggest questions that they believe should be included.
There are conflicting pressures on the 1971 Census to keep the list of items on the Population and Housing Questionnaire and on the Agriculture Questionnaire short enough that the operation is feasible in terms of quality, timeliness and cost, and yet still meet as many vital data needs as possible.
Only questions that will yield information useful to many Canadians are included.
Long before each census is taken, working groups and committees intensively discuss the questions to be recommended for inclusion. Representations are made by federal and provincial government agencies, businesses, universities, town planning experts, financial institutions and many others. The final selection, which must be submitted to the Cabinet for ratification and approval, is made on the basis of the usefulness of a question, the cost involved in asking it, the relative difficulty of getting reliable answers, and the amount of effort it will take for the householder to provide the information.
Because of the recent emphasis on manpower training programmes, new questions will be asked in 1971 about the time people devote to vocational and occupational courses. Town planners have long needed statistics relating place of work to place of residence, so this time the address of the place of work will be requested.
Of course, answers to 500 million questions will be input to computer by complicated high speed automatic optical reading devices and the output of the computer will be statistics in the form of magnetic tapes and computer printouts which will be available themselves and which will form the basis of census publications. The computer technology available for processing the 71 census is vastly in excess of the capability available in 1961 or even 1966 and we will be using computers to make census results available quickly, accurately and cheaply. Results will be in a variety of formats designed to serve the needs simple counts of population on paper, through to complicated five way cross classifications on computer tape. In 1961, very few users of census results had computers available to them, whereas, throughout the 1970's large numbers of users will have high power computer processing capability available and we anticipate that computer tape will be a routine form of releasing 1971 census statistics.
I have reserved till the last, the critically important fact that the whole edifice that I have been talking about has as its foundation the good will and co-operation of the Canadian population, which we believe to be warranted because: first, good census statistics will make a major contribution to good government and good private decisions second, the government is as dependent upon individual Canadians for essential information as it is dependent on them for tax dollars. There is no other source, and the citizen's contribution via his answers to census questions is comparable to the contribution he makes via his tax dollars third, the answers he provides are secret and confidential to the Dominion Bureau of Statistics under penalty of law and in accord with 50 years of unblemished administration of the secrecy provisions of the Statistics Act. The information a citizen provides may never be identified as coming from him and the census files of DBS are not available to any other department or agency of any level of government. Answers to census questions are not sent "to Ottawa". They are sent only to DBS which is obliged under law, to not make available to any other agency except with the written permission of the citizen himself. If I may make a comparison, your answers to census questions are no more distinguishable in any of the products of DBS, or otherwise identifiable, than are the tax dollars you pay. Finally, there is a legal obligation to answer census questions. But that legal obligation will rest lightly on Canadians who have through many censuses demonstrated their willingness to co-operate with DBS, which hardly ever enforces its legal authority, but which builds its statistics on the co-operative understanding of the Canadian people.
We have been testing the 1971 census questions and methods through a series of tests which started some five years ago. These comprehensive tests and a dress rehearsal, involving a total of nearly half a million Canadians, demonstrated that the great majority of Canadian householders would accurately complete the census questionnaires left with them, and mail them back or hold them for an enumerator to call as we request.
I am confident that this coming June, Canadian households and families will make their essential contribution to making available good information for good decisions for a better Canada that will be found the results of the 1971 census--by answering 500 million questions.
Canada counts, count yourself in.
The gratitude of the Club was expressed by Mr. Robert L. Armstrong.