Chapter 4: A stronger agency: 1985 to 1995

Table of Contents

• The passing of the reins
  • The shaping of a man who beat the odds
• Three key cultural changes
  • Investing in people
  • A shrinking public sector
  • A strong and sustained focus on research and development
  • The Data Liberation Initiative
  • Growth in international stature
  • National and international recognition
• Organizational changes
  • Regional offices restructure
  • The National Statistics Council as a guidance mechanism
• Management initiatives
  • Open lines of communication
  • Best employee newspaper
  • University recruitment to attract employees
  • Training
  • Butting out
• Integrated and centralized planning
  • Automated time reporting
• Communications and marketing
  • Inside StatsCan
  • Herby and Elliott ride the halls
  • The Internet emerges as a new tool
  • A stronger corporate identity
• Notable milestones in the statistical system
  • The introduction of longitudinal surveys
  • Acceleration in social statistics programs
  • Social support
  • The General Social Survey: Game changer
  • The evolution of a Canadian health information system
  • Celebrating 50 years
  • New health information initiatives
  • A new national education statistics program
  • Adult literacy around the world
  • Re-examination of Confederation
  • Defending the agency's political neutrality
  • Businesses: Small and large
  • Statistics Canada switches from gross national product to gross domestic product as a prime measure of output
  • Fixed release dates for economic surveys
  • International trade statistics initiatives
  • Free trade
  • Measuring the services sector
  • Remote sensing continues to gain ground
  • New generalized systems
  • Quality guidelines
• Census programs
  • Mr. Census
  • Statistics on Aboriginal peoples
  • Integrating tourism
  • Address Register and Business Register
  • A giant demographic step forward
  • Environment
  • Recycling comes to the agency

The passing of the reins

In 1983, Dr. Wilk arranged for Dr. Fellegi to be promoted to the newly created position of Deputy Chief Statistician in 1984. Not only did he make it obvious that there was a clear candidate for succession, but also the arrangement allowed for a period of understudy. Truth be told, Dr. Fellegi was likely already a prime candidate as, five years earlier, at the time of the search for a new Chief Statistician, the selection board had approached him for an interview, but he had indicated that he did not feel the time was quite right.

In 1985, Dr. Wilk wrote to then-Prime Minister Brian Mulroney to indicate that he wished to retire. Later that summer, Dr. Fellegi was in Bruges, Belgium, with his family, and received a message to call the Clerk of the Privy Council, Gordon Osbaldeston. Mr. Osbaldeston informed him that the Prime Minister wished to know whether he was ready to accept the job of Chief Statistician of Canada. Dr. Fellegi accepted and became Chief Statistician of Canada on September 1, 1985, having already worked at the agency for almost 30 years.

The shaping of a man who beat the odds

Dr. Ivan Peter Fellegi was born in Szeged, Hungary, in 1935. He had an early start at mathematics, already knowing how to count to 1,000 by the time he was four years old. He had had a brain infection at 18 months, and his mother had been advised that, if he survived, he would be seriously impaired for the rest of his life. His family was obviously quite vigilant in the aftermath of such a diagnosis, and taught him by the age of four not only to count, but also to know all the European capitals, as well as to recite poetry. When Germany invaded Hungary, in March 1944, his father was taken to a concentration camp in Austria. The family was later also deported. They returned to Hungary in 1945, where four years later the family business (a stonework factory) was nationalized by the communist regime. Because of his "bourgeois" family origins, Dr. Fellegi was not allowed to go to high school in his home town and so left home at the age of 14 to enroll in high school in the capital of Budapest, where his origins would not be known. The same class-based wall was thrown up when it came time to enter university. However, there was an alternative: exceptional students were allowed university admission irrespective of their family background. All students placing in the top five in the country in national competitions were automatically admitted to university in the field in which they had competed. A poet at heart, Dr. Fellegi was encouraged by his professors to take this competition in Hungarian literature. Instead, he decided that the safest route would be to take his examination in mathematics, with the logic that mathematical answers could not be disputed, while literature was a good deal more subjective. He was successful, and was admitted to the mathematics faculty of Eötvös Loránd University (the University of Budapest).

He had just started his fourth year of studying mathematics, in 1956, when the Hungarian revolution broke out. His father convinced him to leave the country as his sister had done a few years earlier. However, by this time he would never have been legally allowed to do so, so his mother cleverly obtained a medical certificate indicating that he had tuberculosis, and he and a cousin were sent to a sanatorium conveniently located near the Austrian border. They arrived at the border town too late in the day to be admitted to the sanatorium and checked into a local hotel. They set out on foot later that night. It was not a simple trek—they stumbled upon a woodsman who took one look at them, and said that he did not want to know who they were or where they were going but that, if they were interested, there were Russians "that way" and there were no Russians "this way." They did not know when they crossed the border. They just knew that, as soon as they heard voices, their fate would be decided by the language they would hear spoken. When eventually they heard those fateful voices, the language was… German, and they were picked up by an Austrian border patrol.

Dr. Fellegi was reunited with his fiancée, Marika Gulyas, in Vienna, and together they left for Canada, arriving in the dead of winter in January 1957. Their first Canadian address was a prison in Montréal, which at the time was a temporary shelter for hundreds of Hungarian refugees. He later moved in with his sister who was studying medicine in Ottawa. She encouraged him to continue his studies and, within a few weeks, he was enrolled to enter second year of pre-medical studies in the fall at the University of Ottawa.

In the meantime, he sought employment and approached the Public Service Commission. As luck would have it, the public service was short of mathematicians, and as a result had relaxed the rule that only Canadian citizens could be hired by the government. By February, he had been hired by the Dominion Bureau of Statistics as a technical assistant. He soon realized a profound love for statistics, and approached Carleton University to ask whether he could pursue a master's program there in statistical studies. Although the university did not yet have a master's program, it so happened that it was planning on setting one up. The university indicated that it would admit him if he could pass a comprehensive oral examination in mathematics to prove that his knowledge was equivalent to that acquired by a BSc at the university. He was successful and began taking night courses to receive his Master of Science by 1958, becoming the first Carleton University student to earn the honour. He wanted to pursue his PhD in statistics as well, but there was no professor of statistics at Carleton. The mathematics faculty arranged that a professor from Toronto—the esteemed Dr. Donald Alexander Stuart Fraser—would review his thesis when it was finished. Dr. Fellegi riskily pursued his thesis without having an advisor to guide him or even to initially approve his subject. The acknowledgement on his thesis, Sampling without Replacement with Probabilities Proportional to Size, indicates the pertinence of the subject to the Dominion Bureau of Statistics and that it arose in connection with the Labour Force Survey. In 1961, after his thesis was passed by Dr. Fraser and after two and a half hours of oral examination, which the mathematics faculty of the University of Ottawa also participated in administering, he became Carleton University's first PhD graduate.

At the Dominion Bureau of Statistics, he was appointed Chief of Sampling Research and Consultation in 1962, Director in 1965, Director General of Methodology and Systems in 1971, and Assistant Chief Statistician of the Statistical Services Field in 1973. He took a leave of absence from Statistics Canada from May 1978 to March 1979 to serve on President Carter's Commission on the Reorganization of the U.S. Statistical System. He returned to the agency as Assistant Chief Statistician of the Social Statistics Field in 1979, and in January 1984 was appointed as the agency's Deputy Chief Statistician. He served as Chief Statistician of Canada for 23 years, from 1985 to 2008. Upon his retirement, he had worked at the agency for 51 years. He was appointed Chief Statistician of Canada Emeritus in 2008 by Prime Minister Stephen Harper, and continues to return to his office twice a week to make himself available to the agency's employees for professional advice and personal counsel. He has served on the National Statistics Council for 31 years, starting as Chief Statistician and then as a member, since its first meeting in 1986. At the time of this publication, he has served the country's statistical system, and by extension his country, for 61 years.

In addition, Dr. Fellegi has served as President of the Statistical Society of Canada, the International Statistical Institute, and the International Association of Survey Statisticians. He holds six honorary doctorates and is the recipient of numerous medals, citations, and titles, including Fellow of the American Statistical Association, Honorary Member of the International Statistical Institute, Honorary Fellow of the Royal Statistical Society, and gold medalist of the Statistical Society of Canada. He was also awarded the Order of Canada in October 1992 by then-Governor General Ramon Hnatyshyn and was promoted to Officer in 1998. He received the Prime Minister's Outstanding Achievement Award and the Career Achievement Award in 2002. In 2004, he was awarded the Order of Merit of the Hungarian Republic.

Three key cultural changes

While the agency had begun its transformation under the wing of Dr. Wilk, it was under Dr. Fellegi's leadership that many of the initiatives that had been introduced in the first half of the 1980s were strengthened and solidified into the working culture of the agency. Three key elements that were hallmarks of this time period were a continuing and deliberate investment in people, strengthened research and analysis with a focus on sound methodology and collaboration with academia, and a greatly enhanced international reputation.

Investing in people

Dr. Fellegi wholeheartedly agreed with Dr. Wilk's thinking that large-scale reorganizations tended to be disruptive and often counterproductive. He was a proponent of organizational stability yet considerable individual mobility, to develop people and their skills, but also sought the flexibility to redeploy people as needed in the case of changing priorities. These were not easy years during which to lead a government agency as, year after year, relentless cutbacks were whittling away at budgets.

In 1979, when he returned from his leave of absence in the United States, Dr. Fellegi's timing coincided with a major budget cut, and he found himself having to notify some of the staff who were to be laid off. This difficult experience had a profound and lasting impact on him, and he vowed that he would do his best to avoid ever having to do such a thing again. Fast-forward six years to when he was appointed Chief Statistician—this too was on the heels of a large budget cut. His first memo to all staff proposed that, if their jobs were at risk, but they were willing to be flexible, they would be redeployed into different jobs that had become vacant by attrition. They would also be offered the appropriate training and support required. Thus, while the specific work employees were doing may have been terminated, they would not be out of a job. This came to be known across the agency as Dr. Fellegi's "No-layoff Policy." It went a long way in contributing to a feeling of ease across the agency as it faced budget cuts, recessions, pay freezes, and a public service strike. It also made the agency more adaptable to change. Gone were the days when people joined the organization and retired 35 years later from the same area.

As would be the norm for years to come, the need for data was growing exponentially, but the budget for providing such data was shrinking from year to year. A five-year program was implemented in 1985/1986 to reduce spending, as the agency was required to cut annual costs by almost $26 million. With the no-layoff policy in place, although 586 positions were cut, the majority were dealt with through normal attrition. Affected permanent employees were redeployed to other positions, many with the opportunity to upgrade their skills and take on more challenging jobs. Major savings were also being realized in several areas, one of the most significant changes being the establishment of the Operations and Integration Division to centralize data collection operations in 1985, as noted in the previous chapter. The integration and standardization of operations improved efficiency and saved over 175 person-years by centralizing the support staff that mailed out, received, checked and coded the surveys for the agency and by making use of new technologies and improved methods. The new division afforded greater flexibility to the agency, not only because it had the capability to handle variable and peak work periods, but also because its capacity was not dedicated to any one product—such that processing plans for one survey could be substituted for those for another survey at short notice. There would be many other initiatives in the 1980s and 1990s aimed at changing the culture at the agency to encourage corporate-level thinking, but this was the first major one. It bears noting that, as with any major change in a large organization, this was not an easy shift. It was time-consuming and laborious, and was resisted by many line managers who had operated in their own way for decades. Clerical staff were reticent to leave their home divisions to join the new division and their managers did not want to lose them. Logistically establishing the new approach without disrupting ongoing programs was also quite the endeavour. In the end, while the path was difficult, the transition was ultimately successful and created one of the most important functions at the agency.

The other important element to the agency's flexibility and responsiveness in face of unremitting budget cuts was its popular corporate assignments program introduced under Dr. Wilk. The program was flourishing, with about 10% of all employees on a corporate assignment at any time. The program was so useful that 24 other government departments introduced similar programs. The program had been created as a response to the 1981-to-1982 recession and related wage controls, when the agency was coping with severe fiscal restraint. While programs were cut back, there continued to be demand for new information, and so having a pool of employees who were willing to take on new work was invaluable to the agency as well as to those employees who would not lose their jobs. The program felt safe for both the employee and the employer as they had a guarantee that, if an assignment did not work out, employees could return to their previous jobs. In a period of low turnover and scarce promotions, lateral movement was also "the only game in town" for many desiring a change.

In a 1988 interview with SCAN, the employee newsletter, following a budget reduction exercise in which no jobs or programs were terminated, Dr. Fellegi stated, "Our no lay-off policy is not only humane, but I think it contributes to a more productive environment. If people feel reasonably secure in their jobs, they are more ready to take risks—and we want wise risk-takin."

In the midst of the 1990-to-1992 recession, and again facing massive budget reductions, Dr. Fellegi had three principles he wished the agency to follow. First, that everything could not be equally reduced—that some programs would need to be cut in their entirety. Second, that employees would not be laid off, and should feel secure in their jobs, but they should co-operate in redeployment to other jobs. Third, that a budget should be preserved for innovation and improvement. He felt that it was critical to sustain an innovative atmosphere in a period of declining resources, as programs could be re-instated relatively quickly, but that it was much harder to rekindle lost innovation. In fact, an emphasis on analysis and on research and development, even in the midst of successive budget cuts, would be one of the hallmarks of the Fellegi years at the agency.

A strong and sustained focus on research and development

A new annual program of internal research sabbaticals was instigated, whereby mid-level employees could submit research proposals to undertake analytical work or research and development of special interest to Statistics Canada. Up to six employees annually whose proposals were accepted by a panel of experts would be given the opportunity to pursue that work and be relieved of their current duties for up to a year with full salary and benefits. At the end of their sabbaticals, they would return to their jobs, along with enhanced research and analytical skills, and the publishable results of their research. The completed works were refereed and published either as research monographs or in one of the agency's flagship journals, such as the Canadian Economic Observer, Survey Methodology, Perspectives on Labour and Income and Canadian Social Trends. Notably, these were also the years where institutional and peer reviews of analytical pieces became fundamental to the agency's research program.

A lecture series was established at the agency in the late 1980s, covering a number of diverse topics. The lecture series was initially implemented as the vehicle chosen to exemplify the spirit of Celebration 88, a federal program highlighting various achievements in the lead-up to the Olympics being held in Calgary in 1988. Each year, about 15 lectures were presented, some by experts from outside the agency and some by staff members. For example, leading demographer Ansley Coale spoke about population trends in China since 1950, and Professor Anthony Richmond from York University discussed the importance of basic research to public policy in his lecture on demographic research and public policy. Other well-known individuals who gave lectures at the agency included Nathan Keyfitz, Richard Ruggles, and Wassily Leontief (winner of the Nobel prize for economics in 1973). A special lecture on the history of the agency was given in 1988 by recently retired Assistant Chief Statistician David Worton, who was devoting his time to researching and writing a history of the agency, which would be published in 1998.

A renewed focus on analysis was also manifested in changes to the agency's releases in The Daily. The evolution toward publishing more analysis in The Daily was a collaborative corporate effort by managers and analysts from virtually every division. A senior editorial board was established in 1994 to improve the analytical quality of The Daily and transform it into a more readable product by emphasizing the storyline. Dr. Fellegi chaired the Board, which was composed of senior managers and analysts from across the agency. At each meeting, board members were assigned upcoming major releases to review. A few weeks later, the board would meet with analysts to discuss their recommendations. The board met weekly for about eight months. This allowed the board to define what constituted an effective release and helped author divisions redevelop their releases. At the end of the year, new guidelines were put in place for writing an effective release, and a course was offered on writing for The Daily. The course was also very much a success, with notable improvement in releases remarked upon by many regular subscribers.

Growth in international stature

During this time, a great deal of emphasis was placed on the agency's collaboration and cooperation at the international level. This was evidenced not only by the growing stature of the agency internationally, but also by the number of employees elected as members of prestigious international organizations, such as the International Statistical Institute.

A new program was implemented in 1988 to provide technical assistance on a cost-recovery basis to other countries, with a view to supporting Canada's development co-operation program and enhancing international data comparability.

Dr. Fellegi attached a great deal of importance to international collaboration, and would summarize this emphasis as being important for three reasons. First and foremost, it was important for purposes of learning. As a centralized statistical agency, Statistics Canada had no true peer within Canada—so for its employees to be exposed to knowledgeable discussion or critique to inform their professional development, they needed to work with other statistical organizations around the world. The second reason was to broaden staff's perspective by encouraging them to benchmark themselves to others. The third reason was that he wanted the agency to be prominent internationally as he felt that international recognition was crucial to positive perception here in Canada. At the time, international reputation was of great importance to the country.

The other side to the coin was that he felt it was the agency's duty to share the knowledge and experience it gained: not only to collaborate and offer assistance, but also to solidify the agency's knowledge and confidence. In explaining that old adage that in order to best understand a subject one should try to teach it, he would say to SCAN that "When you have to give a paper to an international body, you had better know what you are talking about. I find our international work an incredibly useful stimulus to clarify our own thinking."

Canada was elected to the United Nations Statistical Commission, the highest decision-making body for international statistical activities, in 1988. The Commission consisted of chief statisticians from member states from around the world, and oversees the work of the United Nations Statistics Division. The 24 member countries are elected by the United Nations Economic and Social Council on the basis of equitable geographical distribution, with a term of office of 4 years. Canada was chair of the Commission for the first three sessions, in 1947 and 1948. It has been elected to the commission six more times as of the publication of this piece (1951-to-1959, 1962-to-1969, 1974-to-1981, 1989-to-1992, 2006-to-2009, and 2018-to-2021).

Technical assistance was also a priority for the Canadian government. In response to a request from the United Nations Statistical Office, a six-week course on census sampling was held in Ottawa in 1988, attended by delegates from about 10 sub-Saharan African countries. The agency also conducted a mission to Uganda at the request of the Canadian International Development Agency and the World Bank to assess the possibility of Statistics Canada's involvement in the rehabilitation of their statistics departments. Technical assistance was also provided to Indonesia in the area of environment statistics, and short-term training was given at the agency for statisticians from Indonesia, Egypt and Malaysia.

Starting in 1989/1990, Statistics Canada participated in a major technical assistance program spearheaded by the World Bank, with funding from CIDA, to gradually develop a database to track the social impacts of structural economic adjustments in sub-Saharan Africa. The agency helped design questionnaires, conducted training programs, and got survey programs off the ground.

In 1989, the Department of External Affairs suggested and funded a mission to provide technical assistance on statistical matters to central eastern European countries. A delegation was sent to Hungary to assess that country's most urgent statistical needs, and seminars were prepared on topics such as the legal foundations of a statistical system, the management of statistical offices, ensuring the relevance of statistical products, and respondent relations. For the first time, in 1990/1991, assistance was provided to Eastern Europe, and the agency advised Hungarian officials on the management policies and the legal basis of a national statistical system, as well as on the construction of a register of businesses. Assistance was also provided in setting up a monthly retail trade survey, which would become the country's first modern, sample-based survey.

Other countries receiving technical assistance included the former Czechoslovakia, Senegal, Kenya, Pakistan, Sri Lanka, and Uganda. Barbara Clift, who had played a fundamental role in designing and creating Canada's income and expenditure accounts, spent six months in Uganda in 1989/1990, single-handedly developing a set of production accounts for the country.

In 1991, the agency also participated in the first of what was to be a series of meetings aimed at co-ordinating the World Statistical System. This was a trilateral initiative involving Eurostat, the European Community's Statistical Office, and the American and Canadian national statistical offices to collaborate in helping Eastern European countries adapt their statistical systems to a market economy. The agency also hosted the Economic Commission for Europe seminar in 1990 on the implications of the changing population age structure in industrialized countries.

A bilateral exchange program was initiated with France, Australia and the United Kingdom to develop closer relationships and exchange knowledge. A member ofFrance's Institut national de la statistique et des études économiques came to the agency in 1991 to serve as director of the agency's prices division, while Statistics Canada undertook assignment there. Statistics Canada managers also undertook assignments in Australia and the United Kingdom.

Organizational changes

The following individuals comprised the team of assistant chief statisticians at the end of 1985: Bruce Petrie, Social, Institutions and Labour Statistics; Gordon Brackstone, Informatics and Methodology; Guy Labossière, Management Services; David Worton, Special Assignments; Yvon Fortin, Communications and Operations; Jacob Ryten, Business and Trade Statistics; and Stewart Wells, National Accounts and Analytical Studies.

In October 1985, it was announced that Mr. Harry John Hodder would fill the role of Deputy Chief Statistician. Mr. Hodder had begun his career with the federal government at the Privy Council in 1954, and had served in a number of capacities with the Canadian International Development Agency, Manpower and Immigration, and the Inter-American Development Bank. He had come to Statistics Canada in November 1984 as Assistant Chief Statistician of Business and Trade Statistics. He unfortunately passed away quite suddenly in February 1986. The position of Deputy Chief Statistician remained vacant for a year, until Dr. Fellegi felt that it was no longer required. To this day, there has not been another Deputy Chief Statistician.

Regional offices restructure

In the late 1980s, the regional offices were handling about 400,000 telephone inquiries each year, and carrying out about $1.5 million in custom work for data users. The data collection function was also being increasingly regionalized, as this was found to improve response rates and timeliness. As a result, the sizes of survey samples could be decreased, reducing the number of businesses and organizations needing to participate in surveys. From 1984 to 1989, 39 surveys were regionalized. By 1987/1988, the regional offices were responsible for conducting 100 surveys (up from about 75 in 1985).

The number of regional offices expanded to 10 in 1987 with the addition of the Calgary regional office. In 1988, a study was conducted to assess opportunities to modernize the collection operations of the regional offices, initiated as a result of obsolete mini-computers, the recent successful use of computer-assisted telephone interviewing (CATI), and a growing need to respond quickly to information requests. Shortly thereafter, 1989 would mark the advent of the first collective agreement for interviewers and senior interviewers working in the regional offices (about 1,200 interviewers). Until 1987, the regions were using hand-posted ledgers for their accounting, and forwarding them to Ottawa each month. The Financial Services Division at headquarters had been using an automated accounting system called FINCON for fiscal transactions since 1982. In 1987, this automated system was extended to all the regions.

CATI was first tested in Halifax in March 1986 and, after a few modifications, was implemented across the country in all regional offices. It was a new and exciting development for the agency, as it bypassed all manual processes in data collection, including recording, preliminary editing, and coding. The interviewers were retrained and quickly adapted to the new technology, even though, for many, this was their first time working on computers. Each trainee received a self-study package one week before classroom training, which took place over three days.

Entering the data directly into computers instead of onto a paper form allowed for online edits as soon as the data were entered, prompting immediate resolution of any issues with the respondent. It also allowed for sample file maintenance: the interviewers could update company names, addresses or contacts directly to keep the mailing list clean. The response rates increased by 20% and the per-unit cost of collection dropped by about 27% for the Annual Survey of Manufactures, which was used as a CATI pilot for economic surveys. The CATI software used came from the U.S. Bureau of the Census, and had already been successfully used by the Labour Force Survey as the first CATI pilot for a social survey, in March 1993.

By 1993, interviewers were also testing handheld computers in the field. This was the introduction of computer-assisted personal interviewing (CAPI), which would turn out to be a major challenge. Apparently anything that could go wrong at the outset did, from battery life issues, frozen screens or screens that were hard to read in bright light to painstakingly slow processing speeds. Tenacity paid off, however, as eventually, hand-held computers for the Labour Force Survey would save $1.7 million per year.

The Collection Operations Branch had also recently implemented a new service called "Fast Track Surveys" for some of the survey work that was new to the agency or to the branch. The development, collection and processing for these particular surveys, of which there were about 10 per year, ranged from a few weeks to a maximum of 6 months.

Management initiatives

From the rigid compartmentalization of the agency at the end of the 1970s, the agency was moving into an era where management initiatives were increasingly owned by all employees— they were becoming entrenched in the culture of the organization. These included corporate assignments, open lines of communication, employee opinion surveys, university recruitment, mentoring, and a corporate training program.

Open lines of communication

Dr. Fellegi began a practice of inviting yearly interviews with the employee newspaper, SCAN. These yearly check-ins were published as a special issue and covered much of the accomplishments of the agency over the year, as well as the context in which the agency was operating. State of the Union addresses were also instituted as his opportunity to communicate with all staff members (through his address to senior managers in the Simon Goldberg conference room and the subsequent publication of the material in the employee magazine—SCAN and, later, @StatCan), and relay not only the major accomplishments of the previous year, but also the upcoming challenges to meet for the next.

Statistics Canada's Policy Committee, which had been convened by Dr. Wilk, met weekly. The consensus approach to decision making (as opposed to one-on-one decisions) with all assistant chief statisticians present ensured that all variables from across the agency were taken into account. It was also another way to keep channels of communication open, as all assistant chief statisticians would brief their own managers the next day to ensure that all of senior management was aware not only of the decisions made, but also of the reasons behind them.

Communication was also bolstered through the implementation of employee opinion surveys. These were conducted every three years with anonymous participation to ensure all opinions could be voiced. Directors were required to report and discuss the results from their divisions with their employees, to compare them with those for other areas of the agency, and to establish concrete goals for improvement by the next round of the survey.  In fact, an electronic information network was first proposed at the agency after the 1991 employee opinion survey showed a widespread desire for better internal communications and easier access to information. The Internal Communications Network became operational at headquarters in April 1994 and in the regions that October, and quickly became a pillar in the agency's internal communications program.

Two other initiatives born of the employee opinion surveys were targeted at the career aspirations of employees. The first corporate initiative was skip-level interviews, which were introduced in 1993. These allowed employees to discuss medium-term career aspirations and related training needs with their supervisor's supervisor. The thought was that, while direct supervisors might have a conflict of interest as they may be reticent to lose their best people to career opportunities elsewhere, skip-level supervisors would be more likely to have a broader perspective and a more objective view. They would also be able to advise employees about opportunities elsewhere in the agency as well as provide guidance on longer-term career issues. The second corporate initiative stemming from the surveys was to allow any employee who had been in the same position for four years or more the opportunity to apply for a corporate assignment without their supervisor's permission. Again, this was to balance career advancement with operational needs.

In 1991, Treasury Board set up an interdepartmental working group to identify the best career development approaches, as part of a search for career development models for the public service. After reviewing over 60 private- and public-sector organizations, they chose three—two from the private sector (IBM and Chevron) and one from the public sector (Statistics Canada).

University recruitment to attract employees

At this time, the agency established a corporate university recruitment campaign to attract new employees. While university graduates were hired previously, it had been with little corporate support. Previously, a manager would choose candidates from a Public Service Commission list, invite them to an interview, select successful candidates, and hire them directly. The new corporate program would see the recruits hired by a team to work more generally for Statistics Canada rather than for a particular manager. The Economics, Sociology and Statistics stream (ES) recruitment program was launched in the summer of 1989. All entry-level employees in the ES stream were recruited by a single team of directors that combed universities seeking out the best candidates from across the country. The team would try to hire the best graduating students in December and January, when they were looking for jobs in the largest numbers. For the first 24 months, recruits were given training and a series of corporate assignments engineered to expose them to various facets of work at the agency. They were also assigned mentors to help them acquire a broad corporate perspective.

There was also a drive to increase the amount of part-time work done at the agency, not only because it was felt to provide value for money, but also to foster balance with other life priorities. All directors were urged to increase opportunities for part-time work in all occupational groups in their divisions.

Training

The Survey Skills Development Course was introduced as a pilot project in 1990. A similar course had been offered by the U.S. Census Bureau, and a group of employees (ES recruits) from Statistics Canada had been sent there to take the course in order to investigate its applicability to the Canadian context. Hank Hoffman adapted the course to the Canadian context and to the approaches used at Statistics Canada. The first course took place in 1994, a year after the creation of the Statistics Canada Training Institute. The six-week course for new employees also had a slightly modified version targeted at existing employees. A Survey Support Certificate program was also offered, which provided 3 weeks of training for 250 survey support staff each year. Mr. Hoffman was honoured as the agency's employee of the year in 1993/1994 for his outstanding contributions to the training and career development of agency employees.

A pilot program in 1991 involved the identification of up to four students identified by Carleton University to work on research projects of interest to Statistics Canada. The students were each supervised by an agency employee. A year later, agreements were put in place with Carleton University and the University of Ottawa to offer certificate programs in economics and demographic studies to agency employees. In exchange for guaranteeing a certain enrolment and having Statistics Canada staff teach parts of the course, the universities adjusted some courses to meet the agency's needs. In the regions, several pilot projects were also initiated with local universities, to see what kinds of cooperative programs could be implemented. Although the certificate programs had their place, they were ultimately not as successful as other training initiatives.

Integrated and centralized planning

The regular planning process that had been established under Dr. Wilk continued, and each year the agency would systematically assess its user needs and the budgetary parameters within which it operated. Priorities could therefore be established for the coming years, and resource needs anticipated. Senior management also identified resource savings of at least 1% per year to liberate funds for other needed investments, such as infrastructure investments, new projects, or unexpected budgetary requirements. Each division was expected to become permanently more efficient by 1% each year. The fields could request up-front investments for initiatives, which might help generate such savings, but the investments needed to be "paid back" in three years or less through ongoing savings. The new fund was referred to as the Strategic Investment Opportunity initiative, and divisions were encouraged to make proposals to draw from it. The fund helped the agency expand its analytic capacity, kick-start the renovation of the Business Register, and modernize and streamline the agency's collection activities.

One of the new fundamental culture changes in 1991 at the agency was the introduction of internal cost recovery for many services. For example, the main computer centre would start the year with no budget. Its costs would be covered by the purchase of services by user programs during the year. The aim was to give all program managers financial control over all the resources used by their individual programs, including those needed for collection, mailing, head office operations, telephones, and computing. This made the full costs of programs visible, and allowed program managers, over time, to make the most efficient use of these resources within their total budgets.

At the same time this process would provide service areas with real feedback on service demand so that they could adjust their mix and levels of service provided from one year to the next. In this way, for example, capital costs associated with the agency's computing infrastructure were self-financed through computing costs within program budgets and did not require separate external budgetary submissions. Some managers saw internal cost recovery as a mixed blessing. While it helped them familiarize themselves with the costs of all aspects of their programs, it tied their hands a bit in terms of trading off one cost against another.

Communications and marketing

After consultations with journalists, the two major frontline publications were reviewed and redesigned in 1986: The Daily and the weekly Infomat. The revised publications would better present information to make it more understandable and more user-friendly: they were said to be "humanized." The Daily was modified to highlight the most significant series and place greater emphasis on analysis. Infomat was modified to better display graphics and to include brief analyses of the major social and economic releases of the week.

A new magazine-style quarterly publication on social issues, Canadian Social Trends, was also released in 1986. This compendium presented articles on social issues from a statistical perspective along with key social indicators. It rapidly became one of the agency's bestselling flagship publications, with about 3,500 subscribers by 1987 and 5,000 two years later. Planning was also under way for a new economic flagship publication—the monthly Canadian Economic Observer. First released in 1988, this publication contained an analytical summary of current economic conditions, a feature article, a digest of changes in statistical programs, methodologies and services, as well as statistical tables of key economic variables. These two publications were followed a year later, in 1989, by Perspectives on Labour and Income, a quarterly review in the same magazine format. It focused on contemporary labour and income issues, while featuring new developments in surveys and methodology.

The shift toward user-centric publication was continuing, and this included a growing emphasis on publishing by topic rather than by survey. As a result, the agency was producing more and more compendia of information on a given industry, sector, or policy issue into one publication. While these compendia were labour-intensive to produce, they were very well received by data users. Examples included Shipping in Canada, Health Indicators, Focus on Culture, Youth in Canada, the Family in Canada, and Women in Canada: A Gender-based Statistical Report.

Available exclusively to educational institutions by subscription, the package included a tutorial, a teacher's handbook, a guide to the data, and classroom activities designed by educators. The data initially included were a selection of time series relevant to schools. Later, census data were added, ranging from Jean Talon's Census of New France in 1666 to the latest census information. While the target audience was originally meant to be the secondary school population, it proved useful for all students from senior elementary to postsecondary levels. Schools initially accessed the CANSIM data by modem (to an Ottawa-based computer) using TELIDON and TELICHART technologies. With the advent of CD-ROM technology, a CD of E-Stat was made available in 1992. Telichart allowed the plotting and overlaying of plotted data on different scales to allow students to establish relationships between different phenomena. The goal was to make working with data and learning about Canada easy and fun, and to develop computer literacy skills and analytic abilities. The side benefit was that a generation of Canadians would grow up having relied upon Canada's official data as a basic resource.

Statistics Canada launched its first CD-ROM data product in 1989 and, with the launch of electronic products, felt it had to confront the issue of licensing. From the outset, this was a controversial issue. It was believed that the agency's markets would erode unless it asserted its intellectual property rights and controlled or received compensation for further distribution of its information. Without markets, the agency would not be able to sustain its new products—so it began to license "defensively"—and the practice then morphed into an important revenue source for the cash-strapped agency. Ultimately, the high prices of products likely limited their more widespread use. The agency also launched a "national accounts on diskette" program, which was fairly successful until the technology became more or less obsolete with the arrival of the Internet.

The Internet emerges as a new tool

Computers were becoming much more common—by 1990, there was a computer for almost every employee. In fact, the entire headquarters complex had to be rewired as the use of local and broader networks was increasing. All voice and data cabling in headquarters was replaced, as the capacity was barely enough to meet current requirements of the time. When completed, the network capacity was 20-fold what it had been, and would be vastly more reliable.

In 1986, the first local area network was installed as a computer-linking system to improve the speed and efficiency of communication and work flow. By June 1994, the agency had developed two distinct computer networks to separate the confidential data on one network from the public data on the other. This allowed users direct access to public databases while protecting confidential data from unauthorized access.

The topic of one of the agency's lecture series presentations in 1993 was the Internet, which was just emerging. A SCAN article covering the lecture explained that: "A simple telnet command allows a user to log on to any machine with a public access catalogue….But knowing what you want is only part of the game - finding it is another matter entirely."  These were obviously before the days of today's strong search engines! A later SCAN article, published in 1994, described the Internet to agency employees as "a communications system, research tool, entertainment package and discussion forum rolled into one."

In 1994, there were two Internet "nodes," the first being the "STCINET" node, which let users perform remote log-ons, exchange email, and transfer electronic files around the world. The second node was called "Talon," which informed the Internet community about the agency and its products and services. Any information or data that would not otherwise generate revenue (such as research papers) were eligible for placement on Talon, such as information about the agency, the classification of statistics, and how to find and order publications. Issues of The Daily were also available in keyword searchable format.

Statistics Canada officially launched its website ("StatsCan Online") in March 1995, designed to deliver text, tables and graphics. It provided access to The Daily each morning at 8:30, as well as access to the CANSIM database, the international trade database called Trade Information and Enquiry Retrieval System (TIERS), and 40 virtual publications. After 71 years, the last paper version of The Daily, first published in 1932, would be published in 2003.

A stronger corporate identity

Since the Treasury Board's 1985 cost-recovery directive, which required a move toward full cost recovery as part of a government-wide strategy to reduce deficits, the agency has been developing more responsive products, aggressive marketing, and attractive packaging. The idea of establishing a new, stronger corporate identity was raised at a management conference in Cornwall in 1988. As a result, a new corporate identity was launched in March 1990. The new look for publications used four main elements—a graphic identifier (a little black logo of a graph), a square grid area (where an image would be displayed), standard typography (a typeface called ERAS), and standard positioning of the Federal Identity Program components (the Statistics Canada signature and the Canada wordmark). The agency also came up with a logo for its 75^th anniversary, in 1992, incorporating the colours of the agency, burgundy and blue, which was used on all publications, letterheads and envelopes.

In 1992, Statistics Canada's Marketing Division was created, to improve product planning by reviewing the needs of users and monitoring market response. While marketing certainly was not new at the agency, it was felt that it was time for a stronger, more visible, and more highly coordinated marketing function with consistent pricing. The aim was to become more client-oriented—to learn more about the agency's clients in order to specifically target their needs. By packaging products in a more user-friendly way, more people would be able to locate and use information more easily. It was also seen as a way of generating revenue to help address some of the shortfalls resulting from successive budget cuts. The division was charged with finding the most useful and cost-effective way to deliver products—drawing on all kinds of new and old technology available, including CD-ROM, diskettes, online databases, custom facsimile products, and, of course, paper.

The agency's flagship compendium publication, Canada: A Portrait, which presented the full range of the agency's information in a user-friendly fashion, was also redesigned. When it was released, in late 1992, The Ottawa Citizen called it "one of the most beautiful books ever published in Canada." Statistics Canada won the grand prize for Marketer of the Year from the National Capital Chapter of the American Marketing Association, apparently for "sheer marketing grit" in the face of unforeseen circumstances. The award rewarded marketing excellence and recognized innovative thinking and creativity; the winner was selected by four Toronto-based marketing experts. In essence, the marketing plan had to be changed when budgetary cutbacks throughout the government resulted in a sharp reduction in orders for the book from the usually strong federal market. The marketing coordinators sought out new markets—launching an aggressive pre-publication campaign that sold 15,000 copies before the book was even printed. In fact, it was so popular that an employee from the statistical reference centre kept the centre open late on December 24, 1992, in order to sell publications to frantic last-minute shoppers. Over 900 copies of Canada: A Portrait were sold in December alone, and 40 last-minute shoppers bought copies on December 24. All copies in stock were sold, and one customer even bought the display copy, joking that perhaps it would become a collector's item.

Notable milestones in the statistical system

In the late 1980s, policy concerns were arising for which statistical information was not available, including international competitiveness, impediments to interprovincial trade and their impact on constitutional arrangements, the impact on individuals of social, health and education programs, and the structure and performance of service industries. The agency put forth proposals and received additional funding over five years, some of which would come directly from the budgets of policy departments. The name of the initiative was Data Gaps (later to be referred to as Data Gaps I, when a second iteration came on scene), and it secured funding for the agency's first longitudinal surveys, to strengthen the system of national accounts, to fill gaps with respect to health, education and environment data, and to improve services-sector statistics.

The introduction of longitudinal surveys

In the 1990s, social policy researchers were increasingly feeling the need for micro-level longitudinal data to better understand the effects of social policy. As part of the new analytical focus at the agency, it was beginning in the exploration of longitudinal studies, databases, and surveys for greater analytical insight into changes occurring under the surface of aggregated information. For example, one project was analyzing the economic performance over time of firms engaged in international trade in services. Another was looking at the fate of individual workers faced with job loss, and analyzing changes in earnings and employment over the longer term. This decade saw the launch of many longitudinal surveys at the agency, and many policy departments and academic researchers were strong proponents of this new generation of surveying as a result of the potential for rich analyses of social phenomena.

There were also some "longitudinal-esque" surveys, such as the Labour Market Activity Survey (known as the Work History Survey until 1986), which collected retrospective information for 12 months, and combined this with a second 12-month period collected a year later. The Labour Market Activity Survey was conducted as a supplement to the Labour Force Survey, and was a major development for the household surveys program. Another longitudinal initiative was known as the Longitudinal Administrative Database, derived from administrative records rather than surveys. A 20% sample of Canadian taxfilers was constructed from the information provided annually to the Canada Customs and Revenue Agency in personal income tax forms. The coverage went back to 1982, with additional years added as they became available.

Three major longitudinal surveys were born in the first half of the decade: the Survey of Labour and Income Dynamics (SLID), the National Longitudinal Survey of Children and Youth (NLSCY), and the National Population Health Survey (NPHS).

The Survey of Labour and Income Dynamics (SLID) was funded from the first data gaps initiative (Data Gaps I), and its first period of data collection was in 1993. The survey examined changes in family composition, labour market activities, earnings and family income over time, shedding light on income transitions—such as escaping the "poverty trap." The survey collected information for six years from a sample of households and was designed to continue with overlapping panels. Halfway through the life of the first panel, a second six-year panel would be selected, and so on. The first panel of 15,000 households (about 31,000 individuals) was interviewed in January 1993 to collect preliminary information that would have a continuing influence on labour market activity and personal and family well-being, such as educational attainment, work experience, marital history, and the number of children born and raised.

Labour and income information was collected the following year. The information provided enabled the analysis of changes resulting from the economic climate, shifts in policy, and changes over the lives of Canadians. The first publication was released in 1995 with articles highlighting some of the results from the preliminary interviews. Although the analytical payoff of longitudinal studies was high, such studies were costly in terms of dollars and respondent burden. One of the ways this was partially mitigated was that, for income questions, respondents were given the choice of answering the questions or allowing the agency to access their tax file directly. In 1995, work was also beginning towards integrating SLID and the Survey of Consumer Finances as both collected detailed information on the incomes of Canadians. The integration was complete in 1998, and SLID became the source of both cross-sectional and longitudinal information on income trends.

The National Longitudinal Survey of Children and Youth (NLSCY), funded by and jointly conducted with Human Resources Development Canada (as the department was then called), went into the field in 1994. This was the country's first comprehensive national longitudinal survey dealing exclusively with the development of children. It collected information on a broad range of variables, including family circumstances, parenting style, neighbourhood perception, child-care arrangements, custody history, pre-school vocabulary, behaviour, socio-economic background, health, and school practices and performance. Information would be collected for 25 years on all factors that might influence a child's development, from the child's parents, teachers and principal, and from the children themselves once they reached 10 or 11 years old.

In 1992, developmental work was completed on the new National Population Health Survey (NPHS), and data collection began in June 1994 with funding from the first data gaps initiative. The survey would follow a sample of 17,000 Canadians over time to learn about the health risks they were exposed to, their activities related to health protection, the evolution of their health, their use of the health care system, and the impact of major health interventions. The survey was also designed to serve as a platform for supplementary content and to be linked to routinely-collected administrative data, including vital statistics, environmental measures, and community variables. This was an important new data source at a time when health care reform was a major issue across the country. The resulting information was used to help fuel policy discussions on the possible redesign of the country's health care system.

Acceleration in social statistics programs

A new Social Policy Simulation Database and Model was developed to allow researchers and policy analysts to study the possible effects of alternative tax and social policy changes on income distributions, on different types of families, and on federal and provincial expenditures. The database combined administrative data from personal income tax returns and unemployment claimant histories with survey data on family incomes and expenditures, presenting a detailed portrait of 170,000 "composite" Canadians in a non-confidential, statistically representative sample of Canadians. First released in 1990, it would significantly improve the richness of policy analysis and public debate in the area of public planning and social research.

A new version of the Uniform Crime Reporting survey was launched in 1988; it is referred to as UCR2, or the "incident-based" survey. The survey was designed to replace the aggregate pen- and-paper counts of police-reported crime and to standardize the detailed electronic records from police services on incidents, victims, and the accused. This was accomplished through standard interfaces with police record management systems and was one of the first electronic collections of administrative data for Statistics Canada. By 2009, all major and almost all small police services had converted to the new administrative data survey.

In the early 1990s, two other important social surveys were initiated: the Transition Home Survey and the Violence Against Women Survey. Both were conducted as part of Canada's Family Violence Initiative, the federal government's main collaborative forum for addressing family violence. Still in existence today and coordinated by the Public Health Agency of Canada, the Family Violence Initiative is a collaboration between 15 partner departments and agencies to help prevent and respond to family violence. Statistics Canada's contribution has been through improving the availability of national level data on the nature and extent of family violence.

The Transition Home Survey was first conducted in 1991/1992 as a supplement to the Residential Care Facilities Survey to address the need for information about services for victims of family violence. Starting in 1992/1993, the Transition Home Survey became an independent survey and is the only source of information in the country on shelters for abused women and their children. As a census of residential facilities providing services to women and their children, it collects information for service providers, non-profit organizations and governments, which is used to develop programs and services. It has been conducted every two years since its inception.

The Violence Against Women Survey was first conducted in 1993, and was the first survey measuring the nature and extent of violence against women in Canada. Conducted as a collaboration between the Canadian Centre for Justice Statistics and the Social and Aboriginal Statistics Division, it was one of the most sensitive surveys ever conducted by the agency. It was also one of the first of its kind in the world and therefore generated a great deal of international interest. The interviewers were all women who were carefully screened for sensitivity and were trained by a clinical psychologist before the survey began. Over 12,300 interviews, each lasting from 15 minutes to 3 hours, were conducted with women aged 18 and older. Respondents were given a toll-free number to call back in case they could not comfortably or safely complete the interview. The manager of the survey, Holly Johnson, was presented with the Employee of the Year award recognizing her leadership, professionalism, and outstanding performance in leading the survey. Her advice was subsequently sought for similar surveys in Russia and the United States. She also co-edited a special edition of the Canadian Journal of Criminology dedicated to the survey.

The General Social Survey: Game changer

One of the most important social realm milestones of the 1980s was the establishment of the General Social Survey in 1985. This multi-purpose survey was conceived to allow for the observation of long-term social trends as well as to collect information on social issues of immediate interest. The survey began with a sample size of 10,000, which grew to 25,000 by 1999. Aimed at those 15 years of age and older, the survey investigated five main themes on a rotational basis, including health, time use, victimization, education and work, and family.

The survey emanated from Dr. Fellegi's secondment to the U.S. Census Bureau in the late 1970s. At the time, he realized how much further advanced the Americans were in terms of social statistics, especially as the Canada Health Survey had just been cancelled and the Labour Force Survey was the only source of current data on Canadian households. Upon his return, as Assistant Chief Statistician of Social Statistics, he found funding for the survey, but as budgets were tight, the most feasible approach to collecting such a vast amount of information was on a cyclical basis. In fact, it would become anecdotally known as the "poor man's survey." The General Social Survey collected information on a different topic each year on a five-year cycle, with health and social support networks for the elderly as the focus for Cycle 1 of the survey. The survey has also served a springboard for other new surveys, which initially formed part of a cycle (such as health and education), before becoming independently conducted.

The evolution of a Canadian health information system

In 1988, the National Health Information Council (NHIC) was formed as a joint federal-provincial body by the Conference of Deputy Ministers, with the aim of improving health information through the development and coordination of a national health information system. The NHIC was a body of provincial assistant deputy ministers and their counterparts from Statistics Canada and Health and Welfare Canada. It reported to the provincial and federal deputy ministers of Health and the Chief Statistician on data priorities, the division of responsibilities, and the adequacy of resources. The NHIC's terms of reference included reviewing the concept of a National Centre for Health Information and how Statistics Canada fit into the picture. Statistics Canada proposed renaming its own Health Division to the Canadian Centre for Health Information (CCHI), and giving the NHIC purview over all of its health information programs, including powers to set program priorities and strategic direction. The NHIC agreed, and the CCHI was established in 1989.

The NHIC was intended to develop and maintain an ongoing national consensus on all aspects of national health information systems and to continuously review priorities and plans related to the collection, processing, and release of health information at the national level. The purpose of bringing together representatives from all federal / provincial / territorial health departments, including Health and Welfare Canada, and Statistics Canada, was to facilitate the direct involvement of the provinces and territories, with responsibility for health care, in providing guidance on their priorities. It would also support provincial consensus in terms of the provision of comparable health-related data across the country. There were challenges with roles and responsibilities between the NHIC, Health and Welfare Canada, and Statistics Canada, as the mandate of the NHIC spanned several key areas in both organizations. There was also continued difficulty in achieving collaboration with the provinces and territories.

In April 1990, the NHIC accepted a joint Statistics Canada-Health and Welfare Canada offer to sponsor a consultant to facilitate the further development of the NHIC's plan for the evolution of a national system for health information. The proposal was further endorsed at a June meeting of the deputy ministers of health.

A task force, led by former Chief Statistician Martin Wilk, was created. It tackled a major review of the country's health information requirements to better understand the factors affecting health, and how the government could be spending its money more effectively. Expenditures on health were close to 9% of Canadian gross domestic product (GDP). However, there was very little information on the effectiveness or efficiency of the health system in improving health outcomes. The popular question at the time was "What are we getting for our health dollars?" The task force also formulated planning recommendations for the NHIC toward the objective of establishing a health information system for Canada. In 1991, the task force published A Report by the National Task Force on Health Information, popularly referred to as "The Wilk Report," which described health information in Canada as being in "a deplorable state," lacking coordination, planning, and responsiveness to user needs.

The report recommended the creation of a non-governmental, not-for-profit national organization to coordinate health information. After intense discussions by the ministers of health, approval in principle was given in September 1992 for the creation of the Canadian Institute for Health Information (CIHI), a national mechanism to coordinate the development and maintenance of a comprehensive and integrated health information system for Canada. Its creation allowed for an independent linkage of federal, provincial, and territorial governments with non-governmental health groups. Its mission would be to provide and coordinate the provision of accurate, timely information required for the establishment of sound health policy, for the management of the health system in Canada, and for generating public awareness about the factors affecting good health.

The new institute would amalgamate existing operations and responsibilities (in whole or in part) from four organizations: the Hospital Medical Records Institute, the Management Information Systems Group, Health and Welfare Canada's Health Information Division, and Statistics Canada's Canadian Centre for Health Information. In identifying those activities to be transferred to the CIHI, decisions were based on comparative advantage in terms of cost, infrastructure, or demonstrated competence, as well as legislated mandates. Health and Welfare Canada and Statistics Canada wrote letters of intent delineating their positions with respect to the new institute, including that they would not compete in production of revenue-producing activities or services, subject to being able to fulfil their legislative mandates. The NHIC would be disbanded, although the steering committee continued to act as an interim board of directors to oversee implementation planning for the new institute.

The official transfer of most operations and responsibilities to the CIHI took place in 1994, when it began operating. It was determined that the CCHI would cease to exist once the CCHI was established. However, the production of health statistics at the agency could not disappear altogether as there were functions that were not candidates for transfer or devolution. Statistics Canada transferred a number of activities along with their associated funding to the CIHI, including the following: hospital facilities and operations data, residential care facilities data, hospital inpatient admission / discharge abstracts, patient-oriented database development, information on registered nurses, the tuberculosis registry (and other similar registries), health status analyses, and nosology. As well, to avoid confusion when CIHI became operational, CCHI would be renamed the Health Statistics Division.

New health information initiatives

A number of surveys were conducted in the late 1980s and early 1990s to promote and develop health programs. These covered a wide range of health issues, including legal and illicit drug use, mental health, dietary habits, the quality of family relationships, and mental health conditions and how they affected peoples' lives. Health and Welfare Canada funded the first Drinking and Driving Survey in 1988/1989 to develop information for public awareness campaigns, as well as the 1988 Drug and Alcohol Survey to explore attitudes to drug and alcohol use. The Ontario Ministry of Health funded the 1990 Ontario Health Survey, for which the agency participated in an open bidding process, winning against the private sector for both quality and price. Other surveys included the Survey of Mental Health funded by the Ontario Mental Health Foundation, and the Health Promotion Survey, funded by Health and Welfare Canada.  

After many years of cooperation and collaboration between provincial / territorial cancer registries and Statistics Canada, the Canadian Cancer Registry (CCR) was established in 1992 as the agency's vehicle to collect cancer-related information from all provinces and territories. This replaced the National Cancer Incidence Reporting System (NCIRS), which had been in place since 1969, and had provided a database to study cancer patterns and trends at the tumour level. While the NCIRS was an event-oriented database, the CCR was established as a patient-oriented database, with the advantage that valuable longitudinal data were made available for all Canadians diagnosed with cancer.

A new national education statistics program

Education and lifelong learning were emerging as key determinants of competitiveness, and it was becoming clear that the country needed better measures of the performance of the education system as well as a framework to better understand educational outcomes. Very little information existed on the relationship between education and labour market success, and the ability of individuals to adjust to changing technologies and economic conditions. Similarly, little data existed on transitions such as that from school to labour market, and those between employment and unemployment, or between employment and retirement. While the agency had a relatively complete picture of the operations of the education system, it had very little empirical information on whether the system was producing the right skills for the labour market, the extent and causes of dropping out, the effect of class size or teacher training, and the relationship of education to socio-economic status.

In 1988, provincial deputy ministers of education agreed to a formal working arrangement with the agency to develop a comprehensive national education statistics program. As a result, the Canadian Education Statistics Council (CESC) was established in 1989 as a federal-provincial-territorial partnership between Statistics Canada and the Council of Ministers of Education to govern the education statistics program. A year later, Canada-wide statistical measures for elementary, secondary and postsecondary education were first produced and published.

Adult literacy around the world

The agency's international reputation was growing, and it was increasingly being asked to take the lead on international initiatives. For example, Canada was asked by the United Nations in 1990 to play a leading role in redesigning the United Nations International Crime Survey. The survey collected information for international comparisons, in which Canada was one of several participating nations.

The agency was also asked by the Organization for Economic Co-operation and Development (OECD) to provide international coordination and leadership for a multinational survey of literacy—the world's first large-scale multinational comparative assessment of adult literacy skills. It was to allow the analysis of the relationships between literacy and education, and between labour market success and economic growth. Administered in 1994, 1996 and 1998, the International Adult Literacy Survey was a large-scale co-operative effort by governments, national statistical agencies, research institutions and the OECD. The development and management were coordinated by Statistics Canada and the Educational Testing Service of Princeton, New Jersey, with input from the National Centre for Education Statistics of the United States Department of Education. Nine countries participated in the first cycle, in 1994: Canada, France, Germany, Ireland, the Netherlands, Poland, Sweden, Switzerland, and the United States. Encouraged by the success of the first cycle, five additional countries or territories also administered the next cycle (Australia, the Flemish community in Belgium, Great Britain, New Zealand and Northern Ireland). For the last cycle, an additional nine came on board (Chile, the Czech Republic, Denmark, Finland, Hungary, Italy, Norway, Slovenia and the Italian-speaking region of Switzerland).

The survey was not without controversy. After the first round of data collection and analysis, France withdrew its results before publication. The withdrawal was motivated by concerns with respect to the comparability of the results, specifically that they were thought to underestimate the true literacy skills of the adult population in France relative to those of adults in the other countries. The European Commission commissioned a review of the survey, which was conducted by the United Kingdom Office for National Statistics. Before that review commenced, data had already been collected for the second round (these were published in 1997), and, while the review was under way, the third round of data was collected. The review, which covered only the first cycle, was published as the final analysis was being prepared. While it identified some issues, most had already been dealt with as lessons learned from the first cycle. Over the course of the second and third rounds of data collection, more stringent specifications and quality assurance procedures helped to reduce inter-country variation in survey practice.

Re-examination of Confederation

The government was re-examining Confederation in 1987 and 1991, in the context of the Meech Lake and Charlottetown accords, in an attempt to devolve power to the provinces and territories in the hopes of bringing Quebec into the constitutional fold. In preparation for these discussions, it was necessary to better understand the economic links between the provinces before any changes to the existing arrangements could be proposed. However, 1984 had been the last time a survey of interprovincial trade had been conducted. The agency received funding through the Data Gaps I initiative to carry out a much needed survey on the origin and destination of provincial trade and to update interprovincial input–output tables.

Defending the agency's political neutrality

The year 1991 also marked the introduction of the new 7% Goods and Services Tax to replace the Manufacturer's Sales Tax, which had been levied at the manufacturer's gate. The tax had to be accounted for in the national accounts, and the agency was preparing for its implementation in both its statistics and its operational transactions with the private sector. An attempt was made to suppress the publication of an analytical article after the introduction of the tax.  Prices had risen as expected, and the analytical article attempted to shed some light on how much of this rise was attributable to the new tax versus what inflation would have been without it. Dr. Fellegi categorically refused to put off publication of the article, and refers to the incident as the only incident in his tenure where there was an attempt at interference with the political neutrality of the agency from government.

Businesses: Small and large

The importance of small businesses to the economy continued to be recognized as these firms continued to be the focus of many initiatives at the agency. From 1978 to 1986, the number of questionnaires sent to small businesses was cut in half. As well, a toll-free information service was planned in order to assist businesses across the country in completing questionnaires. A new series of statistical reports on small business was launched in 1986, including Small Business Profiles, which presented a range of financial data (including assets and liabilities, working capital, operating expenses) for various industries, for use by small-business owners and managers. A Small Business and Special Surveys Division was also created in 1987/1988 to carry out customized surveys that could be piggy-backed onto existing surveys at a small additional cost.

The Large Enterprise Statistics Program was established in the early 1990s with 9 profilers from Ottawa, Montreal, Edmonton, and Vancouver. The profilers each had a portfolio of businesses for which they had continuing responsibility and with whom they would liaise and establish working relationships. The profilers would research, visit the businesses, assemble material, and update the Business Register as required. With one profiler gathering all the information required from a given company, data collection was streamlined by means of reduced respondent burden and less duplication of information, resulting in more complete and better-integrated data. The program also served to reinforce good relationships with large companies, which represented about 30% of all business activity in Canada at the time.

As well, the agency instituted a new program to regularly measure response burden. An initiative that contributed substantially to the reduction of response burden was the redesign of the Annual Retail and Wholesale Trade Survey. For years, the agency had conducted an annual census of wholesalers and retailers to obtain detailed information for the calculation of the value added by these industries. It also served as a benchmark for the monthly sample surveys of wholesalers and retailers. Those surveys, however, had recently been redesigned and strengthened to such an extent that they could stand on their own without the need for the benchmarking provided by the annual census. In addition, it was realized that a lot of the more detailed content could be collected more efficiently through a specially expanded version of the January cycle of the sample survey. These changes were implemented for the first time in 1994, greatly reducing the response burden on businesses as well as reducing costs for the agency.

Statistics Canada switches from gross national product to gross domestic product as a prime measure of output

In July 1986, a historical revision of the System of National Accounts was released, after only 2 years instead of the 6 years taken for the last major revision, in 1972. The revision included two major changes—the switch to using gross domestic product instead of gross national product along with reworking the data from 1961 to 1986, and the rebasing of the price indexes to 1981 constant dollars. The most visible difference in the revised accounts was the switch from gross national product to gross domestic product as the prime measure of the country's output. It was felt to be a more appropriate measure of output for the analysis of inflation, production, employment and productivity, given the important contribution of foreign investment to the country's economic development.

While gross national product measures production "owned" by residents, at home or abroad, and hence is a better indicator of the income of Canadian residents, gross domestic product measures domestic production regardless of whether the production is owned by residents or non-residents. In other words, it measures the impact of all production taking place in Canada regardless of who owned the assets. The switch better aligned Canada's practices with international practice and recommendations of the UN System of National Accounts, and allowed the country's accounts to be directly comparable with other measures of economic activity in Canada and with the National Accounts of most other countries. In 1989, the GDP numbers at constant prices were rebased for 1986, and would thereafter be recomputed every 5 years instead of every 10.

International trade statistics initiatives

The agency had been working with the international community since 1971 on the development of a new international commodity description and coding system—called the Harmonized System. The aim was to come up with a standard identification of commodities in trade documents and improve the comparability of international statistics. It was not an easy feat—the system comprised over 5,000 product descriptions and codes, with each definition and classification agreed to in plenary sessions. Created under the authority of the Customs Co-operation Council (now known as the World Customs Organization, an independent intergovernmental organization based in Brussels, Belgium), it was accepted as final in June 1983. It would come into effect almost 5 years later, in 1988. It would eventually be used by over 200 countries and economies as a basis for their customs tariffs and for the collection of international trade statistics. The initiative greatly assisted in trade data comparability and therefore helped to facilitate international trade negotiations. The new classification system was instated at Statistics Canada after a three-month parallel run under both the old and the new systems.

One of Canada's most crucial international statistical relationships was that with its key trading partner, the United States. In 1971, an agreement between Canada and the United States was reached, which provided the basis for the countries to reach mutually agreed-upon estimates for import and export data. This involved an annual reconciliation process of aggregate data five to six months after each calendar year. In July 1987, a new memorandum of understanding was signed on the exchange of import data, considered a world first between any two trading nations. As there was consensus that import data were generally better measured than export data (mostly because imports are the source of customs duty and exports are not), the countries agreed that, by 1990, they would replace their own export figures used to calculate the trade balance with the other country's import figures, in detail, by commodity.

This replacement was dependent on the adoption by both countries of a compatible system of coding commodities, the new Harmonized System, which was implemented at Statistics Canada in January 1988. The system's introduction in the United States ended up being delayed as it formed part of the Omnibus Trade and Competitiveness Act of 1988, which would come into effect one year later, on January 1, 1989. The two countries also agreed upon an official conversion key for industrial classifications that would permit detailed industrial analyses of one country's performance against the other's. This was becoming increasingly important as a result of free trade agreements.

March 19, 1990, marked the first release of data from the Canada-U.S. data exchange. Its implementation at Statistics Canada had the side benefit of resulting in efficiency gains. In 1988, when the Harmonized System was first introduced, the International Trade Division was working to full capacity with over 200 people. By 1990, with a reduced workload in the exports section, and the Harmonized System in place, the division was working efficiently with a complement of 130 people. Outside of the agency, it had an ever bigger impact as it exempted exporters from completing more than 2.5 million documents each year.

The agency was also in the midst of creating a world trade database, the TIERS, to enable the measurement of trade between other countries. For example, trade reconciliation studies with Japan and the European Community were under way. Information about trade among other countries, their output and employment, was added to data the agency already possessed on Canada and its trading partners. This broad international data set was among the first of its kind, allowing Canadian businesses to assess the relative competitiveness of Canada in other markets. In 1992, TIERS included merchandise trade statistics based on customs data for over 16,000 import and 6,000 export commodities of 200 countries under the Harmonized System.

Measuring the services sector

The statistical system was formed and matured around the concept of goods—the primary industries of agriculture, manufacturing, mining, and fishing, for example. The services sector grew in relative importance and, in the post-war period, began to surpass the goods sector. It was not only the fastest-growing sector of the Canadian economy, but also one of the most conceptually difficult to measure. Statistics Canada was therefore keen to improve measurement in this area, but found it difficult to redirect resources from goods surveys to services surveys, not only because of the large shift required internally, but also because of the required budgetary reallocation.

Statistics Canada proposed that statistical agencies of several countries pool resources to develop a classification system for service industries. Each country would write draft classifications for one industry, with Canada drafting the classification for the architects and engineers industry, and later the classifications for the telecommunications industry and the finance and insurance industries. The proposed classification system was officially sanctioned by the UN statistical office in February 1989.

Informal and voluntary meetings of representatives from national statistical agencies to address selected problems in statistical methods have become formally known as "city groups," named after the place where they hold their first meeting. They are an innovative and efficient use of resources to improve international statistical standards. The city group on service statistics became known as the Voorburg Group on Services Statistics, as it was hosted by the Netherlands Statistical Office in January 1987. Nine countries and six agencies gathered to collaborate on a plan to build an internationally accepted classification for service industries. Canada hosted the fourth meeting of the group in October 1989, at which the group discussed 37 papers on topics ranging from international trade in services to proposals for a revised commodity classification for specific industries. One of Statistics Canada's prior chief statisticians, Dr. Sylvia Ostry, took part in the meeting as a panelist for one of the agenda items on the information needs of service industry policies. The group met yearly to examine the progress made in each country. In 1991, it discussed and endorsed a Canadian proposal to conduct an internationally comparable survey of the computer industry.

This services data gap also resulted in a modest expansion in the number of services price indexes. This was not an easy feat, as services indexes are much more difficult to construct than goods indexes owing to the more subjective nature of services. It is harder to hold the quality of services, than the quality of goods, constant, for the purpose of measuring a pure price difference. The development of services indices had begun in the 1980s with the Consulting Engineering Services Price Index. To help address the data gap, new price indexes were constructed to cover topics such as informatics professional services, telecommunications, traveller accommodation, and accounting services.

New generalized systems

In 1987/1988, much of the developmental work in methodology involved the creation and testing of a Generalized Survey System, a set of standard methods for conducting surveys. The system would remove the necessity to design a custom-made program for each survey, and thus would reduce the time, money and energy required, and would promote the use of the best survey methods. The first two modules, the sampling module and the edit and imputation module, were put in place in 1988. Work was continuing on the data capture and tabulation models. In fact, two of the largest applications were introduced in the 1991 Census Program—the generalized system providing automated coding by text recognition for the Census of Population, and the Generalized Edit and Imputation System for the Census of Agriculture. By 1992, the data collection and capture system had been developed to the point where it could be implemented both at headquarters and in regional offices. While use of the generalized systems was a corporate priority, the challenge was to ensure their use in all new or redesigned surveys except under exceptional circumstances. In 1994, a new Generalized Systems and Methods Training Centre was opened, which provided courses and demonstrations of the systems. At the time generalized systems were saving the agency $0.8 million a year on business surveys alone.

Census programs

A significant feature of the 1986 Census was the sheer volume of unpaid advertising and support, including from agency employees. Census workers, along with their friends and relatives, set out to design and create a six-foot-high ice sculpture of the 1986 Census symbol (the "bonhomme") for Winterlude. A team from the Canadian Centre for Justice Statistics also participated in the ice sculpture contest, with a sculpture of the arrest of a suspect. An agency employee wearing a bonhomme costume also made public appearances, including at schools, to encourage children and their families to be "counted in." In addition, famous Canadian personalities donated their time to promote the Census through filmed public service announcements, which were used extensively across the country. These included Andy Moog and Kevin Lowe of the Edmonton Oilers hockey team, Dale Hawerchuck of the Winnipeg Jets, singer Ian Tyson, 1984 Olympic medalists Alwyn Morris and Caroline Waldo, and actor John Vernon. The effectiveness of the unpaid communications strategy depended also on the participation of organizations with communications vehicles that recognized the importance of the census for their own operations. For example, the Safeway chain in Western Canada printed the census logo on their trucks and grocery bags.

The 1986 Census included questions on Aboriginal identity designed to provide better data on the numbers and characteristics of Inuit, status Indian, non-status Indian, and Métis populations in Canada. A separately funded Aboriginal Peoples Program was also established to build statistical capacity among Aboriginal peoples and provide data to Aboriginal communities and organizations for their own purposes (e.g., service delivery and program development).

The 1991 Census welcomed a number of changes, such as collecting information on common-law relationships as well as marriages. As well, mail-back was extended to virtually all areas of the country, while previously only 60% of the population mailed back their questionnaires. The Census also began including non-permanent residents in the target population for the first time. Post-censal surveys for the 1991 Census included a survey of Aboriginal peoples and a survey of people with a disability. The latter was to be similar to the Health and Activity Limitation Survey, conducted in conjunction with the 1986 Census, but would provide more current information and enable an assessment of changes since 1986.

Mr. Census

In November 1992, a man fondly referred to as "Mr. Census" passed away. Dr. Edward T. Pryor was a sociologist, a scholar, and a renowned speaker who had worked on virtually every facet of the Census since 1968, on a total of 6 censuses. His passion for sociology had inspired many others to follow in his footsteps, and he was a mentor for many young students and employees. He served as a consultant for the 1971 Census, and as an advisor on the census reorganization plan. Later, as Director of the Census Characteristics Division, he was in charge of the 1971 Census Products and Services Program. He served as a manager of the 1981 Census, before becoming Director General of the Census and Demographic Statistics Branch. He was also Project Manager for the 1991 Census. He wrote or co-wrote some 30 articles and publications, and was an ambassador for the agency at home and abroad, including assisting China with its 1982 Census. After his passing, he was recognized with Statistics Canada's Career Excellence Award, which Dr. Fellegi presented to his wife who accepted it on his behalf. An Edward T. Pryor Bursary Fund was also established at Carleton University to provide financial assistance to sociology students.

Integrating tourism

In 1983, the General Assembly of the World Tourism Organization made a case for integration of tourism statistics into the System of National Accounts. While the tourism industries (hotels, restaurants, taxis, airlines, tour operators, etc.) were indeed in the accounts, there was no overarching aggregation to show the impact of tourism as a "quasi-industry" unto itself. In response, a national task force on tourism was established in 1985 by federal, provincial and territorial ministers of tourism with retired Chief Statistician Martin Wilk as chair. This was a joint public- / private-sector task force to establish a statistical foundation for the planning, marketing, management and profitability performance of tourism. The agency set up an internal research activity to lend analytical support to the development of a new database and provided the secretariat to the task force, with over 40 employees from different areas of the agency contributing time and experience to the two-year project. The task force worked with more than 50 organizations from all sectors of the industry—government, business, trade associations and universities—and presented its final report to tourism ministers in November 1986. Its recommendations included that existing federal and provincial surveys be improved and that a Tourism Research Institute be established to determine how to meet future data needs. In March 1987, Martin Wilk was awarded with the F.G. Brander Memorial award by the Tourism Industry Association of Canada in recognition of the task force's work. This work would lead to a new tourism satellite account in 1994, which would provide an integrated picture of tourism-related spending to show tourism's overall contribution to output, employment and exports.

Address Register and Business Register

A number of operational databases were established in the late 1980s, including the Address Register, which was established in 1987 and was initially used as a post-listing coverage check for the 1991 Census. Staff produced residential address lists for 23,000 enumeration areas in urban areas of the country, by developing lists from various sources, such as telephone listings and hydro billings, and matching them to census geography. The Address Register was an idea that had been revisited a number of times since the mid-sixties, when it was first investigated for the 1971 Census. Finally, in 1987, as part of the research program for the 1991 Census, the possibility of a register of residential addresses for urban areas was shown to be feasible. Prior to having an address register, each census representative would compile an address list for their enumeration area by hand and deliver a questionnaire to each dwelling. However, it was one thing to hand a respondent a questionnaire and ask this individual to mail it back, and quite another to directly target and mail the questionnaire to the right household. The coverage and quality of the Address Register would improve over ensuing years, until it was of sufficient quality to be used as the basis for the mailing-out of census questionnaires in 2006.

The corporate Business Register, a centrally maintained database containing detailed descriptions of most business entities operating within Canada, also continued to be improved. It would, however, not achieve its full potential until the late 2000s. Statistics Canada took leadership in creating a forum for the international exchange of experience in the area of business registers. The first meeting was held in Ottawa in 1986, with the group referred to originally as the "Roundtable of Business Survey Frames." It was renamed as the Wiesbaden Group on Business Registers in 2007, when it met in Wiesbaden, Germany, to align with the naming convention of other city groups.

The monthly surveys of retail and wholesale trade were the first to use the new Business Register, in 1988. In fact, among the first changes from the Business Survey Redesign Project was a redesign of the Retail Trade Survey to provide better estimates of retail trade activity. The newly redesigned survey would draw its sample from the new Business Register. The next to be added to the Business Register would be the Survey of Employment, Payrolls and Hours. By the end of 1989, most annual surveys would be tied into the Register.

One such survey, the Survey of Employment, Payrolls and Hours, took more than three years, beginning in 1989, to convert to the Business Register. A team of researchers, developers, and support staff worked on this project, which was so large and so complex that it was named "the Big Bang Project."

The work consisted in linking the survey to the new Business Register, smoothing data breaks in the time series, converting data series to the 1980 standard industrial classification, producing seasonally adjusted data, and creating new information products to help users understand the revised data. The redesign was completed in 1994, and it would make use of administrative data on small businesses that were newly available from Revenue Canada's payroll deduction form on the number of employees and monthly payroll. The additional variables were obtained from a small overlap survey done by computer-assisted telephone interviewing of 2,500 businesses. Before the redesign, 30,000 businesses would have been surveyed.

The use of tax records also reduced operating costs by about $0.9 million per year. When the survey was originally designed, in the late 1970s, using administrative data was actually recommended at the time. However, the momentum had not been there to carry through with the recommendation. Contributing to this lack of momentum for an earlier redesign was the agency's relatively poor relationship with other agencies in the late 1970s.

Environment

The topic of the environment was coming to the forefront, in terms of both data needs and daily working life in the public service. In the 1960s, Rachel Carson's revolutionary book Silent Spring and issues such as acid rain spawned a new era of environmental awareness. The Department of the Environment Act established Environment Canada in 1971 in recognition of the need for better environmental management. The agency's environment statistics program was established in the mid-1970s, and at the time it was associated with a program on the development of social indicators. A multi-disciplinary research team of geographers, biologists and economists collaborated to develop the databases that would be required to analyze the interaction between people and their environment.

In 1989/1990, with funding received via Data Gaps I, the agency was developing statistics to assess and track the state of Canada's environment. An August 1989 SCAN article cites that "a new breed of 'environmental statisticians' emerged," housed in the environment and natural resources section of the Analytical Studies Branch. The agency's data on population, agriculture and manufacturing were reorganized by watershed and river basin, to assist users conducting environmental studies in local areas. Data on environmental quality and natural resources were drawn from Environment Canada, Agriculture Canada, and Energy, Mines and Resources, as well as from provincial and territorial environment departments, and integrated into the agency's socio-economic databases. A framework, referred to as the stress-response environmental statistical system (STRESS), was developed and became the basis for the United Nations' statistical framework for the development of environmental statistics. Two important agency publications at this time were the first State of the Environment Report for Canada, written and jointly published with Environment Canada and first released in 1986 at a press conference by then-Minister of the Environment Tom McMillan; and the compendium called Human Activity and the Environment, first published in 1978.

In 1992, the agency received funding through "The Green Plan" for an environmental satellite account program. The Green Plan was an environmental action plan launched by the federal government in 1990 as a national strategy and action plan for sustainable development. It also provided funding for various initiatives over the following five years. The government called upon Statistics Canada to supply a set of indicators to present the state of the environment. In 1993, a National Round Table on the Environment and the Economy would be convened. The agency would rename its Income and Expenditure Accounts Division "National Accounts and Environment Division," and began work toward a satellite account for natural resources. An ad hoc working group was formed in 1993 as another city group to tackle the difficult conceptual issues to do with measuring the environment—the London Group on Environmental Statistics, that first met in London, England, in March 1994. It continues to meet annually as a forum for review, comparison, and discussion of work under way in the area of environmental accounts.