The Sum of the Satisfactions: Canada in the Age of National Accounting.
By DUNCAN McDOWALL. Montreal and Kingston: McGill-Queen's University Press, 2008. Pp. xii, 313, index.
Paul Samuelson described national accounting as a "great invention of the twentieth century--without (which) macroeconomics would be adrift in a sea of unorganized data" (cited on p. 164). Canada has played a leading role in the development and application of the national accounts. It is entirely fitting that the National Accounts Advisory Committee commissioned Duncan McDowall, professor of history at Carleton University, to write a book on Canada in the age of national accounting. McDowall has done exhaustive archival research and interviews and succeeded in presenting complex issues in a generally readable way. The result is not only a highly informative history of the development of national accounts in Canada. It does this in the context of the international intellectual history, of which the accounts form a part, and of their application to broad areas of economic policy in Canada.
The interest in treating national economies in a holistic rather than fragmented way has a long history: McDowall notes such efforts by William Petty in the late 17th century and Francois Quesnay in the mid-18th century. The establishment of the Dominion Bureau of Statistics in 1918 was partly because of the desire to perpetuate the census and the special studies of the agricultural and industrial sectors but also to update the 1915 estimate of wealth and to produce national income data. The lead with respect to the latter moved to the United States, thanks to the establishment of the National Bureau of Economic Research in 1920 and the success of Simon Kuznets in developing a production-driven model of national income in the mid-1930s.
The major impetus for broadly based national accounts came from Keynes' "general theory," which placed analysis of depression in the context of consumption, investment and government variables in the national accounts. The same approach underlay demand management during the war and planning for post-war employment, social welfare and regional equalization as well. James Meade and Richard Stone produced the first credible estimates of national income in 1941, which were shortly after reflected in the budget's attempts to measure and control the wartime inflation gap. The United States was not far behind. A working group led by Donald Macgregor for the Royal Commission on Dominion-Provincial Relations had already produced far better estimates emphasizing provincial and government roles but the Dominion Bureau of Statistics remained wedded to the Kuznets and National Bureau of Economic Research models.
This situation was particularly frustrating for Robert Bryce, a student of Keynes and now Clifford Clark's right-hand man in the Department of Finance and also secretary to the powerful Economic Advisory Committee of the Cabinet. Bryce saw very clearly the need for reliable and comprehensive national accounts if depression and war management and post-war planning were to be effective. At a key meeting in April 1942 of the senior people of the central bank and the government bureaucracy, Clark tried to end the "bureaucratic turf protection and intellectual confusion" (p. 55) regarding attempts to produce reliable national income data by putting the responsibility on the Dominion Bureau of Statistics. Nevertheless, it was not until early in 1944 that the DBS created the Research and Development Staff to work on a more developed set of national accounts, a move brought on in part by the concern about losing this responsibility to the Bank of Canada, where considerable progress had already been made in this area. George Luxton transferred from the Bank to head the new unit. Not long after, Richard Stone and Luxton met Milton Gilbert, the head of the national accounts section of the U.S. Commerce Department. They worked out a common set of standards that set the course for post-war national accounts. Luxton died not long after, but he and his colleague Agatha Chapman of the Bank had done much to improve the national accounts. The "godfather" of the Canadian national accounts, Robert Bryce, kept an eye on developments here, despite his many other responsibilities.
Claude Isbister, then Simon Goldberg, brilliantly followed up the Luxton-Chapman contribution. The time was ripe for major developments in these fields, given such challenges as reconstruction plans, improved social security, and the need to find a substitute for the wartime tax rental agreements with the provinces. And the newly appointed Dominion Statistician, Herbert Marshall, immediately made his priorities clear by declaring the national accounts "the keystone of the statistical arch" (p. 100) and setting up a weekly after-hours study group in his office. Goldberg's remarkable influence over a long period justifies a particular note. Joining DBS in 1945, he became head of the Research and Development Staff in 1950 and Assistant Dominion Statistician in 1954. Despite the broader responsibilities, he never lost his passion for the national accounts in the years up to his resignation in 1972 to head the United Nations Statistics Division. Most of the developments in that long period reflect his emphasis on the national accounts as an integrated system where the interrelationships among economic statistics required consistency in definition and classifications. Typical of his approach was to place DBS's resources at the service of researchers such as William Hood and his colleagues with the Royal Commission on Canada's Economic Prospects, then to use their findings and estimates on such matters as financial flows and fixed capital formation to begin or improve a major series for DBS. In the 1950s, DBS led internationally in some aspects of its work on seasonal adjustment and real output. In the 1960s, there were improvements in both sectoral and regional data as well as productivity measures and assisting with econometric modelling in other departments. An outstanding development was in input-output measures, following on ideas and work initially developed by Wassily Leontief of Harvard University in the 1930s. Goldberg and a notable array of colleagues took a big chance in entering this complex field early, but they knew there were large statistical and policy payoffs.
The mature state of the national accounts was shattered by two developments in the 1970s, reflecting deep social and economic changes. One was the increasing volatility of the economy in response to oil and exchange rate shocks, stagflation and some structural changes. The other was the increased demand that the accounts should reflect a broader view of economic and social welfare, such as the costs associated with resource depletion and pollution, changes in social welfare beyond income, and non-monetary contributions such as unpaid work of housewives in particular.
Statistics Canada (or Statscan), as the Dominion Bureau of Statistics was now called, responded to both of these challenges but with difficulty. In terms of the first challenge, one now saw monthly real domestic product on an industry basis, sectoral productivity estimates, annual input-output data, more detailed balance of payments data and others, all available in a long historical series, thanks to the work by Mac Urquhart and his colleagues, who received Statscan's support. Nevertheless, some important revisions in the data appeared in this decade, as the pressures for early release of a broader set of data mounted. An inquiry by five distinguished non-Canadians, set up by the government, gave high praise in 1980 to Statscan's work and especially the national accounts. However, it also concluded that too much attention had been given to organizational change; the agency needed to get back in closer touch with its private and public users, and there was a serious need to reverse the reduction of its resources. One result was the creation of the National Statistics Council in 1986, which, with various advisory committees, helped the agency on its projects and priorities.
The second challenge was harder to meet, for it involved a difference between those who wanted to maintain the emphasis on consistent and integrated measurement using market prices and those who believed the newer demands justified the use of less easily quantified approaches for variables that lacked markets. Such variables could be measured, nevertheless, even if more approximately than those that had market prices. Some of the housewife's services might be estimated using the market price for them, as with house-cleaning, or the overall service could be estimated by using the income foregone by giving up paid work. And a variety of nonmonetary indicators could be used to greatly clarify an issue such as the physical health of a society or its environmental degradation. It was only in the 1990s that Canadian and French statisticians saw a way to reconcile the two approaches, namely, through a satellite account to the national accounts wherein both monetized and non-monetized data could be brought to bear on such issues as unpaid labour, volunteer activity and many others on which major new studies appeared. In the early 1990s, the Income and Expenditure Accounts Division became the National Accounts and Environment Division. Another milestone appeared with a clear link to the tripartite accord of 1944 attempting to standardize the national accounts. A ten-year international effort culminated in a huge report in 1993, entitled simply System of National Accounts, with the objective of harmonizing the national accounts. Canada not only played a leading role but was the first OECD country to implement this system, with some modifications.
The sense that Canada played a leading and constructive role in the development and application of the national accounts was underlined by The Economist in both 1991 and 1993 when a panel of international experts ranked Canada first among the ten largest OECD countries. My reading of McDowall's book suggests at least five reasons for this result.
First is the quality of the personnel associated with the agency or in an advisory capacity. A large number of highly qualified and dedicated people were attracted to the national accounts section in particular from other agencies of government and from universities or moved on to such organizations after a period there. It was also blessed by the support of some outstanding individuals from Statistics Canada, as well as well-placed and able people in other departments.
Second, the agency was usually closely and directly influenced by those who set the international agenda, and, after a time, it also greatly advanced that agenda. In the early period, for example, Keynes' influence was strongly present through Bryce, who in turn had equally dedicated colleagues such as John Deutsch. Kuznets, Leontief and Stone were all directly involved with those working in Ottawa.
Third, people like Simon Goldberg were not only insistent on the need to think of the accounts as a system encompassing the workings of the entire economy and also as a way of ensuring statistical consistency; they also shared the passion of Bryce and other leading mandarins to use the accounts and the underlying Keynesian theory as a central part of the approach to a wide variety of policy issues. Thus, the improvements in input-output data in the 1980s were a prelude to the pressure in the 1990s for better provincial data to cope with everything from the regional and extra-regional impact of major projects to the impact of the GST generally and of the harmonized sales tax in three eastern provinces particularly. By 2000, Canada had a core national matrix that matched 750 commodities with 300 industries, which could be utilized on a province-by-province basis. This feat, one of the gains from computerization, yielded a unique degree of regional dissection for any country.
Fourth, the agency was given a high degree of independence in that its personal and company data cannot be revealed. It is also expected to maintain a neutral role in its work. Occasionally a particular report or data set is seen by private groups, political parties, governments or the media as something less than neutral--the stakes are often large, after all--but the agency has been largely free of serious crises on this score.
Fifth and finally, the author and many outside observers see centralization in one federal agency as a major source of its success. There are many gains to such an approach such as the avoidance of unnecessary duplication and the focusing of responsibility. There have been heavy costs on occasions. I am still surprised that the most powerful public servants in Ottawa could not persuade the agency to take seriously the need for a broader approach to national income measure and that a serious competitive threat from the Bank of Canada was apparently necessary to persuade it to do so eventually.
This is a splendidly researched and written volume for the specialist and non-specialist alike. Readers will have preferences for different aspects of the broad historic developments McDowall analyses, which go well beyond national accounts to its Canadian and international economic settings. My preference, reflecting perhaps my work experience with the Dominion Bureau of Statistics in the first half of the 1950s, is for the often dramatic wartime and early post-war periods of establishment and adolescence of the accounts. But I was also immensely taken with the book's epilogue, where the author takes us step by step through the last phases of the preparation and release of the quarterly estimate of Gross National Product early in 2006. Here one gets an insight beyond what comes earlier into the dedication and professionalism of the personnel, and even of the "fun" that many of them see in the work.
There are some small slips or oversights in the book. Garrett Hardin's 1968 article on "The tragedy of the commons" (Science, Volume 162, pages 1243-248) is justifiably famous, but John Dales' book Pollution, Property and Prices in the same year, which proposed a way of addressing this and is now widely recognized, deserves mentioning. Stephen Leacock may have been "perhaps Canada's best-known political economist" (p. 80) but not because of his political economy. The state of statistical and economic knowledge before the national accounts is severely downplayed in several places in the introduction (page 4, for example), something that is given a more balanced presentation in what follows. More reference to non-governmental uses of the national accounts would have been helpful. But these are minor matters. McDowall has produced a fine book that will be of lasting significance.
A. Edward Safarian is professor emeritus of business economics, Rotman School of Management, University of Toronto, and was a statistician in the Dominion Bureau of Statistics from 1950 to 1955.
|Printer friendly Cite/link Email Feedback|
|Author:||Safarian, A. Edward|
|Publication:||Canadian Public Administration|
|Date:||Dec 1, 2008|
|Previous Article:||A Canadian Priorities Agenda: Policy Choices to Improve Economic and Social Well-Being.|
|Next Article:||Administration at the Supreme Court of Canada: challenges and change in the Charter era.|