UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Essays of Canadian productivity and international trade Yu, Emily 2010

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
24-ubc_2010_spring_yu_emily.pdf [ 845.87kB ]
Metadata
JSON: 24-1.0069933.json
JSON-LD: 24-1.0069933-ld.json
RDF/XML (Pretty): 24-1.0069933-rdf.xml
RDF/JSON: 24-1.0069933-rdf.json
Turtle: 24-1.0069933-turtle.txt
N-Triples: 24-1.0069933-rdf-ntriples.txt
Original Record: 24-1.0069933-source.json
Full Text
24-1.0069933-fulltext.txt
Citation
24-1.0069933.ris

Full Text

Essays in Canadian Productivity and International Trade by Emily Yu  B.A., The University of British Columbia, 2000 M.A., The University of British Columbia, 2001  A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF Doctor of Philosophy in The Faculty of Graduate Studies (Economics)  The University of British Columbia (Vancouver) April, 2010 c Emily Yu 2010  Abstract This thesis is a collection of three empirical papers that made use of recent Canadian trade and production data. The first chapter “Productivity Performance of Canada” examines Canada’s productivity and changes in terms of trade 1961-2007. These changes have been mostly favourable and have had the same effect on real income growth as Total Factor Productivity improvements of the business sector of the economy. The framework applied is developed by Diewert, Kohli and Morrison and is based on production theory. We utilised published and unpublished data from the Statistics Canada Multifactor Productivity program, which develops “bottom up” estimates of business sector productivity from industry estimates. However, we use in this chapter a “top down” approach which utilises (adjusted) final demand data to form a business sector output aggregate and thus leads to much higher estimates of TFP growth for Canada than the corresponding Statistics Canada estimates. Finally, the new export and import time series are used to determine the contributions to real income growth of changes in these disaggregated export and import prices over the 47 year period. The second chapter “Business Sector Data on Outputs and Inputs for Canada 1961-2007” details the business sector data used in the first chapter and explains the construction of estimates of Canadian final demand expenditures, business sector labour input, business sector capital stock, primary input tax rates, balancing real rates of return and user costs. We also make some recommendations for possible improvements that Statistics Canada could make to its productivity program. ii  Abstract The third chapter “Does Lobbying Affect Antidumping Case Determinations in Canada?” examines whether the mandate of antidumping legislation in Canada was independent of influences other than those allowed. The relation between antidumping case determinations and various determinants is examined, in particular, whether lobbying activities can influence case determinations. Unlike previous studies that constructed political variable with data that have limited information on policy influence targets, this chapter constructs the variable using the Canadian Lobbyists Registration data with detailed information on lobbyists who had indicated they lobbied for administered protection. The current empirical evidence suggests that the antidumping mandate is not apolitical.  iii  Table of Contents Abstract  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  ii  Table of Contents . . . . . . . . . . . . . . . . . . . . . . . . . . . .  iv  List of Tables  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  vi  List of Figures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  ix  Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . .  x  Dedication  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xii  1 Productivity Performance of Canada  . . . . . . . . . . . . .  1  . . . . . . . . . . . . . . . . . . . . . . . . . . .  1  1.1  Introduction  1.2  Output and Input Aggregates and Conventional Productivity Growth in Canada . . . . . . . . . . . . . . . . . . . . . . . .  1.3  Explaining Real Income Growth Generated by the Canadian Business Sector: the Gross Output Approach . . . . . . . . .  1.4  6 14  Explaining Real Income Growth Generated by the Canadian Business Sector: the Net Output Approach . . . . . . . . . .  29  1.5  Productivity Performance Comparison with Other Countries  50  1.6  The Effects of Changing Real Export and Import Prices on Real Income Growth . . . . . . . . . . . . . . . . . . . . . . .  53  1.7  Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . .  73  1.8  Explaining Real Income Growth with The Translog Approach 76 1.8.1  Introduction . . . . . . . . . . . . . . . . . . . . . . .  76  1.8.2  The Production Theory Framework . . . . . . . . . .  77  iv  Table of Contents 1.8.3  The Translog GDP Function Approach  . . . . . . . .  1.8.4  The Translog GDP Function Approach and Changes  88  in the Terms of Trade . . . . . . . . . . . . . . . . . .  91  1.8.5  The Deflated NDP Translog Approach  95  1.8.6  Sectoral Contributions to Real Income Growth . . . . 101  . . . . . . . .  2 Business Sector Data on Outputs and Inputs for Canada 1961-2007 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109 2.1  Introduction  . . . . . . . . . . . . . . . . . . . . . . . . . . . 109  2.2  Estimates of Canadian Final Demand Expenditures  2.3  Business Sector Labour Input Estimates  . . . . . . . . . . . 152  2.4  Business Sector Capital Stock Estimates  . . . . . . . . . . . 158  2.5  Primary Input Tax Rates, Balancing Real Rates of Return  . . . . . 113  and User Costs . . . . . . . . . . . . . . . . . . . . . . . . . . 174 2.6  Sources of Error . . . . . . . . . . . . . . . . . . . . . . . . . 182  2.7  Recommendations for the Statistics Canada Productivity Program  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 185  3 Does Lobbying Affect Antidumping Case Determinations in Canada?  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 189  3.1  Introduction  3.2  Antidumping and Lobbying . . . . . . . . . . . . . . . . . . . 191  3.3  3.4  . . . . . . . . . . . . . . . . . . . . . . . . . . . 189  Empirical Analysis . . . . . . . . . . . . . . . . . . . . . . . . 199 3.3.1  Economic and Political Determinants . . . . . . . . . 200  3.3.2  Econometric Analysis . . . . . . . . . . . . . . . . . . 208  Concluding Remarks  . . . . . . . . . . . . . . . . . . . . . . 217  Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 220  v  List of Tables 1.1  Prices of Canadian Business Sector Output and Input Aggregates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  1.2  Quantities of Canadian Business Sector Output and Input Aggregates, TFP Levels and TFP Growth Rates . . . . . . .  1.3  33  Net Real Income Generated by the Canadian Business Sector and Real Output and Input Prices . . . . . . . . . . . . . . .  1.9  31  Quantities of Canadian Business Sector Net Output and Input Aggregates, TFP Levels and TFP Growth Rates . . . . .  1.8  27  Prices of Canadian Business Sector Net Output and Input Aggregates . . . . . . . . . . . . . . . . . . . . . . . . . . . .  1.7  22  Business Sector Cumulated Growth in Real Income and Cumulated Contribution Factors . . . . . . . . . . . . . . . . . .  1.6  16  Business Sector Year to Year Growth in Real Income and Year to Year Contribution Factors . . . . . . . . . . . . . . .  1.5  11  Gross Real Income Generated by the Canadian Business Sector and Real Output and Input Prices . . . . . . . . . . . . .  1.4  9  38  Business Sector Year to Year Growth in Net Real Income and Net Year to Year Contribution Factors . . . . . . . . . . . . .  40  1.10 Business Sector Cumulated Growth in Net Real Income and Cumulated Contribution Factors . . . . . . . . . . . . . . . .  48  1.11 Average Productivity Growth in Canada and the United States, 1961-2008 . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  51  1.12 Real Income and TFP Growth of Canada, Japan and Australia 52 1.13 Year to Year Export Contribution Factors Using the Gross Output Approach  . . . . . . . . . . . . . . . . . . . . . . . .  62  vi  List of Tables 1.14 Year to Year Import Contribution Factors Using the Gross Output Approach  . . . . . . . . . . . . . . . . . . . . . . . .  65  1.15 Year to Year Export Contribution Factors Using the Net Output Approach  . . . . . . . . . . . . . . . . . . . . . . . . . .  68  1.16 Year to Year Import Contribution Factors Using the Net Output Approach 2.1  . . . . . . . . . . . . . . . . . . . . . . . . . .  71  Housing Value, Quantity and Price Series for Imputed and Paid Rents . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116  2.2  Prices Indexes for Business Sector Outputs: Consumption and Investment . . . . . . . . . . . . . . . . . . . . . . . . . . 120  2.3  Quantity Indexes for Business Sector Outputs: Consumption and Investment . . . . . . . . . . . . . . . . . . . . . . . . . . 122  2.4  Business Sector, Non-business Sector, Government Final Demand and KLEMS Business Sector Price and Quantity Aggregates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126  2.5  Price Indexes for Business Sector Outputs: Net Sales to the Non-business Sector . . . . . . . . . . . . . . . . . . . . . . . 134  2.6  Price Indexes for Eight Commodity Classes of Exports, 19612007 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141  2.7  Quantity Indexes for Eight Commodity Classes of Exports, 1961-2007 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143  2.8  Price Indexes for Seven Commodity Classes of Imports, 19612007 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 148  2.9  Quantity Indexes for Seven Commodity Classes of Imports, 1961-2007 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 150  2.10 Price and Quantity Indexes for Three Types of Business Sector Labour  . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153  2.11 Beginning of Year Asset Values for Residential Structures and Land and Six Business Sector Capital Stocks . . . . . . . . . 162  vii  List of Tables 2.12 Smoothed Geometric Depreciation Rates for ICT, Non-ICT Machinery and Equipment, Non-residential Structures and Residential Structures Capital Stocks Implied by the Balance Sheets and Investment Flow Data  . . . . . . . . . . . . . . . 167  2.13 Prices for Residential Structures and Land and Six Business Sector Capital Stocks  . . . . . . . . . . . . . . . . . . . . . . 169  2.14 Quantities of Residential Structures and Land and Six Business Sector Capital Stocks . . . . . . . . . . . . . . . . . . . . 171 2.15 Business Sector Tax Rates, Balancing Real Rates of Return and User Costs . . . . . . . . . . . . . . . . . . . . . . . . . . 179 3.1  Antidumping Cases in Canada 1990-2006 . . . . . . . . . . . 194  3.2  Antidumping Cases in Canada by the Named Country’s Economic Status 1990-2004 . . . . . . . . . . . . . . . . . . . . . 195  3.3 3.4  Active Lobbyist Registrations in Canada 1996-2005 . . . . . . 197 Percentage of Lobbyists that Lobbied Canadian International Trade Tribunal 1996-2003 . . . . . . . . . . . . . . . . . . . . 198  3.5  Regression of Lobbying Efforts . . . . . . . . . . . . . . . . . 206  3.6  Expected Signs of Determinants . . . . . . . . . . . . . . . . . 209  3.7  Negative Binomial Regression on Antidumping Case Determinations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 211  3.8  Negative Binomial Regression on Antidumping Case Determinations (Cont’d) . . . . . . . . . . . . . . . . . . . . . . . . 212  3.9  Negative Binomial Regression on Antidumping Petitions . . . 216  viii  List of Figures 1.1  Real Income Change and Terms of Trade Contribution 19622007 (Gross Output Approach) . . . . . . . . . . . . . . . . .  24  1.2  Level of Total Factor Productivity in Canada 1961-2007 . . .  36  1.3  Rate of Productivity Growth in Canada 1962-2007 . . . . . .  37  1.4  Real Income Generated by the Business Sector in Canada 1961-2007 . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  1.5  43  Real Income Change and Terms of Trade Contribution 19622007 (Net Output Approach) . . . . . . . . . . . . . . . . . .  44  1.6  Canadian Export Values 1961-2007 . . . . . . . . . . . . . . .  55  1.7  Canadian Export Prices 1961-2007 . . . . . . . . . . . . . . .  56  1.8  Canadian Export Quantities 1961-2007 . . . . . . . . . . . . .  57  1.9  Canadian Import Values 1961-2007 . . . . . . . . . . . . . . .  58  1.10 Canadian Import Prices 1961-2007 . . . . . . . . . . . . . . .  59  1.11 Canadian Import Quantities 1961-2007 . . . . . . . . . . . . .  60  3.1  Antidumping Cases in Canada by the Named Countries 19902006 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 196  ix  Acknowledgements First and foremost I owe my sincerest gratitude to my thesis supervisor, Professor Brian Copeland, who has supported me throughout my thesis with his advices, guidance and patience. I am extremely grateful to his incredible patience with me, without which this thesis could not have happened. He was also the reason I have chosen my field to be International Trade. I thank him for giving me a direction to the very interesting field of Economics. I would also like to thank my other thesis supervisor, Professor Erwin W. Diewert who constantly assisted me on my research, who gave me plenty of useful suggestions and not to mention the enormous amount of encouragements when I was feeling stressful. He has taught me many many things that I would not have learnt if not for him. The first and second chapters of this thesis are inspired by his earlier work and the results of a joint research with him. I am indebted to Dr Werner Antweiler who served on my thesis committee and gave me many sound advices during the early formation stage of this thesis. I like to thank my first year advisor in the doctoral program, Professor Margaret Slade, who told me there is no limit to what I can do if I dare to dream. Many thanks also to the librarian at Koerner’s Library, Mary Luebbe who assisted me in data search for more than two years. I also like to acknowledge the KLEMS program of Statistics Canada for providing the data used in this thesis. I am grateful to many of the fellow schoolmates who lent me their ears when I needed someone to discuss my work with. I also have been blessed with many wonderful friends, Dr. Edmond Ng, Kenneth Mak, Iris Yeung, Jasmine Poon, Steve (Chubby) Xu, Steve Yong and Markus von Wartburg, to name a few, thanks for the occasional amateur therapy sessions, supports x  Acknowledgements and kind (sometimes harsh) words. Their existences remarkably brightened my life as a graduate student. I wholeheartedly thank my best friend at UBC, Carl Ruest, for spending a lot of his precious time to help me improve the thesis. He was undoubtedly my best lunch companion during all those rainy days at school. Lastly, and most importantly, I wish to thank my parents, Wing Chow Yu and Tsang Dor Sin. They sacrificed a lot to give me the best. Without their sacrifices, I will not be who I am today.  xi  Dedication To my parents and my family, I dedicate this thesis.  xii  Chapter 1  Productivity Performance of Canada 1.1  Introduction  Gains in productivity are an important factor in leading higher living standards. However, the calculation of productivity is not a straightforward matter. Statistics Canada calculates business sector Multifactor Productivity (MFP) growth or Total Factor Productivity (TFP) growth using a “bottom up” approach; i.e., separate TFP growth estimates are made for each major industrial sector in the Canadian business sector and then these sector estimates are aggregated to provide an estimate of total business sector TFP growth. But aggregate TFP growth can also be calculated using a “top down” approach; i.e., instead of aggregating over business sector industry outputs and intermediate inputs, it is possible to use estimates of deliveries to final demand as the measure of aggregate output growth and this aggregate output information can be combined with information on aggregate primary input usage to give an alternative approach to the measurement of aggregate TFP growth. In this chapter, using the “top down” approach and using new data from Statistics Canada, we show that the productivity performance of the Canadian business sector appears to be much better than the lacklustre official Statistics Canada estimates of TFP growth, which are based on the “bottom up” approach.1 This indicates the “productivity gap” between the official estimates by Statistics Canada and 1 However, we will show later that the main source of difference between the two methods for computing TFP growth appears to be in the aggregation of capital services rather than in differences in the aggregation of outputs.  1  Chapter 1. Productivity Performance of Canada the U.S. may not be as wide as we have come to believe. Diewert (2006b) (2007c) had shown that if effects of indirect tax can be ignored and that data from the input output tables are consistent then the “top down” approach used in this chapter and the “bottom up” approach used by Statistics Canada should give similar measures of productivity. The “top down” approach can provide a comparison to the official estimates. Furthermore, it also allows us to estimate individual contribution of export and import commodity classes which is not possible using the “bottom up” approach. Thus, the “top down” approach makes it possible to carry out the calculation of terms of trade contribution to real income growth. In addition to developing new estimates of TFP growth for the Canadian business sector for the years 1962-2007, this chapter will provide new estimates for the growth in real income generated by the Canadian business sector and provide estimates of how various growth factors contributed to this overall real income growth. There are three main factors which explain the growth in real income: • TFP growth; • Growth in primary inputs used by the business sector and • Changes in real output prices. Included in the last explanatory factor are changes in real export and real import prices. We have developed some new time series for disaggregated components of Canadian exports and imports so that the gains in real income generated by increasing real export prices or decreasing real import prices can be traced back to particular classes of exports and imports. Note that this type of analysis cannot be done using the “bottom up” approach to TFP measurement because the industry input output tables do not have an accurate breakdown of industrial outputs into exports and deliveries to domestic demanders or a breakdown of intermediate inputs into imports and domestically produced intermediates. 2  Chapter 1. Productivity Performance of Canada When economists measure TFP growth, the output concept is almost always the gross product produced by the sector or economy and a traditional Jorgenson and Griliches (1967) user cost of capital is used to measure the contribution of capital inputs. The traditional user cost of capital has three main components: • A nominal interest rate component; • A depreciation component and • A (negative) price appreciation term. Sometimes the first and third components listed above are combined into a single real interest rate component. The important points to note are that (i) depreciation appears in the user cost of capital and (ii) a gross output concept is used in the traditional approach to the measurement of TFP growth. However, depreciation is not a source of income; households cannot consume depreciation. Depreciation should be treated as a charge against income instead of a component of income. Thus instead of using a gross output approach to the measurement of output and a user cost of capital that includes depreciation, when attempting to measure the income generated by the private production sector, the depreciation component in the user cost should be removed and treated as an (intertemporal) intermediate input, leading to a net output approach to the measurement of output and a waiting services approach to the user cost of capital. We will implement both the traditional approach to TFP measurement as well as the net output approach and compare the two approaches using Canadian business sector data over the period 1961-2007. As noted above, it is not easy to measure the exact magnitude of productivity gains or of terms of trade. Diewert (1983), Diewert and Morrison (1986), Diewert, Mizobuchi and Nomura (2005), Diewert and Lawrence (2006), Morrison and Diewert (1990) and Kohli (1990, 1991, 2003, 2004a, 2004b, 2006, 2008) have developed methodologies based on production theory that allow the contribution of each type of gain be represented by an 3  Chapter 1. Productivity Performance of Canada index number estimate. In Section 1.8, we outline the Diewert, Mizobuchi and Nomura (2005) and Diewert and Lawrence (2006) methodology and show how it can be used to measure the determinants of growth of an economy’s gross and net real income. In Sections 1.2-1.4, we apply this same methodology to the business sector of the Canadian economy over the years 1962-2007. The details of how the Canadian business sector data was developed from Statistics Canada sources are described in Chapter 2 “Business Sector Data on Outputs and Inputs for Canada 1961-2007”. Section 1.2 aggregates up the data from Chapter 2 and develops conventional measures of Canadian business sector Total Factor Productivity (TFP) for the years 1961-2007. However, productivity growth, while perhaps the most important source of growth in living standards, is not the entire story. If a country’s export prices increase more rapidly than its import prices, then it is well known that this has an effect that is similar to a productivity improvement.2 Thus in Section 1.3, we measure the relative contributions of productivity improvements, changes in real export and import prices and growth of labour and capital inputs to the growth of (gross) real income generated by the business sector in Canada using the methodology explained in Subsections 1.8.2-1.8.5. This is still not the end of the story; GDP is an imperfect measure of productive potential, not welfare.3 For welfare measurement purposes, it is generally conceded that Net Domestic Product (NDP) is a better measure of output, since investment that just equals depreciation means that society is not made any better off from the viewpoint of sustainable final consumption possibilities. Hence in Section 1.4, we subtract depreciation off from gross investment and use consumption plus sales to the non-business sector plus net investment plus the trade balance as our business sector output concept. Thus depreciation will be treated as an intermediate input in this production model. 2  See for example Diewert and Morrison (1986). For a more extensive discussion of the merits of GDP versus net income, see Diewert (2006a). 3  4  Chapter 1. Productivity Performance of Canada Subsection 1.8.6 explains this real net output approach and adapts a translog model of production based on the work of Diewert and Morrison (1986) and Kohli (1990) to this new model of market sector real net income generation.4 This approach is implemented for the Canadian business sector in Section 1.4. The main determinants of growth in real net income generated by the business or market sector of the economy are: • Technical progress or improvements in Total Factor Productivity; • Growth in domestic output prices or the prices of internationally traded goods and services relative to the price of consumption and • Growth in primary inputs. It turns out that productivity growth becomes a more important factor for explaining real net income growth compared to explaining real gross income growth. Also the importance of capital deepening is greatly reduced in the net output framework compared to the gross output framework. Somewhat surprisingly, for the years 2000-2007, improvements in the terms of trade made almost the same contribution to real income growth as capital deepening in the gross output framework and in the net output framework, the effects of falling real import prices contributed substantially more to real income growth than capital deepening over the period 2000-2007. Section 1.5 discusses how our estimates compare to the productivity estimates of other countries. We compare our Canadian real income growth estimates using the “top down” approach with the estimations that were done for Australia and Japan. It can be seen that the overall average income growth of the three countries are similar, but the productivity performances are very different. Canada had benefited from terms of trade improvement and thus even without a stellar productivity performance, the real income growth was as much as the two other countries that have relatively higher productivity growths. Then in Section 1.6, we look at changes in real export and import 4 For previous implementations of this model of real net income to Japan and Australia, see Diewert, Mizobuchi and Nomura (2005) and Diewert and Lawrence (2006).  5  Chapter 1. Productivity Performance of Canada prices by commodity class and their effects on real income growth in both the gross output and the net output frameworks. These new results are somewhat surprising. Finally Section 1.7 concludes.5  1.2  Output and Input Aggregates and Conventional Productivity Growth in Canada  In Chapter 2, we constructed price and quantity series for 23 net outputs and 9 primary inputs for the business sector of the Canadian economy for the years 1961-2007. The 23 net outputs are, • Q1 , Domestic consumption (excluding market residential rents and the services of owner occupied housing); • Q2 , Real sales of goods and services by the business sector to the nonmarket sector less real sales of goods and services from the non-market sector to the business sector; • Q3 , Government investment; • Q4 , Business sector investment in residential structures; • Q5 , Business sector investment in information and communication technology (ICT) machinery and equipment; • Q6 , Business sector investment in non-ICT machinery and equipment; • Q7 , Business sector investment in non-residential structures; • Q8 , Inventory change; • Q9 , Exports of agricultural and fish products; • Q10 , Exports of energy products; 5 The final part of Section 1.8 has some new material on how the real net output model used in the present paper can be extended from a single production sector to the case of many industries.  6  Chapter 1. Productivity Performance of Canada • Q11 , Exports of forest products; • Q12 , Exports of industrial goods and materials (excluding energy and forest product exports); • Q13 , Exports of machinery and equipment (excluding automotive products); • Q14 , Exports of automotive products; • Q15 , Exports of other consumer goods (excluding automotive products); • Q16 , Exports of services; • Q17 , Imports of agricultural and fish products; • Q18 , Imports of energy products; • Q19 , Imports of industrial goods and materials (including imports of forest products but excluding imports of energy products); • Q20 , Imports of machinery and equipment (excluding automotive products); • Q21 , Imports of automotive products; • Q22 , Imports of other consumer goods and • Q23 , Import of services. The nine primary inputs into the business sector are, • Q24 , The labour services of workers with primary or secondary education; • Q25 , The labour services of workers with some or completed secondary certificate or diploma;  7  Chapter 1. Productivity Performance of Canada • Q26 , The labour services of workers with a university degree or above;6 • Q27 , The stock of ICT machinery and equipment available to the business sector at the start of each year; • Q28 , The stock of non-ICT machinery and equipment available to the business sector at the start of each year; • Q29 , The starting stock of business sector non-residential structures; • Q30 ,The starting stocks of inventories used by the business sector; • Q31 , The stock of agricultural land use by the business sector and • Q32 , The stock of non-agricultural, non-residential land used by the business sector. As explained in Chapter 2, user cost prices for the last six primary inputs were constructed using balancing or endogenous real rates of return that made the value of net output produced by the business sector equal to the value of primary inputs used by the business sector. Details of the construction are given in Chapter 2 where the price and quantity series constructed from the 23 net outputs and 9 primary inputs are also shown. In this section, we will aggregate the above net outputs and primary inputs into D, domestic output, equal to an aggregate of the first eight net outputs; X, exports equal to an aggregate of the eight types of exports of goods and services; M , imports equal to an aggregate of the seven types of imports of goods and services; L, labour services equal to an aggregate of the three types of labour services; K, capital services equal to an aggregate of the six types of capital services. Once these aggregates have been constructed, we further aggregate the three net outputs D + X − M into real gross domestic product Y and aggregate the two inputs K and L into domestic input Z and finally construct a conventional measure of productivity Y /Z. 6  These three types of labour services input data are taken directly from Statistics Canada KLEMS program; see Baldwin, Gu and Yan (2007) for description.  8  Chapter 1. Productivity Performance of Canada The aggregations were done using chained T¨ornqvist price indexes,7 and the resulting prices and quantities are shown in Tables 1.1 and 1.2.  Table 1.1: Prices of Canadian Business Sector Output and Input Aggregates  Year  t PC  t PD  t PX  t PM  t PL  1961  1.00000  1.00000  1.00000  1.00000  1.00000  1962  1.00538  1.00538  1.02992  1.05787  1963  1.02055  1.02264  1.04054  1964  1.02437  1.02959  1965  1.03690  1.05092  1966  1.07553  1967  t PK  t PY  t PZ  1.00000  1.00000  1.00000  1.03625  1.01691  0.99683  1.03034  1.09648  1.06604  1.15494  1.00691  1.09305  1.05692  1.10526  1.11049  1.23317  1.01577  1.14776  1.07980  1.10214  1.18272  1.25857  1.04400  1.20580  1.08861  1.12451  1.11930  1.25929  1.34193  1.08930  1.28444  1.11050  1.12289  1.15421  1.14426  1.33342  1.23578  1.12505  1.30269  1968  1.15168  1.15976  1.20119  1.16726  1.41620  1.30595  1.16924  1.38147  1969  1.18980  1.19995  1.23088  1.19648  1.52197  1.34691  1.20982  1.46616  1970  1.22208  1.23858  1.26710  1.21965  1.61278  1.40223  1.25289  1.54532  1971  1.24828  1.28233  1.28552  1.24798  1.72528  1.39719  1.29339  1.61828  1972  1.29847  1.34008  1.33325  1.27498  1.86392  1.49709  1.35869  1.74411  1973  1.38744  1.44587  1.51489  1.35954  2.03837  2.06511  1.49788  2.05198  1974  1.58382  1.65206  1.91283  1.64641  2.35044  2.47332  1.73334  2.39718  1975  1.82198  1.87090  2.16694  1.89029  2.70332  2.21008  1.95274  2.53719  1976  1.90726  1.97372  2.29702  1.92853  3.10646  2.39955  2.09001  2.86510  1977  2.03175  2.09967  2.50231  2.17241  3.38889  2.74058  2.19373  3.17107  1978  2.19264  2.26119  2.73837  2.41667  3.53495  3.09969  2.34654  3.39783  1979  2.40645  2.48060  3.20786  2.73027  3.78520  3.80986  2.60728  3.82051  1980  2.69497  2.75772  3.73464  2.97957  4.11781  3.97428  2.97313  4.09308  1981  2.95335  3.03059  3.99821  3.26618  4.59295  4.05499  3.23724  4.42379  1982  3.22860  3.28484  4.08926  3.43918  5.02021  3.61928  3.46983  4.51822  Continued on Next Page. . . 7 More specifically, the chained Divisia option in Shazam was used to do the aggregations.  9  Chapter 1. Productivity Performance of Canada Table 1.1 – Continued t PY  t PZ  4.44415  3.68247  4.97537  5.48101  4.94318  3.81283  5.33580  3.67711  5.75670  5.17976  3.91833  5.59935  3.74437  5.90250  5.08340  3.98140  5.65108  4.45792  3.69150  6.11054  5.74526  4.14995  6.03889  4.01307  4.47080  3.60108  6.51242  5.76285  4.30933  6.29679  4.11690  4.12618  4.56005  3.59364  6.78864  5.58998  4.47092  6.40227  1990  4.35206  4.28654  4.52868  3.64367  7.04667  5.31935  4.60964  6.45499  1991  4.59099  4.42296  4.37107  3.57992  7.34245  4.50786  4.72731  6.30363  1992  4.65258  4.47108  4.49573  3.72567  7.48023  4.98747  4.75269  6.58795  1993  4.74252  4.55596  4.69389  3.92460  7.45961  5.15125  4.82121  6.64462  1994  4.77089  4.62372  4.97322  4.16089  7.41314  6.03823  4.89058  6.98418  1995  4.79147  4.65787  5.29132  4.27688  7.53622  6.56286  5.02353  7.27300  1996  4.88952  4.71581  5.32097  4.22185  7.63205  6.99768  5.13502  7.50733  1997  4.96547  4.77523  5.32718  4.23677  7.91076  7.00934  5.19090  7.68381  1998  5.03224  4.84195  5.31558  4.37960  8.13918  6.93415  5.15347  7.79482  1999  5.12045  4.91274  5.37870  4.36044  8.33890  7.36196  5.27960  8.08916  2000  5.25425  5.02564  5.71039  4.44227  8.74780  8.33839  5.54730  8.73198  2001  5.40970  5.14618  5.80311  4.58416  8.97770  8.04237  5.62691  8.75403  2002  5.47743  5.22044  5.68895  4.61494  9.09489  8.53655  5.61217  9.02599  2003  5.61543  5.29558  5.64768  4.31493  9.26253  8.32732  5.87497  9.04388  2004  5.69551  5.37836  5.78629  4.20643  9.48779  9.25929  6.12630  9.55745  2005  5.81654  5.49862  5.95274  4.15511  9.84265  9.80219  6.39760  9.99376  2006  5.92386  5.63197  5.97000  4.12736  10.29074  9.81133  6.58063  10.27222  2007  6.02712  5.75706  6.02590  4.02177  10.66121  6.84476  10.63228  Year  t PC  t PD  t PX  t PM  t PL  1983  3.46323  3.46344  4.15051  3.42324  5.22085  1984  3.61506  3.60128  4.29656  3.58225  1985  3.72257  3.71138  4.38071  1986  3.80422  3.80294  4.37060  1987  3.89726  3.90844  1988  4.00205  1989  t PK  10.14050 5  10  Chapter 1. Productivity Performance of Canada Table 1.2: Quantities of Canadian Business Sector Output and Input Aggregates, TFP Levels and TFP Growth Rates  QtD  QtX  QtM  QtK  QtY  QtZ  Tt  1961  28752  6867  −7897  19202  8520  27722  27722  1.00000  −  1962  30578  7195  −8033  20042  8739  29750  28782  1.03361  1.03361  1963  32261  7832  −8031  20574  9007  32113  29582  1.08555  1.05024  1964  34602  9105  −8989  21446  9324  34766  30768  1.12995  1.04090  1965  37904  9418  −10180  22416  9751  37149  32164  1.15499  1.02216  1966  40831  10696  −11579  23550  10328  39949  33880  1.17914  1.02091  1967  41117  11827  −12306  24056  11057  40656  35112  1.15790  0.98198  1968  42835  12910  −13527  24158  11627  42246  35756  1.18152  1.02040  1969  46062  13802  −15377  24718  12059  44521  36737  1.21189  1.02571  1970  46017  15211  −15293  24798  12569  45988  37285  1.23341  1.01776  1971  48326  15929  −16480  25333  13007  47844  38238  1.25120  1.01442  1972  52098  17257  −18892  26101  13416  50590  39410  1.28367  1.02595  1973  59241  19008  −21754  27591  13866  56663  41362  1.36992  1.06719  1974  65163  18347  −23977  28558  14615  59579  43080  1.38298  1.00953  1975  63453  16951  −23228  28530  15571  57119  43961  1.29930  0.93949  1976  67488  18390  −24774  28499  16309  61084  44559  1.37086  1.05508  1977  70950  19678  −24836  28805  17018  65759  45492  1.44552  1.05446  1978  73085  21544  −26197  30018  17690  68588  47367  1.44801  1.00173  1979  78467  22467  −28092  31737  18344  72879  49736  1.46532  1.01195  1980  77309  22548  −28715  32833  19285  71254  51757  1.37669  0.93951  1981  80810  23012  −30716  33723  20148  73083  53480  1.36653  0.99262  1982  70930  22882  −25710  32059  21330  68632  52707  1.30214  0.95288  1983  75524  24326  −28444  32283  21742  72008  53296  1.35109  1.03759  1984  80477  28444  −33270  33497  22102  76807  54884  1.39943  1.03578  1985  85491  29938  −35548  34871  22585  81087  56743  1.42902  1.02114  1986  88667  31456  −37965  36384  23167  83519  58843  1.41937  0.99325  1987  94688  32933  −39889  38166  23747  89072  61210  1.45517  1.02522  Year t  QtL  τt  Continued on Next Page. . .  11  Chapter 1. Productivity Performance of Canada Table 1.2 – Continued QtZ  Tt  τt  93146  63746  1.46120  1.00414  25638  94280  65839  1.43198  0.98000  41030  26721  93557  66811  1.40032  0.97789  39707  27455  87853  65884  1.33345  0.95224  −51473  39311  27831  91077  65705  1.38615  1.03952  45382  −55461  40184  28050  92144  66858  1.37821  0.99427  103227  51076  −60606  41745  28099  97970  68602  1.42809  1.03619  1995  105681  55452  −64385  42958  28426  101581  70163  1.44779  1.01380  1996  108585  58646  −67340  44212  28923  105126  71906  1.46199  1.00981  1997  116675  63457  −77378  45636  29440  109300  73839  1.48025  1.01249  1998  120947  69086  −81755  47078  30519  115417  76307  1.51254  1.02182  1999  124963  76337  −88261  48781  31631  121154  79074  1.53216  1.01297  2000  132088  83350  −95661  50512  32736  128862  81864  1.57409  1.02737  2001  132099  80654  −90649  51183  33919  130142  83653  1.55574  0.98834  2002  140327  81599  −92347  52290  34561  137310  85376  1.60829  1.03377  2003  142198  79268  −96341  53129  35172  133617  86799  1.53939  0.95716  2004  150705  83281  −104558  55049  35675  139173  89210  1.56007  1.01343  2005  158450  84730  −112492  55875  36549  141962  90878  1.56211  1.00131  2006  165871  85022  −117772  56905  37720  145225  93035  1.56098  0.99928  2007  173734  86002  −124556  58347  38998  148654  95699  1.55335  0.99511  Year t  QtD  QtX  1988  101144  35371  1989  104645  1990 1991  QtM  QtL  QtK  −45163  39912  24550  35434  −47820  40981  102201  37556  −48551  96067  38167  −49281  1992  98558  40921  1993  98528  1994  QtY  12  Chapter 1. Productivity Performance of Canada Note that we have also listed the price of our household consumption aggregate, PCt , in Table 1.1, which will play a role in the subsequent sections. The productivity level in year t of the Canadian business sector, T t , can be defined as the aggregate year t output, QtY divided by the aggregate year t  input, QtZ ,8  Tt ≡  QtY QtZ  t = 1961, . . . , 2007  (1.1)  Productivity growth for year t, τ t , is defined as the productivity level in year t, T t , divided by the productivity level from the previous year, τt ≡  Tt T t−1  t = 1962, . . . , 2007  (1.2)  Table 1.2 lists the quantities that match up to the prices in Table 1.1 and it also lists productivity levels and growth rates. The average rate of total factor productivity (TFP) growth over the 46 years (1962-2007) is 1.01% per year,9 which is much higher than the 0.5 to 0.7% per year range that Diewert and Lawrence (2000) found over the period 1962-1996. The present 1.01% average rate of TFP growth can also be compared with Statistics Canada’s recent KLEMS program average Multifactor Productivity Growth over the same years of 0.38% per year,10 which is a rather substantial difference.11 8  This is known as Multifactor Productivity or Total Factor Productivity. This rate of TFP growth is reasonably close to the average rate of productivity growth of Australia obtained by Diewert and Lawrence (2006) using a similar methodology and over a similar period. The Diewert and Lawrence market sector average rate of TFP growth for Australia over the period of 1961-2004 was 1.49% per year. However, there is an upward bias in the Diewert and Lawrence results due to the fact that they essentially used hours worked as their measure of labour input instead of a quality adjusted measure of labour input for Australia (which was not available). 10 See CANSIM II series V41712881, Canada, Multifactor Productivity, Business Sector, Table 3830021, Multifactor Productivity, Value Added, Capital Input and Labour Input in the Aggregate Business Sector and Major Sub-Sectors. Comparing levels of TFP with the starting level being 1 in 1961, our TFP ended up at 1.553 in 2007 whereas KLEMS Multifactor Productivity ended up at 1.184 in 2007. This is a very substantial difference. 11 Our measures of business sector output and capital input were different from the KLEMS measures because we excluded rental housing from our measure of capital services, whereas the KLEMS measures included rental housing in their output and capital input measures. Our measures of labour input were identical and it turned out that the 9  13  Chapter 1. Productivity Performance of Canada During the golden years of 1962-1973, Canada’s TFP growth averaged about 2.68% per year; over the dismal years of 1974-1991, the average TFP growth was essentially 0 (−0.086 per year); TFP growth nicely recovered during 1992-1999 to an average of 1.76% per year. Finally, from 2000-2007, the average TFP growth fell to only 0.20% per year. There were two years of poor productivity growth, 2001 and 2003, where drops of 1.17% and 4.28% occurred. After some increases in productivity growth in 2004 and 2005, slight drops in productivity growth again occurred in 2006 and 2007. If the two poor-performing years are excluded, the productivity growth during this period is on average 1.4% per year. Productivity growth does not necessarily represent the whole story behind the growth in living standards. If the prices of Canadian exports increases more rapidly than the prices of Canadian imports, then the real income generated by the business sector increases. This terms of trade effect is not taken into account in the conventional productivity growth computation. Thus in the following section, we implement the translog real income methodology explained in Subsections 1.8.2-1.8.5 and this approach will enable us to assess the contribution to Canadian living standards of improvements in Canada’s terms of trade.  1.3  Explaining Real Income Growth Generated by the Canadian Business Sector: the Gross Output Approach  The basic methodology used in this section can easily be explained in non-technical terms. The business sector faces exogenous domestic and international prices for the net outputs it produces: domestic outputs, exports and (minus) imports. It also utilises inputs of labour and capital in order to produce its outputs. The value of outputs produced by the business sector average rate of growth of our business sector value added measure was very close to the corresponding KLEMS average growth rate. The capital services growth rates differed substantially.  14  Chapter 1. Productivity Performance of Canada less the value of imports used (value added) must eventually flow back to the labour and capital primary inputs that were used to produce the value added. This is the gross income generated by the business sector. In order to turn this into real income ρt , we divide the gross income in year t by the price of consumption in year t, PCt . This real income is the number of consumption bundles that could be purchased by the owners of the labour and capital inputs that were used in year t by the Canadian business sector. We also divide each of the prices of domestic output, export, import, labour t , P t , P t and P t ) by the price of consumpand capital services (PDt , PX M L K  tion, PCt , to form the corresponding real output and input prices facing the Canadian business sector in each year. Our measures of gross real income generated by the business sector, ρt and the corresponding real output and input prices are shown in Table 1.3.  15  Chapter 1. Productivity Performance of Canada Table 1.3: Gross Real Income Generated by the Canadian Business Sector and Real Output and Input Prices  Year t  ρt  t PD Pt C  t PX Pt C  t PM Pt C  t PL Pt C  t PK Pt C  1961  27722  1.00000  1.00000  1.00000  1.00000  1.00000  1962  29497  1.00000  1.02441  1.05221  1.03070  1.01147  1963  31684  1.00205  1.01959  1.07440  1.04457  1.13169  1964  34474  1.00509  1.03178  1.07897  1.08407  1.20383  1965  37404  1.01352  1.04137  1.06292  1.14063  1.21379  1966  40460  1.01216  1.04554  1.04070  1.17085  1.24769  1967  41188  1.01115  1.03936  1.03040  1.20074  1.11281  1968  42890  1.00701  1.04299  1.01353  1.22968  1.13395  1969  45270  1.00853  1.03453  1.00561  1.27918  1.13205  1970  47147  1.01350  1.03684  0.99801  1.31970  1.14742  1971  49573  1.02727  1.02983  0.99976  1.38213  1.11929  1972  52936  1.03205  1.02679  0.98191  1.43547  1.15296  1973  61173  1.04211  1.09186  0.97989  1.46916  1.48843  1974  65204  1.04308  1.20773  1.03952  1.48403  1.56162  1975  61218  1.02685  1.18933  1.03749  1.48373  1.21301  1976  66937  1.03485  1.20436  1.01115  1.62875  1.25811  1977  71002  1.03343  1.23160  1.06923  1.66796  1.34888  1978  73403  1.03126  1.24889  1.10217  1.61219  1.41368  1979  78961  1.03081  1.33303  1.13456  1.57294  1.58319  1980  78608  1.02328  1.38578  1.10560  1.52796  1.47470  1981  80108  1.02615  1.35379  1.10592  1.55517  1.37301  1982  73760  1.01742  1.26657  1.06522  1.55492  1.12101  1983  76566  1.00006  1.19845  0.98845  1.50751  1.28324  1984  81009  0.99619  1.18852  0.99092  1.51616  1.36739  1985  85351  0.99699  1.17680  0.98779  1.54643  1.39145  1986  87409  0.99966  1.14888  0.98427  1.55157  1.33625  1987  94847  1.00287  1.14386  0.94720  1.56791  1.47418  Continued on Next Page. . .  16  Chapter 1. Productivity Performance of Canada Table 1.3 – Continued Year t  ρt  t PD Pt C  t PX Pt C  t PM Pt C  t PL Pt C  t PK Pt C  1988  100298  1.00275  1.11713  0.89981  1.62727  1.43997  1989  102387  1.00225  1.10764  0.87290  1.64897  1.35781  1990  99095  0.98495  1.04058  0.83723  1.61916  1.22226  1991  90462  0.96340  0.95210  0.77977  1.59932  0.98189  1992  93036  0.96099  0.96629  0.80078  1.60776  1.07198  1993  93673  0.96066  0.98975  0.82753  1.57292  1.08618  1994  100428  0.96915  1.04241  0.87214  1.55383  1.26564  1995  106501  0.97212  1.10432  0.89260  1.57284  1.36970  1996  110404  0.96447  1.08824  0.86345  1.56090  1.43116  1997  114262  0.96169  1.07285  0.85325  1.59315  1.41162  1998  118198  0.96219  1.05630  0.87031  1.61741  1.37794  1999  124920  0.95943  1.05044  0.85157  1.62855  1.43776  2000  136049  0.95649  1.08681  0.84546  1.66490  1.58698  2001  135368  0.95129  1.07272  0.84740  1.65956  1.48666  2002  140688  0.95308  1.03862  0.84254  1.66043  1.55850  2003  139793  0.94304  1.00574  0.76841  1.64948  1.48294  2004  149700  0.94432  1.01594  0.73855  1.66584  1.62572  2005  156144  0.94534  1.02342  0.71436  1.69218  1.68523  2006  161326  0.95073  1.00779  0.69673  1.73717  1.65624  2007  168821  0.95519  0.99980  0.66728  1.76887  1.68249  17  Chapter 1. Productivity Performance of Canada The results show that the gross real income generated by the Canadian business sector has grown from $27,722 million dollars worth of 1961 consumption bundles in 1961 to $168,821 in 2007, a 6.09 fold increase. The real price of domestic output has fallen to 0.9552 times the starting level (due to the fact that machinery and equipment prices have risen less rapidly than the prices of consumption) and the real price of exports has fallen slightly to 0.9998 times of the starting level while the real price of imports has fallen substantially to 0.6673 times the starting level. The quality adjusted real wage of business sector workers have risen to 1.77 times their initial 1961 levels. The real price of capital services has risen 1.68 fold, this reflects rapidly rising prices of agricultural land and non-agricultural business land as well as upward trends in machinery and equipment depreciation rates and in real rates of return. The details are discussed in Chapter 2.12 There are six quantitative factors that can be used to explain the real income ρt generated by the business sector in year t: 1. The price of domestic production (an aggregate of C + I + G) relative to the price of consumption in year t,  t PD t ; PC  2. The price of exports relative to the price of consumption in year t,  t PX t ; PC  3. The price of imports relative to the price of consumption in year t, t PM t ; PC  4. The quantity of labour used by the business sector in year t, QtL ; 5. The quantity of capital used by the business sector in year t, QtK and 6. The level of technology of the business sector in year t. 12 The volatility of the real price of capital services reflects the fact that we have used balancing real rates of return in our user costs and these real rates are subject to a considerable amount of measurement error. One would expect the aggregate real price of capital services to decline, reflecting the decline in the real price of machinery and equipment, but this decline is offset by a large increase in the real price of land services.  18  Chapter 1. Productivity Performance of Canada The formal model outlined in Section 1.8, based on the work of Diewert and Morrison (1986) and Kohli (1990), allows us to decompose the growth ρt , into multiplicative ρt−1 t t t t t αD , αX , αM , βL , βK and τ t that describe  of real income from year t − 1 to t,  year to year  contribution factors  the effects of  changes in the six quantitative factors listed above going from year t − 1 to t. The following equation which decomposes the year to year growth in real income generated by the business sector, year explanatory contribution  ρt , ρt−1  into a product of six year to  factors:13  ρt t t t t = τ t αD αX αM βLt βK ρt−1  t = 1962, 1963, . . . , 2007.  (1.3)  t is greater than one, this means that the domestic price of output Thus if αD t grew faster than the price of consumption going from year t − 1 to t and αD  measures the contribution of rising real domestic output prices to the growth t is greater than one, this means Canadian in real income. Similarly, if αX  export prices grew faster than the price of consumption going from year t measures the contribution of rising real export prices t − 1 to t and αX  to the growth in real income generated by the Canadian business sector. t is larger than one, this means the Canadian import prices However, if αM  did not increase as quickly as the price of consumption going from year t − 1 t measures the contribution of falling real import prices to the to t and αM  growth in real income generated by the Canadian business sector. If βLt  is larger than one, then the labour input in the business sector increased going from year t − 1 to t and βLt measures the contribution of the increase in labour input to the growth in real income generated by the Canadian t , if it is greater than one, then the business business sector. Similarly for βK  t sector capital service input increased going from year t − 1 to t and βK  measures the contribution of the increase in capital input to the growth in real income generated by the business sector. Finally, if τ t is larger than one, then the efficiency of the Canadian business sector increased from year t − 1 13  See Equations (1.50), (1.59) and (1.64) in Section 1.8 in order to derive this equation. All of the variables in Equation (1.3) can be identified using data in Chapter 2.  19  Chapter 1. Productivity Performance of Canada to t and τ t measures the contribution of the efficiency increase to the growth in real income generated by the Canadian business sector.14 The year to year contribution factors and the averages of them are listed in Table 1.4. The periodic averages of the year to year growth in real income show that the gross real income generated by the Canadian business sector over the entire sample period grew at 4.10 percent per year over the 47 years (1961-2007). The biggest contributor to this growth was the growth of quality adjusted labour input at 1.60 percentage points per year. Next was capital services input, which contributed on average 1.13 percentage points per year, followed by TFP growth (1.01 percentage points per year) and declines in real import prices (0.43 percentage points per year). Declines in real domestic output prices and real export prices give rise to negative average contribution factors, −0.09 and −0.03 percentage points per year respectively. The last column in Table 1.4 gives the product of the real t , which is defined as, export and real import price contribution factors, αXM t t t αXM ≡ αX αM  (1.4)  t is the terms of trade contribution factor, it gives Roughly speaking, αXM  the contribution to real income growth of the combined effects of real changes in the international prices faced by the Canadian business sector.15 It can be seen that the effects of changing real international prices are not negligible for Canada: on average, changing real export and import prices contributed 0.38 percentage points per year to real income growth over the entire samThe productivity growth rates, τ t , computed here do not completely agree with the ones computed in the last section. The discrepancy arises because the input aggregates in calculating τ t here is a direct T¨ ornqvist quantity index whereas an implicit quantity index is used in the earlier computation. 15 Ulrich Kohli has pointed out that this is a slight abuse of terminology. Strictly speaking, the terms of trade is the price of exports over the price of imports and hence involves t only two prices. Our definition of αXM involves three prices: the price of exports, the price of imports and the price of domestic consumption. Our terms of trade contribution factor is the rate of change counterpart to Kohli’s(2006; 50) trading gains factor. 14  20  Chapter 1. Productivity Performance of Canada ple period.16 However, for shorter periods, the effects of changing real international prices can be far more important in explaining changes in the real income generated by the market sector of the economy. If the attention is restricted to the recent years (2000-2007), it can be seen that the improvements in Canadian terms of trade become larger than the average contribution of capital deepening. During this period, the average annual growth in the real income generated by the Canadian business sector was 3.88% per year, which can be explained by the following factors: decreases in the real price of imports (1.63 percentage points), increases in quality adjusted labour input (1.40 percentage points), increases in capital services input (1.01 percentage points) and improvements in TFP (0.20 percentage points). There were also small negative contributors to market sector real income growth during the naughts: decreases in the real price of domestically produced goods and services (−0.05 percentage points) and decreases in real prices of exports (−0.34 percentage points). Thus decreases in the real price of imports proved to be the most important factor in explaining the growth in real income generated by the market sector during this period. Overall, the joint effects of changes in real export and import prices contributed about 1.28 percentage points per year on average to the growth of market sector real income during this period, which was larger than the capital deepening contribution of 1.01 percentage points per year.17  16 Thus the contribution of falling real import prices outweighs the effects of the falling export prices. 17 These results are very similar to the results obtained for Australia using a similar framework by Diewert and Lawrence (2006); i.e. both Australia and Canada have had very favourable changes in their terms of trade in recent years which contributed greatly to real income growth during the naughts.  21  Chapter 1. Productivity Performance of Canada Table 1.4: Business Sector Year to Year Growth in Real Income and Year to Year Contribution Factors  Year t  ρt ρt−1  τt  αtD  αtX  αtM  t βL  t βK  αtXM  1962  1.06402  1.03361  1.00000  1.00602  0.98556  1.03028  1.00773  0.99150  1963  1.07414  1.05024  1.00210  0.99882  0.99418  1.01820  1.00943  0.99301  1964  1.08806  1.04090  1.00308  1.00312  0.99883  1.02848  1.01127  1.00194  1965  1.08499  1.02216  1.00854  1.00248  1.00428  1.03048  1.01448  1.00677  1966  1.08172  1.02091  0.99862  1.00108  1.00622  1.03428  1.01841  1.00730  1967  1.01799  0.98199  0.99899  0.99830  1.00302  1.01478  1.02126  1.00131  1968  1.04133  1.02040  0.99587  1.00107  1.00519  1.00296  1.01535  1.00627  1969  1.05548  1.02571  1.00153  0.99744  1.00259  1.01608  1.01117  1.00003  1970  1.04146  1.01776  1.00497  1.00073  1.00253  1.00223  1.01266  1.00326  1971  1.05145  1.01443  1.01352  0.99775  0.99943  1.01508  1.01033  0.99717  1972  1.06785  1.02595  1.00468  0.99901  1.00617  1.02134  1.00911  1.00518  1973  1.15561  1.06713  1.00988  1.02092  1.00072  1.03876  1.01043  1.02166  1974  1.06589  1.00953  1.00096  1.03484  0.97865  1.02287  1.01826  1.01275  1975  0.93887  0.93953  0.98361  0.99488  1.00076  0.99933  1.02109  0.99563  1976  1.09342  1.05508  1.00821  1.00415  1.00992  0.99926  1.01435  1.01412  1977  1.06072  1.05446  0.99858  1.00755  0.97932  1.00735  1.01349  0.98671  1978  1.03382  1.00172  0.99784  1.00495  0.98842  1.02793  1.01294  0.99331  1979  1.07573  1.01194  0.99955  1.02461  0.98852  1.03661  1.01293  1.01285  1980  0.99553  0.93951  0.99259  1.01519  1.01049  1.02181  1.01843  1.02584  1981  1.01907  0.99262  1.00286  0.99086  0.99988  1.01743  1.01559  0.99074  1982  0.92076  0.95289  0.99143  0.97430  1.01502  0.96690  1.01927  0.98894  1983  1.03805  1.03756  0.98324  0.97884  1.02800  1.00458  1.00660  1.00625  1984  1.05802  1.03578  0.99618  0.99668  0.99903  1.02358  1.00609  0.99572  1985  1.05360  1.02114  1.00080  0.99590  1.00130  1.02563  1.00803  0.99719  1986  1.02411  0.99325  1.00270  0.99013  1.00150  1.02750  1.00923  0.99162  1987  1.08509  1.02522  1.00323  0.99823  1.01598  1.03099  1.00898  1.01417  1988  1.05747  1.00414  0.99988  0.99069  1.02084  1.02900  1.01207  1.01133  Continued on Next Page. . .  22  Chapter 1. Productivity Performance of Canada Table 1.4 – Continued Year t  ρt ρt−1  τt  αtD  αtX  αtM  t βL  t βK  αtXM  1989  1.02083  0.98000  0.99949  0.99669  1.01242  1.01743  1.01513  1.00907  1990  0.96784  0.97789  0.98239  0.97601  1.01721  1.00080  1.01395  0.99280  1991  0.91289  0.95226  0.97770  0.96525  1.03013  0.97776  1.00854  0.99433  1992  1.02845  1.03952  0.99745  1.00613  0.98853  0.99310  1.00421  0.99460  1993  1.00684  0.99427  0.99965  1.01091  0.98478  1.01498  1.00253  0.99552  1994  1.07211  1.03618  1.00887  1.02652  0.97368  1.02550  1.00059  0.99949  1995  1.06047  1.01379  1.00300  1.03239  0.98772  1.01850  1.00418  1.01971  1996  1.03665  1.00981  0.99248  0.99158  1.01786  1.01829  1.00644  1.00929  1997  1.03495  1.01249  0.99721  0.99167  1.00659  1.02019  1.00656  0.99820  1998  1.03444  1.02182  1.00051  0.99062  0.98839  1.02012  1.01304  0.97912  1999  1.05687  1.01297  0.99722  0.99650  1.01318  1.02301  1.01296  1.00963  2000  1.08909  1.02737  0.99710  1.02251  1.00432  1.02211  1.01289  1.02693  2001  0.99499  0.98834  0.99495  0.99152  0.99867  1.00825  1.01348  0.99021  2002  1.03930  1.03377  1.00177  0.98014  1.00323  1.01340  1.00711  0.98331  2003  0.99364  0.95716  0.98994  0.98132  1.05112  1.00995  1.00664  1.03148  2004  1.07087  1.01343  1.00129  1.00574  1.02093  1.02224  1.00542  1.02679  2005  1.04305  1.00131  1.00104  1.00412  1.01731  1.00912  1.00950  1.02150  2006  1.03319  0.99928  1.00552  0.99168  1.01286  1.01119  1.01240  1.00443  2007  1.04646  0.99511  1.00461  0.99587  1.02185  1.01543  1.01301  1.01763  1962-2007  1.04100  1.01010  0.99904  0.99969  1.00430  1.01600  1.01130  1.00380  1962-1973  1.06870  1.02680  1.00350  1.00220  1.00070  1.02110  1.01260  1.00290  1974-1991  1.02340  0.99914  0.99562  0.99665  1.00540  1.01320  1.01310  1.00190  1992-1999  1.04130  1.01760  0.99955  1.00580  0.99509  1.01670  1.00630  1.00070  2000-2007  1.03880  1.00200  0.99953  0.99661  1.01630  1.01400  1.01010  1.01280  Averages  23  Chapter 1. Productivity Performance of Canada Figure 1.1 shows the contribution of the combined effects of real changes t − 1) and the changes in real income in export and import prices (αXM  ([ρt /ρt−1 ] − 1) in 1962-2007. As shown in the figure, the movements of the  two series start out quite differently, but are very similar from 1994 onward. In fact, as we mention earlier, during the years 2000-2007, much of the changes in real income can be explained by the combined changes in the international prices, largely coming from the decreases in import prices. Figure 1.1: Real Income Change and Terms of Trade Contribution 1962-2007 (Gross Output Approach)  20  15  Real income TOT  5  0 19 62 19 64 19 66 19 68 19 70 19 72 19 74 19 76 19 78 19 80 19 82 19 84 19 86 19 88 19 90 19 92 19 94 19 96 19 98 20 00 20 02 20 04 20 06  Percentage points  10  -5  -10 Year  24  Chapter 1. Productivity Performance of Canada The various growth factors for the four subperiods, as listed in Table 1.4 are: • The 12 golden years for the Canadian economy, 1962-1973, when the real income generated by the business grew by 6.87% per year and TFP growth was a stellar 2.68% per year; • The 18 dismal years for the Canadian economy, 1974-1991, characterised by stagflation, oil shocks and rapidly increasing tax rates when the real income generated by the business sector grew by 2.34% per year and TFP growth was essentially zero; • The 8 years after the recession of 1991, 1992-1999, when real income growth recovered to 4.13% per year and TFP growth recovered to 1.76% per year and • The 8 years in this century, 2000-2007, when TFP growth dropped to 0.20% per year but real income growth was strong at 3.88% due to the very strong contribution made by falling real import prices during this period, which contributed on average 1.63% per year to real income growth. The annual changes presented in Table 1.4 can be converted into levels using Equations (1.54) with extensions to multiple inputs and outputs. Let t and At T t , AtD , AtX , AtM , BLt , BK XM be the cumulated products of the annual t , αt , αt , β t , β t and αt link factors τ t , αD X M L K XM respectively. Using these  definitions and cumulating Equation (1.3) leads to the following equation, which explains the cumulative growth in real gross income generated by the Canadian business sector relative to the base year 1961: ρt ρ1961  t = T t AtD AtX AtM BLt BK ;  t = 1962, 1963, . . . , 2007.  (1.5)  The cumulated variables that appear in Equation (1.5) are presented in Table 1.5 along with the cumulated terms of trade contribution factor, AtXM 25  Chapter 1. Productivity Performance of Canada defined to be the product of the two cumulated international price factors, AtX and AtM . Table 1.5 shows that the gross real income generated by the business sector grew 6.09 fold over the years 1961-2007. The main factors explaining this growth are growth of quality adjusted labour input (with cumulative growth factor 2.06), productivity increases (with cumulative growth factor 1.55), growth of capital services (with cumulative growth factor 1.67) and lower import prices (with cumulative growth factor 1.21). There were negative contributions from declining real domestic output prices (with cumulative growth factor 0.96) and declining real export prices (with cumulative growth factor 0.98). The real prices of Canadian raw material exports have increased dramatically in recent years, but these increases do not show up in the AtX column of Table 1.5, i.e. the overall real price of Canadian exports have remained relatively constant in the recent years. This apparent contradiction can be explained by the falling real prices for Canadian exports of manufactured goods. As already noted above, in the same period, the effects of the falling real import prices have been substantial.  26  Chapter 1. Productivity Performance of Canada Table 1.5: Business Sector Cumulated Growth in Real Income and Cumulated Contribution Factors  Year t  ρt ρ1961  Tt  AtD  AtX  AtM  t BL  t BK  AtXM  1961  1.00000  1.00000  1.00000  1.00000  1.00000  1.00000  1.00000  1.00000  1962  1.06402  1.03361  1.00000  1.00602  0.98556  1.03028  1.00773  0.99150  1963  1.14291  1.08554  1.00210  1.00483  0.97983  1.04903  1.01723  0.98456  1964  1.24355  1.12994  1.00519  1.00796  0.97868  1.07891  1.02869  0.98648  1965  1.34925  1.15498  1.01378  1.01046  0.98288  1.11179  1.04359  0.99316  1966  1.45950  1.17913  1.01238  1.01155  0.98899  1.14991  1.06280  1.00042  1967  1.48576  1.15790  1.01136  1.00983  0.99197  1.16691  1.08540  1.00172  1968  1.54716  1.18151  1.00719  1.01091  0.99713  1.17037  1.10206  1.00800  1969  1.63300  1.21189  1.00873  1.00832  0.99971  1.18918  1.11437  1.00803  1970  1.70070  1.23340  1.01374  1.00905  1.00224  1.19184  1.12848  1.01131  1971  1.78820  1.25120  1.02745  1.00678  1.00167  1.20981  1.14014  1.00845  1972  1.90953  1.28367  1.03226  1.00578  1.00785  1.23563  1.15052  1.01367  1973  2.20668  1.36984  1.04246  1.02683  1.00857  1.28352  1.16252  1.03563  1974  2.35207  1.38289  1.04346  1.06260  0.98704  1.31287  1.18375  1.04883  1975  2.20829  1.29927  1.02636  1.05716  0.98778  1.31200  1.20871  1.04425  1976  2.41459  1.37083  1.03478  1.06155  0.99759  1.31103  1.22605  1.05899  1977  2.56120  1.44548  1.03331  1.06956  0.97696  1.32066  1.24259  1.04492  1978  2.64781  1.44798  1.03108  1.07485  0.96565  1.35754  1.25867  1.03793  1979  2.84833  1.46527  1.03062  1.10130  0.95457  1.40724  1.27495  1.05127  1980  2.83559  1.37664  1.02298  1.11803  0.96458  1.43794  1.29844  1.07844  1981  2.88968  1.36648  1.02590  1.10781  0.96447  1.46300  1.31869  1.06845  1982  2.66070  1.30211  1.01711  1.07935  0.97896  1.41457  1.34410  1.05663  1983  2.76193  1.35101  1.00006  1.05651  1.00637  1.42105  1.35297  1.06324  1984  2.92218  1.39934  0.99624  1.05300  1.00540  1.45455  1.36120  1.05869  1985  3.07883  1.42893  0.99704  1.04868  1.00671  1.49183  1.37214  1.05571  1986  3.15306  1.41928  0.99972  1.03833  1.00822  1.53286  1.38481  1.04686  1987  3.42136  1.45508  1.00295  1.03649  1.02432  1.58036  1.39724  1.06170  Continued on Next Page. . .  27  Chapter 1. Productivity Performance of Canada Table 1.5 – Continued Year t  ρt ρ1961  Tt  AtD  AtX  AtM  t BL  t BK  AtXM  1988  3.61799  1.46111  1.00284  1.02684  1.04567  1.62620  1.41411  1.07373  1989  3.69336  1.43190  1.00233  1.02344  1.05865  1.65455  1.43550  1.08347  1990  3.57459  1.40024  0.98467  0.99889  1.07687  1.65588  1.45553  1.07567  1991  3.26319  1.33340  0.96272  0.96417  1.10931  1.61905  1.46796  1.06957  1992  3.35604  1.38609  0.96026  0.97009  1.09659  1.60787  1.47414  1.06379  1993  3.37900  1.37814  0.95993  0.98067  1.07990  1.63196  1.47788  1.05903  1994  3.62268  1.42800  0.96844  1.00667  1.05148  1.67357  1.47875  1.05849  1995  3.84175  1.44770  0.97135  1.03928  1.03856  1.70452  1.48493  1.07936  1996  3.98254  1.46190  0.96404  1.03053  1.05711  1.73569  1.49449  1.08938  1997  4.12172  1.48016  0.96135  1.02195  1.06407  1.77074  1.50430  1.08743  1998  4.26368  1.51245  0.96184  1.01236  1.05172  1.80636  1.52391  1.06472  1999  4.50616  1.53206  0.95917  1.00882  1.06558  1.84792  1.54366  1.07497  2000  4.90761  1.57400  0.95639  1.03152  1.07018  1.88878  1.56356  1.10392  2001  4.88304  1.55565  0.95156  1.02278  1.06876  1.90437  1.58464  1.09311  2002  5.07494  1.60819  0.95324  1.00247  1.07221  1.92989  1.59591  1.07486  2003  5.04267  1.53930  0.94365  0.98374  1.12702  1.94909  1.60651  1.10869  2004  5.40004  1.55997  0.94487  0.98939  1.15061  1.99244  1.61521  1.13840  2005  5.63249  1.56201  0.94585  0.99346  1.17052  2.01060  1.63056  1.16287  2006  5.81943  1.56088  0.95107  0.98519  1.18558  2.03310  1.65078  1.16802  2007  6.08978  1.55325  0.95545  0.98112  1.21149  2.06447  1.67226  1.18862  28  Chapter 1. Productivity Performance of Canada As discussed in Subsection 1.8.6, the concept of income used in this section is biased upwards. The problem is that depreciation payments are part of the user cost of capital for each asset but depreciation does not provide households with any sustainable purchasing power. Hence, the measure of real income, ρt that is used in this section is overstated. In the following section, we implement the net real income model that is described in more detail in Subsection 1.8.6.  1.4  Explaining Real Income Growth Generated by the Canadian Business Sector: the Net Output Approach  The overstatement of income problem that is implicit in the gross output approach used in the previous section can readily be remedied: all we need to do is take the user cost formula for an asset that has investment price PIt in year t and decompose it into two parts: • One part that represents depreciation and foreseen obsolescence, δPIt K t , and • The remaining part that is the reward for postponing consumption, rt PIt K t . The depreciation part δPIt K t will be removed from the user cost and treated as an intermediate input as an offset to gross investment. The user costs in the previous section took the form of: U t = (rt + δ t + τBt )PIt  (1.6)  where rt is the period t balancing real rate of interest, δ t is a geometric depreciation rate for period t, τBt is an appropriate business taxation rate  on the asset (including property taxes if applicable) and PIt is the period t  investment price for the asset. However, in the net output approach to the 29  Chapter 1. Productivity Performance of Canada measurement of income,18 we split up each (gross product) user cost times the beginning of the period stock K t into the depreciation component δPIt K t  and the remaining term (rt + τBt )PIt K t and we regard the second term as a genuine income component but we treat the first term as an intermediate input cost for the business sector and as an offset to gross investment made by the business sector during the year under consideration. Thus in this section, the new aggregate for domestic output will aggregate C + I + G like before, but the depreciation series for business structures, ICT and non-ICT machinery and equipment as negative outputs of the business sector. The ICT and non-ICT machinery and equipment and non-residential structures user costs are also changed because now the depreciation terms are omitted. The new investment aggregate I is a net investment aggregate (gross investment components were indexed with a positive sign in the aggregate and depreciation components were indexed with a negative sign in the aggregate) and the new capital services aggregate is now a “reward for waiting” capital services aggregate.19 By including these changes, the aggregate data series would have changed. The new net product prices, quantities and real income series (counterparts to Tables 1.1 to 1.3) are shown in Tables 1.6 to 1.8.20  18 See Diewert (2006a) for a more detailed discussion of the net output approach to income measurement. 19 This approach seems to be broadly consistent with an approach advocated by Rymes (1968, 1983), who stressed the role of waiting services: “Second, one can consider the ’waiting’ or ’abstinence’ associated with the net returns to capital as the non-labour primary input.” T.K. Rymes (1968, p.362). Denison (1974) also advocated a net output approach to productivity measurement. 20 The TFP growth rate τ t in Tables 1.7 and 1.9 differ slightly because when calculating τ t in Table 1.9, the input aggregate is a direct T¨ ornqvist quantity index whereas in Table 1.7, the input aggregate is an implicit quantity index; i.e. the value of inputs was deflated by the T¨ ornqvist input price index. Both the direct and implicit T¨ ornqvist indexes are superlative and hence will generally approximate each other very closely; see Diewert (1978).  30  Chapter 1. Productivity Performance of Canada Table 1.6: Prices of Canadian Business Sector Net Output and Input Aggregates  Year  t PC  t PD  t PX  t PM  t PL  t PK  t PY  t PZ  1961  1.00000  1.00000  1.00000  1.00000  1.00000  1.00000  1.00000  1.00000  1962  1.00538  1.00477  1.02992  1.05787  1.03625  1.01517  0.99506  1.03177  1963  1.02055  1.01951  1.04054  1.09648  1.06604  1.21066  1.00167  1.09654  1964  1.02437  1.02570  1.05692  1.10526  1.11049  1.32368  1.01000  1.15544  1965  1.03690  1.04437  1.07980  1.10214  1.18272  1.32976  1.03635  1.21382  1966  1.07553  1.08188  1.12451  1.11930  1.25929  1.43384  1.08239  1.29617  1967  1.11050  1.11723  1.15421  1.14426  1.33342  1.23143  1.11942  1.31158  1968  1.15168  1.15726  1.20119  1.16726  1.41620  1.33236  1.16779  1.39834  1969  1.18980  1.19642  1.23088  1.19648  1.52197  1.35774  1.20736  1.48633  1970  1.22208  1.23208  1.26710  1.21965  1.61278  1.39712  1.24801  1.56563  1971  1.24828  1.27432  1.28552  1.24798  1.72528  1.33648  1.28652  1.63857  1972  1.29847  1.33283  1.33325  1.27498  1.86392  1.45722  1.35367  1.77330  1973  1.38744  1.44088  1.51489  1.35954  2.03837  2.36282  1.49982  2.11395  1974  1.58382  1.64631  1.91283  1.64641  2.35044  2.88611  1.73846  2.47391  1975  1.82198  1.86583  2.16694  1.89029  2.70332  2.22381  1.95864  2.59155  1976  1.90726  1.96938  2.29702  1.92853  3.10646  2.45812  2.10208  2.95469  1977  2.03175  2.09474  2.50231  2.17241  3.38889  2.93446  2.20157  3.28528  1978  2.19264  2.25525  2.73837  2.41667  3.53495  3.40699  2.35179  3.51206  1979  2.40645  2.47513  3.20786  2.73027  3.78520  4.47127  2.61939  3.96092  1980  2.69497  2.75871  3.73464  2.97957  4.11781  4.51719  3.00661  4.22499  1981  2.95335  3.03348  3.99821  3.26618  4.59295  4.37549  3.27072  4.55078  1982  3.22860  3.29615  4.08926  3.43918  5.02021  3.31517  3.50723  4.60204  1983  3.46323  3.51354  4.15051  3.42324  5.22085  4.86150  3.76691  5.15888  1984  3.61506  3.65788  4.29656  3.58225  5.48101  5.66773  3.90201  5.56258  1985  3.72257  3.76949  4.38071  3.67711  5.75670  5.97348  4.00777  5.84765  1986  3.80422  3.86708  4.37060  3.74437  5.90250  5.68664  4.07122  5.88540  1987  3.89726  3.98569  4.45792  3.69150  6.11054  6.89128  4.26595  6.34107  Continued on Next Page. . .  31  Chapter 1. Productivity Performance of Canada Table 1.6 – Continued t PZ  Year  t PC  t PD  t PX  t PM  t PL  t PK  t PY  1988  4.00205  4.09782  4.47080  3.60108  6.51242  6.81320  4.44386  6.62912  1989  4.11690  4.21841  4.56005  3.59364  6.78864  6.34569  4.62309  6.72753  1990  4.35206  4.39764  4.52868  3.64367  7.04667  5.68557  4.77719  6.76258  1991  4.59099  4.59362  4.37107  3.57992  7.34245  4.19549  4.95539  6.60905  1992  4.65258  4.64682  4.49573  3.72567  7.48023  5.09947  4.98029  6.94680  1993  4.74252  4.73209  4.69389  3.92460  7.45961  5.28267  5.04452  6.97845  1994  4.77089  4.78795  4.97322  4.16089  7.41314  6.90899  5.10110  7.35256  1995  4.79147  4.81456  5.29132  4.27688  7.53622  7.86053  5.24941  7.68006  1996  4.88952  4.87348  5.32097  4.22185  7.63205  8.64495  5.37425  7.94364  1997  4.96547  4.93146  5.32718  4.23677  7.91076  8.52931  5.42748  8.13030  1998  5.03224  4.99337  5.31558  4.37960  8.13918  8.19333  5.36027  8.22664  1999  5.12045  5.08140  5.37870  4.36044  8.33890  9.06262  5.51701  8.58639  2000  5.25425  5.20795  5.71039  4.44227  8.74780  11.02286  5.83458  9.36276  2001  5.40970  5.34378  5.80311  4.58416  8.97770  10.23916  5.91943  9.35715  2002  5.47743  5.42904  5.68895  4.61494  9.09489  11.21972  5.89394  9.67799  2003  5.61543  5.56047  5.64768  4.31493  9.26253  10.98273  6.26151  9.75266  2004  5.69551  5.66394  5.78629  4.20643  9.48779  12.99255  6.57696  10.39319  2005  5.81654  5.80418  5.95274  4.15511  9.84265  14.07954  6.90744  10.91980  2006  5.92386  5.95701  5.97000  4.12736  10.29074  13.95161  7.12388  11.23987  2007  6.02712  6.10280  6.02590  4.02177  10.66121  14.56298  7.44707  11.66971  32  Chapter 1. Productivity Performance of Canada Table 1.7: Quantities of Canadian Business Sector Net Output and Input Aggregates, TFP Levels and TFP Growth Rates  QtD  QtX  QtM  QtK  QtY  QtZ  Tt  τt  1961  25452  6867  −7897  19202  5220  24422  24422  1.00000  0.00000  1962  27156  7195  1963  28698  7832  −8033  20042  −8031  20574  5348  26328  25391  1.03690  1.03690  5509  28554  26083  1.09471  1.05576  1964  30880  9105  −8989  21446  5701  31052  27143  1.14400  1.04503  1965  33935  9418  −10180  22416  5925  33184  28332  1.17124  1.02381  1966  36532  10696  −11579  23550  6230  35653  29773  1.19750  1.02242  1967  36362  11827  −12306  24056  6592  35906  30646  1.17166  0.97842  1968  37696  12910  −13527  24158  6852  37114  30995  1.19743  1.02199  1969  40646  13802  −15377  24718  7070  39110  31769  1.23106  1.02808  1970  40280  15211  −15293  24798  7342  40264  32096  1.25450  1.01904  1971  42286  15929  −16480  25333  7549  41816  32831  1.27365  1.01526  1972  45763  17257  −18892  26101  7730  44262  33788  1.30999  1.02854  1973  52582  19008  −21754  27591  7933  49995  35471  1.40947  1.07594  1974  58017  18347  −23977  28558  8319  52422  36838  1.42305  1.00963  1975  55709  16951  −23228  28530  8833  49405  37340  1.32313  0.92979  1976  59148  18390  −24774  28499  9121  52781  37551  1.40561  1.06233  1977  62034  19678  −24836  28805  9410  56883  38119  1.49224  1.06164  1978  63659  21544  −26197  30018  9728  59212  39650  1.49336  1.00075  1979  68536  22467  −28092  31737  10037  62995  41659  1.51215  1.01259  1980  66718  22548  −28715  32833  10516  60768  43244  1.40523  0.92929  1981  69487  23012  −30716  33723  10875  61904  44491  1.39137  0.99013  1982  58666  22882  −25710  32059  11336  56604  43138  1.31216  0.94307  1983  62713  24326  −28444  32283  11395  59449  43409  1.36953  1.04372  1984  67214  28444  −33270  33497  11520  63785  44743  1.42557  1.04092  1985  71770  29938  −35548  34871  11757  67611  46338  1.45908  1.02351  1986  74416  31456  −37965  36384  12018  69537  48102  1.44561  0.99077  Year t  QtL  Continued on Next Page. . .  33  Chapter 1. Productivity Performance of Canada Table 1.7 – Continued Year t  QtD  QtX  QtM  QtL  QtK  QtY  QtZ  Tt  τt  1987  79795  32933  −39889  38166  12246  74450  50086  1.48644  1.02824  1988  85403  35371  −45163  39912  12556  77740  52113  1.49175  1.00357  1989  87899  35434  −47820  40981  12974  77984  53590  1.45520  0.97550  1990  84568  37556  −48551  41030  13358  76421  53985  1.41560  0.97278  1991  77953  38167  −49281  39707  13573  70326  52730  1.33371  0.94215  1992  79864  40921  −51473  39311  13581  72949  52299  1.39486  1.04585  1993  79507  45382  −55461  40184  13599  73662  53248  1.38337  0.99177  1994  83827  51076  −60606  41745  13567  79042  54838  1.44137  1.04192  1995  85817  55452  −64385  42958  13673  82146  56148  1.46303  1.01503  1996  88197  58646  −67340  44212  13898  85143  57603  1.47809  1.01029  1997  95522  63457  −77378  45636  14100  88674  59196  1.49799  1.01346  1998  98627  69086  −81755  47078  14461  93588  60980  1.53474  1.02454  1999  101444  76337  −88261  48781  14834  98099  63031  1.55635  1.01408  2000  107179  83350  −95661  50512  15180  104411  65066  1.60470  1.03107  2001  106033  80654  −90649  51183  15588  104590  66165  1.58075  0.98507  2002  113127  81599  −92347  52290  15743  110657  67391  1.64202  1.03876  2003  114374  79268  −96341  53129  16011  106675  68489  1.55756  0.94856  2004  121796  83281  −104558  55049  16135  111286  70423  1.58024  1.01457  2005  128251  84730  −112492  55875  16435  113118  71554  1.58088  1.00040  2006  134178  85022  −117772  56905  16858  115217  73025  1.57777  0.99804  2007  140386  86002  −124556  58347  17304  117368  74899  1.56702  0.99318  34  Chapter 1. Productivity Performance of Canada Comparing Table 1.6 with Table 1.1, we see that the 2007 price of domestic absorption, PD has increased to 6.10 from the gross approach level of 5.76. This is because net investment is considerably smaller than gross investment and so the relatively low inflation prices of ICT and non-ICT machinery and equipment get much smaller weights in net domestic absorption compared to their weights in gross domestic absorption. The other striking difference between the results of the two approaches is that the price of waiting services, t , in Table 1.6 grew 14.56 fold over the sample period whereas the price PK t , in Table 1.1 grew only 10.14 fold. The of traditional capital services, PK  difference in growth rate can be explained by the fact that the prices of ICT, machinery and equipment services get much lower weights in the calculation of capital services aggregate for Table 1.6 than in the calculation for Table 1.1 because the corresponding user cost for the net concept of capital services now excludes the very large depreciation terms in the net user cost. Thus the prices of agricultural land and business non-agricultural land get much higher weights in the net user cost compared to the gross concept user cost.21 However, even though the land components now get a much higher weight in waiting services compared to machinery and equipment, the overall price increase in input prices has only increased to 11.67 fold (compared to the gross output model 10.63 fold increase in input prices) over the sample period due to the fact that the importance of capital services dramatically shrinks relative to labour services in the net output framework. Thus the level of business sector TFP using the net approach increased 1.57 fold over the period 1961-2007 and the average rate of net TFP growth was 1.04 percent per year. Recall that using the gross approach, the level of business sector TFP increased 1.55 fold over the period 1961-2007 and the average rate of gross product TFP growth was 1.01 percent per year. Therefore switching to the more appropriate net approach does not substantially increase Canadian business sector TFP growth on average. Table 1.9 shows 21 From Table 2.13, we estimate the price of agricultural land increased 19.26 fold and the price of business non-agricultural land increased 52.46 fold over the period 1961-2007. For comparison purposes, the price of residential land increased 88.93 fold over this period.  35  Chapter 1. Productivity Performance of Canada the breakdown of the net TFP growth and the averages over subperiods. Figures 1.2 and 1.3 show the levels of productivity and the rates of productivity growth using the two approaches. Throughout the sample period, the level of productivity calculated using the net output approach is constantly higher than the one using the gross output approach. However, the difference between the rates of productivity growth is not as clear.  Figure 1.2: Level of Total Factor Productivity in Canada 1961-2007  1.7  1.6  1.5  1.4 gross output net output 1.3  1.2  1.1  19 61 19 63 19 65 19 67 19 69 19 71 19 73 19 75 19 77 19 79 19 81 19 83 19 85 19 87 19 89 19 91 19 93 19 95 19 97 19 99 20 01 20 03 20 05 20 07  1  Year  36  Chapter 1. Productivity Performance of Canada  Figure 1.3: Rate of Productivity Growth in Canada 1962-2007  1.08  1.06  1.04  1.02 gross output net output  1  0.98  0.96  0.94  19 62 19 64 19 66 19 68 19 70 19 72 19 74 19 76 19 78 19 80 19 82 19 84 19 86 19 88 19 90 19 92 19 94 19 96 19 98 20 00 20 02 20 04 20 06  0.92  Year  37  Chapter 1. Productivity Performance of Canada The net counterpart to Table 1.3 is Table 1.8; ρt now represents the net real income generated by the Canadian business sector in year t.  Table 1.8: Net Real Income Generated by the Canadian Business Sector and Real Output and Input Prices  Year t  ρt  t PD Pt C  t PX Pt C  t PM Pt C  t PL Pt C  t PK Pt C  1961  24422  1.00000  1.00000  1.00000  1.00000  1.00000  1962  26058  0.99939  1.02441  1.05221  1.03070  1.00974  1963  28025  0.99899  1.01959  1.07440  1.04457  1.18628  1964  30616  1.00130  1.03178  1.07897  1.08407  1.29219  1965  33166  1.00720  1.04137  1.06292  1.14063  1.28244  1966  35880  1.00591  1.04554  1.04070  1.17085  1.33315  1967  36195  1.00606  1.03936  1.03040  1.20074  1.10890  1968  37633  1.00484  1.04299  1.01353  1.22968  1.15688  1969  39687  1.00556  1.03453  1.00561  1.27918  1.14115  1970  41119  1.00818  1.03684  0.99801  1.31970  1.14323  1971  43096  1.02086  1.02983  0.99976  1.38213  1.07066  1972  46143  1.02647  1.02679  0.98191  1.43547  1.12226  1973  54045  1.03851  1.09186  0.97989  1.46916  1.70300  1974  57540  1.03945  1.20773  1.03952  1.48403  1.82225  1975  53111  1.02407  1.18933  1.03749  1.48373  1.22055  1981  58173  1.03257  1.20436  1.01115  1.62875  1.28882  1977  61637  1.03101  1.23160  1.06923  1.66796  1.44430  1978  63509  1.02855  1.24889  1.10217  1.61219  1.55383  1979  68569  1.02854  1.33303  1.13456  1.57294  1.85804  1980  67795  1.02365  1.38578  1.10560  1.52796  1.67616  1981  68556  1.02713  1.35379  1.10592  1.55517  1.48153  1982  61489  1.02092  1.26657  1.06522  1.55492  1.02681  1983  64662  1.01453  1.19845  0.98845  1.50751  1.40375  1984  68848  1.01184  1.18852  0.99092  1.51616  1.56781  Continued on Next Page. . .  38  Chapter 1. Productivity Performance of Canada Table 1.8 – Continued Year t  ρt  t PD Pt C  t PX Pt C  t PM Pt C  t PL Pt C  t PK Pt C  1985  72791  1.01260  1.17680  0.98779  1.54643  1.60467  1986  74417  1.01652  1.14888  0.98427  1.55157  1.49482  1987  81494  1.02269  1.14386  0.94720  1.56791  1.76824  1988  86322  1.02393  1.11713  0.89981  1.62727  1.70243  1989  87573  1.02466  1.10764  0.87290  1.64897  1.54138  1990  83886  1.01047  1.04058  0.83723  1.61916  1.30641  1991  75908  1.00057  0.95210  0.77977  1.59932  0.91385  1992  78088  0.99876  0.96629  0.80078  1.60776  1.09605  1993  78353  0.99780  0.98975  0.82753  1.57292  1.11390  1994  84512  1.00358  1.04241  0.87214  1.55383  1.44816  1995  89997  1.00482  1.10432  0.89260  1.57284  1.64053  1996  93584  0.99672  1.08824  0.86345  1.56090  1.76806  1997  96925  0.99315  1.07285  0.85325  1.59315  1.71772  1998  99689  0.99228  1.05630  0.87031  1.61741  1.62817  1999  105696  0.99237  1.05044  0.85157  1.62855  1.76989  2000  115943  0.99119  1.08681  0.84546  1.66490  2.09789  2001  114445  0.98781  1.07272  0.84740  1.65956  1.89274  2002  119071  0.99117  1.03862  0.84254  1.66043  2.04835  2003  118949  0.99021  1.00574  0.76841  1.64948  1.95581  2004  128508  0.99446  1.01594  0.73855  1.66584  2.28119  2005  134333  0.99788  1.02342  0.71436  1.69218  2.42060  2006  138557  1.00560  1.00779  0.69673  1.73717  2.35515  2007  145019  1.01256  0.99980  0.66728  1.76887  2.41624  39  Chapter 1. Productivity Performance of Canada Table 1.9: Business Sector Year to Year Growth in Net Real Income and Net Year to Year Contribution Factors  Year t  ρt ρt−1  τt  αtD  αtX  αtM  t βL  t βK  αtXM  1962  1.06695  1.03690  0.99936  1.00682  0.98365  1.03439  1.00509  0.99037  1963  1.07552  1.05573  0.99958  0.99866  0.99342  1.02062  1.00654  0.99209  1964  1.09244  1.04503  1.00235  1.00352  0.99868  1.03219  1.00817  1.00219  1965  1.08330  1.02381  1.00602  1.00279  1.00483  1.03441  1.00908  1.00764  1966  1.08183  1.02242  0.99868  1.00121  1.00702  1.03875  1.01165  1.00824  1967  1.00877  0.97846  1.00016  0.99807  1.00342  1.01676  1.01232  1.00148  1968  1.03974  1.02199  0.99878  1.00122  1.00592  1.00337  1.00799  1.00714  1969  1.05457  1.02808  1.00073  0.99708  1.00296  1.01835  1.00651  1.00003  1970  1.03608  1.01904  1.00263  1.00083  1.00289  1.00256  1.00771  1.00372  1971  1.04810  1.01527  1.01250  0.99741  0.99934  1.01734  1.00547  0.99676  1972  1.07070  1.02854  1.00555  0.99887  1.00709  1.02456  1.00446  1.00595  1973  1.17124  1.07552  1.01191  1.02388  1.00082  1.04429  1.00568  1.02472  1974  1.06467  1.00963  1.00093  1.03955  0.97585  1.02594  1.01228  1.01445  1975  0.92302  0.93008  0.98430  0.99415  1.00087  0.99924  1.01408  0.99501  1981  1.09531  1.06233  1.00882  1.00478  1.01144  0.99915  1.00651  1.01627  1977  1.05956  1.06162  0.99842  1.00869  0.97623  1.00846  1.00663  0.98472  1978  1.03037  1.00074  0.99754  1.00571  0.98666  1.03229  1.00764  0.99229  1979  1.07966  1.01254  0.99999  1.02844  0.98677  1.04236  1.00801  1.01484  1980  0.98872  0.92929  0.99516  1.01757  1.01214  1.02525  1.01249  1.02992  1981  1.01123  0.99015  1.00348  0.98937  0.99986  1.02031  1.00834  0.98923  1982  0.89691  0.94322  0.99391  0.96965  1.01780  0.96092  1.00885  0.98690  1983  1.05161  1.04339  0.99387  0.97483  1.03347  1.00546  1.00113  1.00745  1984  1.06473  1.04092  0.99739  0.99609  0.99886  1.02789  1.00279  0.99495  1985  1.05728  1.02351  1.00075  0.99518  1.00152  1.03017  1.00531  0.99670  1986  1.02234  0.99078  1.00390  0.98843  1.00176  1.03236  1.00552  0.99017  1987  1.09509  1.02822  1.00612  0.99793  1.01871  1.03633  1.00477  1.01659  1988  1.05925  1.00358  1.00122  0.98918  1.02427  1.03381  1.00644  1.01319  Continued on Next Page. . .  40  Chapter 1. Productivity Performance of Canada Table 1.9 – Continued Year t  ρt ρt−1  τt  αtD  αtX  αtM  t βL  t βK  αtXM  1989  1.01449  0.97551  1.00072  0.99614  1.01449  1.02035  1.00782  1.01058  1990  0.95790  0.97280  0.98583  0.97186  1.02025  1.00094  1.00640  0.99155  1991  0.90490  0.94235  0.98998  0.95890  1.03585  0.97366  1.00297  0.99327  1992  1.02871  1.04579  0.99815  1.00731  0.98635  0.99178  1.00010  0.99356  1993  1.00340  0.99177  0.99902  1.01303  0.98186  1.01790  1.00025  0.99466  1994  1.07861  1.04182  1.00581  1.03168  0.96871  1.03047  0.99950  0.99940  1995  1.06490  1.01503  1.00121  1.03853  0.98545  1.02197  1.00188  1.02342  1996  1.03985  1.01029  0.99235  0.99006  1.02114  1.02164  1.00419  1.01099  1997  1.03570  1.01346  0.99657  0.99019  1.00777  1.02386  1.00370  0.99788  1998  1.02852  1.02454  0.99914  0.98892  0.98629  1.02383  1.00616  0.97536  1999  1.06026  1.01408  1.00009  0.99586  1.01562  1.02730  1.00618  1.01141  2000  1.09695  1.03105  0.99888  1.02656  1.00509  1.02609  1.00605  1.03178  2001  0.98708  0.98508  0.99688  0.99002  0.99844  1.00973  1.00708  0.98847  2002  1.04043  1.03876  1.00315  0.97657  1.00382  1.01586  1.00263  0.98029  2003  0.99897  0.94856  0.99909  0.97802  1.06051  1.01173  1.00451  1.03720  2004  1.08037  1.01455  1.00406  1.00672  1.02453  1.02607  1.00212  1.03142  2005  1.04532  1.00040  1.00326  1.00479  1.02017  1.01062  1.00538  1.02506  2006  1.03145  0.99804  1.00745  0.99032  1.01498  1.01303  1.00744  1.00516  2007  1.04664  0.99318  1.00676  0.99519  1.02549  1.01799  1.00754  1.02056  1962-2007  1.04070  1.01040  1.00030  0.99958  1.00510  1.01850  1.00620  1.00450  1962-1973  1.06910  1.02920  1.00320  1.00250  1.00080  1.02400  1.00760  1.00340  1974-1991  1.02090  0.99781  0.99791  0.99591  1.00650  1.01530  1.00710  1.00210  1992-1999  1.04250  1.01960  0.99904  1.00690  0.99415  1.01980  1.00270  1.00080  2000-2007  1.04090  1.00120  1.00240  0.99602  1.01910  1.01640  1.00530  1.01500  Averages  41  Chapter 1. Productivity Performance of Canada Note that from Tables 1.8, the starting level of net real income in 1961, $24,422 million, is less than the corresponding starting value of gross real income in 1961 from Tables 1.3, which was $27,722 million (see Figure 1.4). This makes sense since we now subtract depreciation from the previous estimates of gross income. Net real income generated by the Canadian business sector grew 5.94 fold over the period 1961-2007 which is 2.5 percent less than the 6.09 fold growth of gross real income. The real price of waiting capital services grew 2.42 fold, which is more rapid than the previous 1.68 fold increase in the real price of gross capital services. This difference is due to the fact that depreciation gave the prices of ICT and other machinery and equipment (which decreases in real terms) a larger role in the price of gross capital services but when depreciation is regarded as an intermediate input, the price of land (which increases in real terms) gets a much larger weight in the price of waiting capital services. The same translog contributions methodology explained in Section 1.8 can be applied to the net output model used in this section. Thus Equation (1.3) in the last section is applicable to our new measure of real income generated by the Canadian business sector and Table 1.9 shows the net income generated (counterpart to Table 1.4). The contribution of the combined effects t − 1) and the changes in real of real changes in international prices (αXM  income ([ρt /ρt−1 ] − 1) in 1962-2007 using the net approach are also depicted in Figure 1.5.  42  Chapter 1. Productivity Performance of Canada  Figure 1.4: Real Income Generated by the Business Sector in Canada 19612007  180000 160000 140000  100000 Gross Net  80000 60000 40000 20000 0 19 61 19 63 19 65 19 67 19 69 19 71 19 73 19 75 19 77 19 79 19 81 19 83 19 85 19 87 19 89 19 91 19 93 19 95 19 97 19 99 20 01 20 03 20 05 20 07  Million Dollars  120000  Year  43  Chapter 1. Productivity Performance of Canada  Figure 1.5: Real Income Change and Terms of Trade Contribution 1962-2007 (Net Output Approach)  20  15  10  5 Real income TOT  19 62 19 64 19 66 19 68 19 70 19 72 19 74 19 81 19 78 19 80 19 82 19 84 19 86 19 88 19 90 19 92 19 94 19 96 19 98 20 00 20 02 20 04 20 06  0  -5  -10  -15 Year  44  Chapter 1. Productivity Performance of Canada The net real income generated by the Canadian business sector grew at an annual rate of 4.07 percent on average over the 47 year period (19612007), which is slightly less than the gross real income growth rate of 4.10 percent. Real domestic output prices averaged a tiny positive contribution to the growth in real net income of 0.03 per year and falling real export prices made a tiny negative contribution of −0.04 per year. Positive average contributions to the growth of real net income were due to productivity improvements (1.04 per year compared to 1.01 in the gross output framework), growth of labour input (1.85 per year compared to the previous gross income 1.60), growth of capital input (0.62 per year compared to the previous 1.13) and falls in real import prices (0.51 per year compared to the previous 0.43). Comparing these average contribution growth rates in the gross and net real income frameworks leads to the following observations: • The role of productivity improvements is magnified in the net output framework compared to the gross output framework;22 • The role of increases in labour input is also magnified in the net output framework; • The role of increases in capital input (capital deepening) is greatly diminished in the net income framework and • The role of falling real import prices is also magnified in the net output framework. During the naughts, the average contribution factor for changes in real export and import prices together was 1.28 percentage points per year in 22 This phenomenon is reasonably well known and is explained in Schreyer (2001): as the input denominator in a total factor productivity measure shrinks (by treating inputs as negative outputs and placing them in the net output numerator), the resulting measure of TFP will increase. This magnification effect shows up most clearly during periods of large productivity growth; i.e. during the period 1962-1973, the average net TFP growth was 2.92% per year compared to the average gross rate of 2.68% and during the period 1992-1999, the average net TFP growth rate was 1.96% per year compared to the average gross rate of 1.76%.  45  Chapter 1. Productivity Performance of Canada the gross framework and 1.50 percentage points per year in the net framework. The corresponding contribution factor for capital growth during the naughts was 1.01 percentage points in the gross framework and 0.53 percentage points in the net framework. Looking at the contribution of falling import prices alone in the net output framework, during the entire sample period, falling import prices contributed about 0.51 percentage points per year to real income growth whereas the effects of net capital accumulation contributed about 0.62 percentage points per year. During the years of the present decade, falling import prices contributed a very large 1.91 percentage points per year to real income growth whereas the effects of net capital accumulation contributed only 0.53 percentage points per year and TFP improvements contributed only 0.12 percentage points per year. Thus for short periods, changes in the real export or import prices that a country faces can have substantially larger effects on living standards than the effects of net capital accumulation or improvements in TFP. The average annual rate of TFP growth in the net output framework was a satisfactory 1.04 percentage points per year. As usual, there are considerable variations in the average over different periods. During the golden years, 1962-1973, TFP growth averaged a spectacular 2.92 percentage points per year. During the dismal years, 1974-1991, TFP growth actually averaged −0.22 percentage points per year. The TFP growth recovered during 19921999 to average a respectable 1.96 percentage points per year. However, during the years 2000-2007, net TFP growth fell to 0.12 percentage points per year, and the decline can be explained all by two poor performance years (2001 and 2003). When these two years are excluded, the TFP growth averaged 1.27 percentage points. The year to year growth presented in Table 1.9 can be cumulated and the decomposition given by Equation (1.5) in the last section is applied to the new net data. The cumulated variables are shown in Table 1.10.  46  Chapter 1. Productivity Performance of Canada The net real income generated by the business sector grew 5.94 fold over the years 1961-2007. The main factors explaining this growth are growth of labour input (cumulative growth factor factor 2.31), productivity increases (cumulative growth factor 1.57), growth of waiting capital services (cumulative growth factor 1.33), lower real import prices (cumulative factor 1.25)23 and higher real domestic output prices (cumulative growth factor 1.01). There was a small negative contribution from declining real export prices (cumulative growth factor 0.97).  23  Note that most of this growth took place over the years 2001-2007.  47  Chapter 1. Productivity Performance of Canada Table 1.10: Business Sector Cumulated Growth in Net Real Income and Cumulated Contribution Factors  Year t  ρt ρ1961  Tt  AtD  AtX  AtM  t BL  t BK  AtXM  1961  1.00000  1.00000  1.00000  1.00000  1.00000  1.00000  1.00000  1.00000  1962  1.06695  1.03690  0.99936  1.00682  0.98365  1.03439  1.00509  0.99037  1963  1.14753  1.09468  0.99895  1.00548  0.97718  1.05572  1.01166  0.98254  1964  1.25360  1.14397  1.00130  1.00902  0.97589  1.08970  1.01993  0.98469  1965  1.35802  1.17121  1.00732  1.01184  0.98060  1.12720  1.02920  0.99221  1966  1.46915  1.19747  1.00599  1.01306  0.98749  1.17088  1.04118  1.00039  1967  1.48203  1.17168  1.00615  1.01111  0.99086  1.19050  1.05401  1.00187  1968  1.54093  1.19745  1.00492  1.01234  0.99672  1.19452  1.06243  1.00902  1969  1.62502  1.23108  1.00565  1.00939  0.99967  1.21644  1.06935  1.00906  1970  1.68365  1.25452  1.00830  1.01022  1.00256  1.21955  1.07760  1.01281  1971  1.76462  1.27368  1.02091  1.00761  1.00190  1.24070  1.08349  1.00953  1972  1.88938  1.31002  1.02657  1.00647  1.00901  1.27116  1.08832  1.01553  1973  2.21292  1.40896  1.03879  1.03050  1.00983  1.32747  1.09450  1.04063  1974  2.35603  1.42253  1.03975  1.07126  0.98545  1.36190  1.10794  1.05567  1975  2.17468  1.32306  1.02343  1.06499  0.98630  1.36086  1.12354  1.05040  1976  2.38194  1.40553  1.03246  1.07008  0.99758  1.35971  1.13085  1.06749  1977  2.52380  1.49215  1.03083  1.07939  0.97387  1.37121  1.13835  1.05118  1978  2.60045  1.49325  1.02829  1.08555  0.96087  1.41549  1.14704  1.04308  1979  2.80761  1.51198  1.02828  1.11643  0.94816  1.47544  1.15623  1.05856  1980  2.77594  1.40507  1.02330  1.13605  0.95967  1.51269  1.17067  1.09023  1981  2.80710  1.39124  1.02686  1.12397  0.95954  1.54342  1.18044  1.07849  1982  2.51771  1.31225  1.02061  1.08985  0.97661  1.48310  1.19088  1.06436  1983  2.64765  1.36919  1.01435  1.06242  1.00930  1.49120  1.19222  1.07230  1984  2.81903  1.42521  1.01171  1.05826  1.00815  1.53279  1.19555  1.06689  1985  2.98050  1.45871  1.01246  1.05316  1.00969  1.57904  1.20190  1.06336  1986  3.04707  1.44526  1.01641  1.04097  1.01146  1.63013  1.20853  1.05291  1987  3.33683  1.48605  1.02263  1.03882  1.03038  1.68935  1.21430  1.07038  Continued on Next Page. . .  48  Chapter 1. Productivity Performance of Canada Table 1.10 – Continued Year t  ρt ρ1961  Tt  AtD  AtX  AtM  t BL  t BK  AtXM  1988  3.53455  1.49137  1.02388  1.02758  1.05539  1.74646  1.22212  1.08450  1989  3.58575  1.45484  1.02462  1.02362  1.07068  1.78199  1.23167  1.09597  1990  3.43478  1.41527  1.01011  0.99481  1.09237  1.78368  1.23956  1.08670  1991  3.10813  1.33368  0.99998  0.95392  1.13153  1.73670  1.24324  1.07939  1992  3.19738  1.39475  0.99813  0.96090  1.11609  1.72243  1.24336  1.07244  1993  3.20824  1.38326  0.99715  0.97342  1.09584  1.75326  1.24367  1.06672  1994  3.46043  1.44111  1.00295  1.00426  1.06155  1.80668  1.24305  1.06607  1995  3.68503  1.46276  1.00416  1.04295  1.04611  1.84638  1.24539  1.09104  1996  3.83187  1.47781  0.99648  1.03258  1.06822  1.88633  1.25061  1.10302  1997  3.96868  1.49771  0.99306  1.02245  1.07652  1.93133  1.25523  1.10069  1998  4.08186  1.53446  0.99220  1.01112  1.06176  1.97735  1.26297  1.07356  1999  4.32783  1.55606  0.99229  1.00693  1.07835  2.03132  1.27078  1.08582  2000  4.74740  1.60437  0.99119  1.03367  1.08383  2.08432  1.27846  1.12032  2001  4.68605  1.58043  0.98809  1.02335  1.08214  2.10460  1.28752  1.10741  2002  4.87549  1.64169  0.99121  0.99937  1.08627  2.13798  1.29090  1.08559  2003  4.87047  1.55724  0.99030  0.97741  1.15200  2.16307  1.29673  1.12597  2004  5.26189  1.57990  0.99433  0.98398  1.18026  2.21947  1.29948  1.16135  2005  5.50038  1.58054  0.99756  0.98870  1.20406  2.24303  1.30647  1.19045  2006  5.67335  1.57743  1.00500  0.97913  1.22210  2.27225  1.31619  1.19659  2007  5.93796  1.56668  1.01179  0.97442  1.25325  2.31313  1.32611  1.22119  49  Chapter 1. Productivity Performance of Canada  1.5  Productivity Performance Comparison with Other Countries  Canada’s official productivity growth rate of 0.70 percent per year (19622008) had led to worries of low productivity performance comparing to fellow industrial countries and stimulated many discussions on how to close the wide “productivity gap” between Canada and the U.S. According to Statistics Canada official estimates, between 1961 and 2008 (shown in Table 1.11), productivity in Canada rose 2.0% a year on average, while productivity grew 2.3% on average in the United States. This comparison shows that the average labour productivity growth in Canada has been satisfactory as compared to the United States in the last four decades. However, when the focus is restricted to the present decade, the productivity growth rate is estimated to be only 0.7% per year in Canada while the United States grew 2.6% per year. Statistics Canada decomposed the labour productivity growth into three components: capital deepening, labour composition change and multifactor productivity growth. The results of their decomposition shows that contributions of the first two components are similar in the two countries, and the slow growth in productivity in Canada post-2000 was entirely contributed by the slow growth in multifactor productivity. In the previous section, we had shown that it is possible to estimate well using the “top down” approach and our estimates seem reasonable. Our TFP (multifactor) growth by the net approach is 1.04% (1961-2007) while the official estimate is 0.38% (1961-2007). The large difference between our estimate and the official estimate suggests that there may be possibility of miscalculation in the official estimates. Now if this is truly the case, then the so called “productivity gap” between Canada and the U.S. should in fact be narrower than we had initially thought. How does Canada compare with other developed countries? Base on the same net output approach used in this chapter, Diewert, Mizobuchi and Nomura (2005) had estimated the productivity performances in Japan (195650  Chapter 1. Productivity Performance of Canada  Table 1.11: Average Productivity Growth in Canada and the United States, 1961-2008 1961-2008  2000-2008 percent per yeara  Canada  United States  a  Output per hour worked  2.0  0.7  Contribution of capital deepening  1.3  1.1  Contribution of labour compensation  0.4  0.3  Multifactor productivity growth  0.3  −0.6  Output per hour worked  2.3  2.6  Contribution of capital deepening  0.8  1.0  Contribution of labour compensation  0.2  0.2  Multifactor productivity growth  1.2  1.4  Source: The Canadian Productivity Review, 15-206-X no.025, Statistics Canada  2003) and Diewert and Lawrence (2006) for Australia (1961-2004). Table 1.12 summarises the estimated results of Canada, Japan and Australia in the same subperiods. Overall, the average real income growth rates of all three countries are very close, all of which had an estimated rate of approximately 4.00 percent. In all subperiods, Canada had been similar to Australia in terms of real income growth rate. Japan, on the other hand, had a substantial decline in real income growth since 1992. However, Canada is the lowest in the three countries for TFP growth. In the subperiod 1962-1973, Canada had a higher TFP growth rate than Australia. But the situation had reversed since; Australia had outperformed Canada in all the remaining subperiods. Japan had experienced low growth in TFP in 1992-1999 but had somewhat recovered in 2000 onward. Thus, the three countries have similar average real income growth but very different TFP growth. The average TFP growth in Canada had been reasonable compared to other countries, and certainly not as poor as official estimates had shown. 51  Chapter 1. Productivity Performance of Canada  Table 1.12: Real Income and TFP Growth of Canada, Japan and Australia Average Growth Rate (%)  ρt ρt−1  τt  1962-  1962-1973  1974-1991  1992-1999  2000-  Canada  4.07  6.91  2.09  4.25  4.09  Japan  3.94  8.79  3.21  0.02  0.47  Australia  3.78  5.66  2.52  3.52  4.26  Canada  1.04  2.92  0.00  1.96  0.12  Japan  2.86  6.08  2.33  0.13  1.03  Australia  1.86  2.62  1.12  2.80  1.16  Recall that we had shown earlier that Canada had benefited most from increases in labour and capital inputs, and productivity improvement. The contribution of terms of trade improvements was also significant and positive for all subperiods. However, this is not the case in Australia and Japan. Diewert and Lawrence (2006) showed that terms of trade improvement was transitory in Australia: in 1974-1991 and 1992-1999, the contributions of terms of trade were negative. Similar results were found in Japan by Diewert, Mizobuchi and Nomura (2005). This transitory property of terms of trade contributions had caused the overall effect of international prices changes to be much smaller as compared to what was found in Canada. The comparison between these countries shows that Canada had generated similar real income growth rates as other developed countries through the additional benefit of terms of trade improvement over short period of time. However, it is still lagging behind in the TFP improvements. Higher TFP growth is important for income growth to be sustainable. Thus, the worries of the Canadian government are not totally unfounded, policies to encourage technology advancement are needed in the long term. In the following section, we will use our disaggregated data on exports and imports and our net output methodology to determine the effects of 52  Chapter 1. Productivity Performance of Canada changing disaggregated real export and import prices on the growth of real income generated by the production sector.  1.6  The Effects of Changing Real Export and Import Prices on Real Income Growth  We generate real price series for eight export sectors and seven import sectors using value and quantity series from disaggregated trade data. Details describing the actual construction of these price indexes are shown in Chapter 2. The export commodity classes are as follows, • X1 , Exports of agricultural and fish products; • X2 , Exports of energy products; • X3 , Exports of forest products; • X4 , Exports of industrial goods and materials (excluding energy and forest product exports); • X5 , Exports of machinery and equipment (excluding automotive products); • X6 , Exports of automotive products; • X7 , Exports of other consumer goods (excluding automotive products); • X8 , Exports of services; and the import commodity classes are, • M1 , Imports of agricultural and fish products; • M2 , Imports of energy products; • M3 , Imports of industrial goods and materials (including imports of forest products but excluding imports of energy products); 53  Chapter 1. Productivity Performance of Canada • M4 , Imports of machinery and equipment (excluding automotive products); • M5 , Imports of automotive products; • M6 , Imports of other consumer goods and • M7 , Imports of services. The values, prices and quantities of Canadian export that we constructed are depicted in Figures 1.6-1.8; and the values, prices and quantities of Canadian imports are shown in Figures 1.9-1.11. As can be seen in Figure 1.6, the export values of forest products, agricultural products and other consumer goods remained rather steady over the sample period. The services sector had grown a lot during 1980s and 1990s in its value of export but the growth seemed to have stopped since. On the other hand, industrial goods and materials (excluding energy and forest products), machinery and equipments, energy products and automotive products all had increased quite a great deal. For the export prices, most of the sectors had steady growth throughout the sample period as shown in Figure 1.7. The only exception is the energy sector which experienced huge price increase in the recent years. The export quantities depicted in Figure 1.8 show that machinery and equipment had been the fastest growing in export quantity component. Automotive products and industrial goods were also impressive in terms of growth. The other sectors had been rather slow in growth or stagnant through out the sample period. As can be seen in Figure 1.9, the most rapid growing value of imports was machinery and equipment but the growth leveled off during the last ten years. Imports of industrial goods and materials, services imports and automotive imports all grew at similarly moderate rate, while imports of other consumer goods, energy products and agricultural products all experienced little growth only. As energy prices are determined internationally, the price of energy imports grew most as in the case of exports. Prices of other imports as shown in Figure 1.10 show they grew much slower and at similar 54  Chapter 1. Productivity Performance of Canada rate. For import quantities (shown in Figure 1.11), machinery and equipment grew the fastest, this reflects Statistics Canada had quality adjusted the prices of information and computer technological devices. The rest of the import sectors seemed to be growing only moderately and at similar rate in quantity. Comparing Figures 1.7 and 1.10, it can be seen that the export prices increase more rapidly than the import prices, and this leads to the improvement in Canadian terms of trade and in living standards.  Figure 1.6: Canadian Export Values 1961-2007  120000  100000  VX1 VX2 VX3 VX4 VX5 VX6 VX7 VX8  60000  40000  20000  0 19 61 19 63 19 65 19 67 19 69 19 71 19 73 19 75 19 77 19 79 19 81 19 83 19 85 19 87 19 89 19 91 19 93 19 95 19 97 19 99 20 01 20 03 20 05 20 07  Millions of dollars  80000  Year  55  Chapter 1. Productivity Performance of Canada  Figure 1.7: Canadian Export Prices 1961-2007  25  20  PX1 PX2 PX3 PX4 PX5 PX6 PX7 PX8  15  10  5  5  3  1  9  7  5  3  1  7 20 0  20 0  20 0  20 0  19 9  19 9  19 9  19 9  9 19 8  19 9  5  3  1  9  7  5  3  1  9  7  5  3  7 19 8  19 8  19 8  19 8  19 7  19 7  19 7  19 7  19 7  19 6  19 6  19 6  19 6  19 6  1  0  Year  56  Chapter 1. Productivity Performance of Canada  Figure 1.8: Canadian Export Quantities 1961-2007  40000  35000  QX1 QX2 QX3 QX4 QX5 QX6 QX7 QX8  25000  20000  15000  10000  5000  20 07  20 05  20 03  20 01  19 99  19 97  19 95  19 93  19 91  19 89  19 87  19 85  19 83  19 81  19 79  19 77  19 75  19 73  19 71  19 69  19 67  19 65  19 63  0 19 61  Millions of 1961 dollars  30000  Year  57  Chapter 1. Productivity Performance of Canada  Figure 1.9: Canadian Import Values 1961-2007  140000  120000  VM1 VM2 VM3 VM4 VM5 VM6 VM7  80000  60000  40000  20000  20 05 20 07  20 01 20 03  19 99  19 97  19 95  19 93  19 91  19 89  19 87  19 85  19 83  19 81  19 79  19 77  19 75  19 73  19 71  19 69  19 67  19 63 19 65  0 19 61  Million of dollars  100000  Year  58  Chapter 1. Productivity Performance of Canada  Figure 1.10: Canadian Import Prices 1961-2007  35  30  25  VP1 VP2 VP3 VP4 VP5 VP6 VP7  20  15  10  5  5  3  7 20 0  20 0  20 0  9  7  1 20 0  19 9  19 9  3  1  9  7  5  3  5 19 9  19 9  19 9  19 8  19 8  19 8  19 8  9  7  5  1 19 8  19 7  19 7  19 7  1  9  7  5  3  3 19 7  19 7  19 6  19 6  19 6  19 6  19 6  1  0  Year  59  Chapter 1. Productivity Performance of Canada  Figure 1.11: Canadian Import Quantities 1961-2007  100000  90000  80000  VQ1 VQ2 VQ3 VQ4 VQ5 VQ6 VQ7  60000  50000  40000  30000  20000  10000  20 07  20 05  20 03  20 01  19 99  19 97  19 95  19 93  19 91  19 89  19 87  19 85  19 83  19 81  19 79  19 77  19 75  19 73  19 71  19 69  19 67  19 65  19 63  0 19 61  Millions of 1961 dollars  70000  Year  60  Chapter 1. Productivity Performance of Canada The methodology explained in Section 1.8 can be used to work out the contribution of each change in the real export price for our eight classes of exports to the business sector real income growth. Table 1.13 can be viewed t which as a decomposition of the aggregate export contribution factor αX  appeared in Table 1.4 of the previous section into eight commodity specific t t , which multiply together to yield the overall export to αX8 factors, αX1  t . contribution factor αX  t t t t t t t t t αX = αX1 αX2 αX3 αX4 αX5 αX6 αX7 αX8  (1.7)  The arithmetic averages of the contribution factors for the years 1962-2007 are listed in the last row of Table 1.13. The average contribution factors for the eight classes of exports are as follows. The increases in real energy products export prices had contributed an average of 0.16 percentage points to real income growth. Real export price increases in materials, services and forest products had contributed on average 0.04, 0.03 and 0.01 percentage points respectively. The big negative contributors were machinery and equipment and autos which by their increases in export prices contributed −0.14 and −0.11 percentage points on average. Both agricultural products and other consumer goods had minor negative contributions (−0.02 and −0.006 percentage points). Overall the cumulative effects of real export price changes were not large.  61  Chapter 1. Productivity Performance of Canada Table 1.13: Year to Year Export Contribution Factors Using the Gross Output Approach  Year t  αtX1  αtX2  αtX3  αtX4  αtX5  αtX6  αtX7  αtX8  1961  1.00000  1.00000  1.00000  1.00000  1.00000  1.00000  1.00000  1.00000  1962  1.00246  0.99979  1.00175  1.00105  1.00047  1.00001  0.99998  1.00050  1963  0.99909  1.00013  0.99984  0.99946  0.99993  1.00000  0.99999  1.00038  1964  1.00024  0.99981  1.00077  1.00085  1.00029  1.00003  1.00005  1.00108  1965  1.00005  1.00009  1.00025  1.00082  1.00027  0.99990  0.99999  1.00110  1966  1.00105  0.99965  0.99910  1.00106  0.99995  0.99962  0.99995  1.00069  1967  0.99912  0.99892  0.99937  0.99955  1.00024  0.99947  0.99995  1.00169  1968  0.99842  1.00004  0.99947  1.00204  1.00080  0.99926  0.99996  1.00108  1969  0.99775  1.00020  1.00063  0.99980  0.99970  0.99862  1.00003  1.00070  1970  0.99799  0.99978  0.99837  1.00243  1.00121  0.99960  0.99985  1.00150  1971  1.00017  1.00058  1.00064  0.99524  0.99947  1.00035  0.99997  1.00133  1972  1.00062  0.99975  1.00186  0.99786  0.99977  0.99863  0.99993  1.00059  1973  1.01170  1.00227  1.00621  1.00512  0.99895  0.99633  0.99997  1.00025  1974  1.01167  1.01381  1.00379  1.00950  0.99948  0.99661  0.99995  0.99960  1975  0.99357  1.00875  1.00085  0.99584  0.99932  0.99734  0.99982  0.99942  1976  0.99550  1.00655  0.99938  1.00068  0.99921  1.00080  1.00007  1.00193  1977  0.99542  1.00491  1.00152  1.00382  0.99996  1.00128  0.99998  1.00067  1978  1.00112  1.00101  1.00138  1.00165  0.99880  1.00162  0.99994  0.99942  1979  1.00390  1.00475  1.00535  1.01165  0.99911  1.00036  0.99984  0.99948  1980  1.00059  1.01020  0.99756  1.01039  0.99739  0.99936  1.00002  0.99968  1981  0.99965  0.99911  0.99821  0.99362  0.99822  1.00063  1.00008  1.00132  1982  0.99470  0.99797  0.99410  0.98862  0.99867  0.99986  0.99987  1.00027  1983  0.99600  0.99596  0.99645  0.99553  0.99710  0.99767  0.99988  1.00007  1984  0.99986  0.99692  1.00309  0.99859  0.99680  1.00148  0.99992  1.00005  1985  0.99878  0.99680  0.99926  0.99713  0.99824  1.00461  1.00007  1.00103  1986  0.99854  0.98521  1.00399  1.00146  1.00474  0.99414  1.00048  1.00147  1987  0.99813  0.99830  1.00334  1.00036  0.99942  0.99806  1.00009  1.00053  Continued on Next Page. . .  62  Chapter 1. Productivity Performance of Canada Table 1.13 – Continued Year t  αtX1  αtX2  αtX3  αtX4  αtX5  αtX6  αtX7  αtX8  1988  1.00074  0.99338  1.00021  1.00308  0.99894  0.99445  1.00004  0.99985  1989  1.00108  1.00160  1.00074  0.99741  0.99889  0.99644  1.00017  1.00036  1990  0.99633  1.00251  0.99411  0.99146  0.99753  0.99582  0.99965  0.99836  1991  0.99523  0.99360  0.99252  0.99124  0.99535  0.99815  0.99975  0.99890  1992  1.00301  1.00080  1.00079  0.99933  0.99883  1.00391  0.99988  0.99959  1993  1.00162  1.00137  1.00417  0.99950  0.99939  1.00426  0.99997  1.00061  1994  1.00218  0.99856  1.00769  1.00981  1.00161  1.00497  1.00029  1.00116  1995  1.00412  0.99903  1.01136  1.01013  1.00106  1.00404  1.00031  1.00199  1996  1.00197  1.00840  0.99328  0.99105  0.99645  1.00045  0.99995  1.00012  1997  0.99693  1.00055  0.99879  0.99775  0.99595  1.00104  0.99984  1.00081  1998  0.99868  0.99053  1.00139  0.99571  0.99862  1.00535  1.00003  1.00032  1999  0.99868  1.00929  0.99979  0.99644  0.99571  0.99687  0.99991  0.99985  2000  0.99949  1.02921  0.99874  1.00405  0.99534  0.99595  0.99967  1.00076  2001  1.00108  0.99990  0.99941  0.99601  0.99679  1.00079  0.99974  0.99778  2002  0.99956  0.98790  0.99586  0.99798  0.99896  0.99959  0.99985  1.00029  2003  0.99818  1.01374  0.99544  0.99746  0.99209  0.98638  0.99949  0.99862  2004  0.99897  1.00663  1.00271  1.00759  0.99623  0.99343  0.99976  1.00050  2005  0.99670  1.01989  0.99665  1.00302  0.99619  0.99198  0.99970  1.00026  2006  0.99901  0.99604  0.99714  1.00937  0.99629  0.99433  0.99978  0.99976  2007  1.00169  0.99898  0.99747  1.00571  0.99727  0.99492  0.99971  1.00013  Average  0.99981  1.00160  1.00010  1.00040  0.99859  0.99889  0.99994  1.00030  63  Chapter 1. Productivity Performance of Canada As in the case of export prices, the import price cumulative contribution t from Table 1.4 can be decomposed as a multiplicative contribufactor, αM  tion factors of the seven sectors as follows, t t t t t t t t = αM αM 1 αM 2 αM 3 αM 4 αM 5 αM 6 αM 7  (1.8)  Table 1.14 lists the decomposition of the aggregate import contribution factor from 1961 to 2007 constructed using the gross output approach. We calculate the arithmetic averages of these series to find the individual contribution factor of each import sector over the sample period. The biggest contribution was by the lower machinery and equipment real import prices which contributed 0.42 percentage points to the real income growth in the business sector. The lower real import prices in autos, other consumer goods, material and forest products and agricultural products also provided small contributions to the real income growth (0.05, 0.05, 0.04 and 0.02 percentage points respectively). However, the higher real energy products import prices had a negative contribution of −0.11 percentage points. Real services import prices also had −0.05 percentage points contribution. Overall, the cumulative effects of real import price changes using the gross output approach were rather large but most of the effects can be explained by the large contribution of lower machinery and equipment real import prices.  64  Chapter 1. Productivity Performance of Canada Table 1.14: Year to Year Import Contribution Factors Using the Gross Output Approach  Year t  αtM 1  αtM 2  αtM 3  αtM 4  αtM 5  αtM 6  αtM 7  1961  1.00000  1.00000  1.00000  1.00000  1.00000  1.00000  1.00000  1962  0.99874  0.99961  0.99670  0.99415  0.99894  0.99981  0.99754  1963  0.99376  1.00056  0.99909  1.00100  0.99996  1.00035  0.99948  1964  1.00066  1.00014  0.99901  1.00027  0.99953  0.99988  0.99934  1965  1.00535  0.99987  0.99962  0.99936  1.00083  1.00036  0.99889  1966  1.00137  1.00013  1.00229  1.00074  1.00077  1.00055  1.00035  1967  1.00143  1.00134  1.00032  1.00008  1.00020  1.00037  0.99926  1968  1.00014  0.99971  1.00310  1.00207  1.00021  1.00061  0.99935  1969  1.00060  1.00103  1.00046  1.00118  1.00054  1.00038  0.99839  1970  0.99914  1.00024  1.00083  1.00175  1.00102  1.00052  0.99903  1971  0.99983  0.99895  1.00293  0.99961  0.99933  1.00039  0.99839  1972  0.99925  0.99947  1.00294  1.00253  1.00108  1.00014  1.00075  1973  0.99638  0.99828  0.99795  1.00345  1.00322  1.00104  1.00042  1974  0.99773  0.97757  0.99099  1.00353  1.00350  1.00143  1.00348  1975  1.00262  0.99610  1.00285  0.99800  0.99926  1.00095  1.00098  1976  1.00315  1.00033  1.00209  1.00360  0.99988  0.99990  1.00093  1977  0.99634  0.99794  0.99672  0.99895  0.99575  0.99803  0.99541  1978  0.99914  0.99912  0.99549  1.00532  0.99538  0.99812  0.99583  1979  0.99941  0.99529  0.99284  1.00335  0.99939  0.99986  0.99835  1980  1.00068  0.98932  0.99713  1.02330  1.00048  0.99939  1.00024  1981  1.00000  0.99625  1.00065  1.01201  0.99413  0.99851  0.99842  1982  1.00303  1.00359  1.00474  1.00039  1.00076  1.00142  1.00102  1983  1.00245  1.00362  1.00491  1.01029  1.00281  1.00233  1.00131  1984  0.99957  1.00058  1.00070  1.00194  0.99914  0.99895  0.99815  1985  1.00136  0.99998  1.00327  1.00153  0.99795  0.99986  0.99737  1986  0.99894  1.00798  0.99979  1.00109  0.99793  0.99820  0.99758  1987  1.00113  0.99970  1.00175  1.00762  1.00328  1.00074  1.00166  Continued on Next Page. . .  65  Chapter 1. Productivity Performance of Canada Table 1.14 – Continued Year t  αtM 1  αtM 2  αtM 3  αtM 4  αtM 5  αtM 6  αtM 7  1988  1.00021  1.00303  0.99992  1.00726  1.00483  1.00124  1.00421  1989  1.00090  0.99950  1.00203  1.00592  1.00120  1.00092  1.00190  1990  1.00120  0.99742  1.00494  1.00733  1.00273  1.00159  1.00189  1991  1.00138  1.00371  1.00632  1.00840  1.00454  1.00210  1.00332  1992  1.00034  1.00025  0.99958  0.99776  0.99624  0.99815  0.99616  1993  1.00038  1.00079  0.99927  0.99694  0.99632  0.99799  0.99302  1994  0.99816  1.00010  0.99494  0.99505  0.99449  0.99722  0.99344  1995  0.99853  0.99957  0.99249  1.00373  0.99720  0.99875  0.99740  1996  1.00087  0.99777  1.00560  1.01056  1.00119  1.00137  1.00040  1997  0.99962  1.00043  1.00162  1.00582  1.00034  1.00057  0.99819  1998  1.00056  1.00344  0.99860  0.99798  0.99625  0.99712  0.99442  1999  1.00139  0.99781  1.00394  1.00745  1.00194  1.00104  0.99956  2000  1.00077  0.99235  0.99897  1.00809  1.00286  1.00123  0.99998  2001  0.99976  1.00165  0.99912  1.00127  1.00032  0.99895  0.99762  2002  0.99993  0.99961  1.00220  1.00182  1.00012  1.00052  0.99902  2003  1.00174  0.99824  1.00786  1.01891  1.00875  1.00741  1.00725  2004  1.00109  0.99661  0.99805  1.01224  1.00458  1.00520  1.00304  2005  1.00114  0.99293  1.00044  1.01076  1.00552  1.00355  1.00289  2006  1.00116  0.99622  0.99800  1.00870  1.00420  1.00347  1.00111  2007  0.99972  0.99996  1.00341  1.00795  1.00516  1.00359  1.00190  Average  1.00020  0.99887  1.00040  1.00420  1.00050  1.00050  0.99954  66  Chapter 1. Productivity Performance of Canada t in TaThe decomposition of the aggregate export contribution factor αX  ble 1.9 (net output approach) is listed in Table 1.15. The increases in real energy products export prices had contributed an average of 0.18 percentage points to real income growth. Real export price increases in materials, services and forest products had contributed 0.04, 0.04 and 0.01 percentage points respectively to the growth. Machinery and equipment and autos both contributed negatively, −0.17 and −0.13 percentage points, to the real income growth. Agricultural products and other consumer goods had minor negative contributions (−0.02 and −0.007 percentage points). Again, overt , were not large, all the cumulative effects of real export price changes, αX  but it can be seen that the net output approach has magnified the results obtained from the gross output approach.  67  Chapter 1. Productivity Performance of Canada Table 1.15: Year to Year Export Contribution Factors Using the Net Output Approach  Year t  αtX1  αtX2  αtX3  αtX4  αtX5  αtX6  αtX7  αtX8  1961  1.00000  1.00000  1.00000  1.00000  1.00000  1.00000  1.00000  1.00000  1962  1.00279  0.99976  1.00198  1.00119  1.00053  1.00001  0.99998  1.00056  1963  0.99897  1.00014  0.99982  0.99939  0.99992  1.00000  0.99998  1.00043  1964  1.00027  0.99979  1.00087  1.00096  1.00032  1.00003  1.00005  1.00122  1965  1.00006  1.00010  1.00028  1.00093  1.00031  0.99989  0.99998  1.00124  1966  1.00119  0.99960  0.99899  1.00120  0.99994  0.99957  0.99994  1.00078  1967  0.99900  0.99878  0.99929  0.99949  1.00028  0.99940  0.99994  1.00191  1968  0.99820  1.00005  0.99939  1.00233  1.00091  0.99916  0.99995  1.00123  1969  0.99744  1.00023  1.00072  0.99977  0.99966  0.99843  1.00003  1.00080  1970  0.99770  0.99975  0.99814  1.00278  1.00139  0.99954  0.99983  1.00172  1971  1.00020  1.00067  1.00074  0.99453  0.99939  1.00040  0.99996  1.00153  1972  1.00072  0.99971  1.00214  0.99755  0.99974  0.99842  0.99993  1.00068  1973  1.01333  1.00258  1.00708  1.00584  0.99881  0.99581  0.99997  1.00029  1974  1.01322  1.01566  1.00429  1.01077  0.99941  0.99617  0.99994  0.99954  1975  0.99265  1.01001  1.00097  0.99525  0.99922  0.99696  0.99979  0.99934  1976  0.99481  1.00755  0.99928  1.00078  0.99909  1.00093  1.00008  1.00222  1977  0.99473  1.00566  1.00175  1.00440  0.99995  1.00148  0.99998  1.00077  1978  1.00129  1.00116  1.00160  1.00190  0.99862  1.00187  0.99993  0.99933  1979  1.00450  1.00548  1.00617  1.01346  0.99898  1.00042  0.99981  0.99940  1980  1.00069  1.01179  0.99718  1.01202  0.99699  0.99926  1.00003  0.99963  1981  0.99959  0.99896  0.99792  0.99258  0.99793  1.00073  1.00009  1.00154  1982  0.99373  0.99760  0.99302  0.98655  0.99842  0.99984  0.99985  1.00032  1983  0.99523  0.99518  0.99578  0.99467  0.99654  0.99722  0.99985  1.00008  1984  0.99984  0.99636  1.00365  0.99833  0.99623  1.00174  0.99990  1.00005  1985  0.99857  0.99624  0.99913  0.99663  0.99793  1.00542  1.00008  1.00121  1986  0.99829  0.98267  1.00468  1.00171  1.00557  0.99313  1.00056  1.00172  1987  0.99782  0.99802  1.00391  1.00042  0.99932  0.99774  1.00010  1.00062  Continued on Next Page. . .  68  Chapter 1. Productivity Performance of Canada Table 1.15 – Continued Year t  αtX1  αtX2  αtX3  αtX4  αtX5  αtX6  αtX7  αtX8  1988  1.00086  0.99231  1.00024  1.00358  0.99877  0.99355  1.00005  0.99983  1989  1.00126  1.00187  1.00086  0.99698  0.99871  0.99586  1.00020  1.00041  1990  0.99568  1.00295  0.99309  0.98997  0.99710  0.99509  0.99959  0.99808  1991  0.99434  0.99241  0.99113  0.98961  0.99449  0.99781  0.99971  0.99869  1992  1.00359  1.00095  1.00094  0.99920  0.99860  1.00466  0.99986  0.99951  1993  1.00193  1.00163  1.00497  0.99940  0.99927  1.00509  0.99996  1.00073  1994  1.00260  0.99828  1.00917  1.01170  1.00192  1.00593  1.00035  1.00138  1995  1.00489  0.99885  1.01349  1.01202  1.00125  1.00479  1.00037  1.00236  1996  1.00232  1.00993  0.99206  0.98943  0.99581  1.00054  0.99994  1.00015  1997  0.99638  1.00064  0.99857  0.99735  0.99522  1.00123  0.99981  1.00095  1998  0.99844  0.98881  1.00164  0.99493  0.99837  1.00633  1.00003  1.00038  1999  0.99843  1.01101  0.99976  0.99579  0.99493  0.99629  0.99989  0.99983  2000  0.99940  1.03446  0.99851  1.00477  0.99451  0.99523  0.99961  1.00090  2001  1.00128  0.99988  0.99931  0.99531  0.99622  1.00093  0.99970  0.99738  2002  0.99948  0.98571  0.99511  0.99761  0.99878  0.99952  0.99982  1.00034  2003  0.99785  1.01620  0.99463  0.99701  0.99068  0.98397  0.99940  0.99837  2004  0.99880  1.00776  1.00317  1.00888  0.99559  0.99231  0.99972  1.00058  2005  0.99616  1.02318  0.99610  1.00351  0.99557  0.99067  0.99965  1.00030  2006  0.99885  0.99539  0.99668  1.01090  0.99568  0.99340  0.99974  0.99972  2007  1.00197  0.99881  0.99706  1.00665  0.99682  0.99409  0.99967  1.00015  Average  0.99977  1.00180  1.00010  1.00040  0.99834  0.99871  0.99993  1.00040  69  Chapter 1. Productivity Performance of Canada t in TaThe decomposition of the aggregate import contribution factor αM  ble 1.9 is shown in Table 1.16. The lower machinery and equipment real import prices contributed 0.49 percentage points to the real income growth. The lower real import prices in autos, other consumer goods, material and forest products and agricultural products contributed minor to the real income growth (0.06, 0.06, 0.05 and 0.03 percentage points respectively). The higher real energy products import prices had a −0.13 percentage points contribution whereas the higher real services import prices had a −0.06 percentage points contribution. Overall, the cumulative effects of real import t using the net output approach were quite large (0.51 perprice changes, αM  centage points) but again mostly can be explained by the large contribution of lower machinery and equipment prices. As in the case of export prices, the net output approach has magnified the results of the gross output approach.  70  Chapter 1. Productivity Performance of Canada Table 1.16: Year to Year Import Contribution Factors Using the Net Output Approach  Year t  αtM 1  αtM 2  αtM 3  αtM 4  αtM 5  αtM 6  αtM 7  1961  1.00000  1.00000  1.00000  1.00000  1.00000  1.00000  1.00000  1962  0.99858  0.99956  0.99626  0.99337  0.99879  0.99978  0.99722  1963  0.99295  1.00063  0.99897  1.00113  0.99996  1.00040  0.99941  1964  1.00075  1.00016  0.99888  1.00030  0.99947  0.99986  0.99925  1965  1.00604  0.99986  0.99958  0.99928  1.00094  1.00040  0.99875  1966  1.00154  1.00014  1.00259  1.00083  1.00087  1.00062  1.00039  1967  1.00162  1.00152  1.00036  1.00009  1.00023  1.00042  0.99916  1968  1.00016  0.99967  1.00353  1.00236  1.00024  1.00069  0.99926  1969  1.00068  1.00118  1.00053  1.00135  1.00061  1.00043  0.99817  1970  0.99902  1.00027  1.00095  1.00200  1.00116  1.00060  0.99889  1971  0.99980  0.99880  1.00337  0.99956  0.99923  1.00045  0.99815  1972  0.99913  0.99939  1.00338  1.00291  1.00124  1.00016  1.00086  1973  0.99587  0.99804  0.99767  1.00393  1.00367  1.00119  1.00048  1974  0.99743  0.97463  0.98980  1.00400  1.00397  1.00162  1.00394  1975  1.00300  0.99554  1.00325  0.99771  0.99916  1.00109  1.00112  1976  1.00363  1.00038  1.00241  1.00414  0.99987  0.99988  1.00108  1977  0.99578  0.99763  0.99622  0.99880  0.99511  0.99773  0.99472  1978  0.99900  0.99898  0.99479  1.00614  0.99467  0.99783  0.99519  1979  0.99932  0.99457  0.99174  1.00386  0.99929  0.99983  0.99810  1980  1.00078  0.98766  0.99669  1.02697  1.00055  0.99930  1.00028  1981  1.00000  0.99563  1.00075  1.01400  0.99317  0.99826  0.99816  1982  1.00359  1.00424  1.00560  1.00046  1.00090  1.00169  1.00121  1983  1.00292  1.00432  1.00586  1.01228  1.00334  1.00278  1.00156  1984  0.99950  1.00068  1.00083  1.00230  0.99899  0.99876  0.99782  1985  1.00159  0.99997  1.00384  1.00180  0.99760  0.99984  0.99691  1986  0.99876  1.00937  0.99975  1.00128  0.99757  0.99789  0.99716  1987  1.00132  0.99965  1.00205  1.00892  1.00383  1.00087  1.00194  Continued on Next Page. . .  71  Chapter 1. Productivity Performance of Canada Table 1.16 – Continued Year t  αtM 1  αtM 2  αtM 3  αtM 4  αtM 5  αtM 6  αtM 7  1988  1.00024  1.00353  0.99990  1.00844  1.00562  1.00144  1.00490  1989  1.00104  0.99942  1.00236  1.00690  1.00140  1.00107  1.00222  1990  1.00141  0.99697  1.00581  1.00862  1.00320  1.00187  1.00222  1991  1.00163  1.00441  1.00750  1.00998  1.00539  1.00249  1.00394  1992  1.00041  1.00030  0.99949  0.99733  0.99553  0.99780  0.99542  1993  1.00045  1.00094  0.99913  0.99635  0.99561  0.99761  0.99168  1994  0.99780  1.00012  0.99398  0.99410  0.99343  0.99668  0.99219  1995  0.99826  0.99949  0.99110  1.00443  0.99668  0.99852  0.99691  1996  1.00103  0.99736  1.00662  1.01249  1.00140  1.00162  1.00048  1997  0.99955  1.00051  1.00191  1.00686  1.00040  1.00068  0.99787  1998  1.00067  1.00407  0.99834  0.99761  0.99557  0.99660  0.99341  1999  1.00165  0.99741  1.00466  1.00883  1.00229  1.00124  0.99948  2000  1.00091  0.99101  0.99879  1.00954  1.00338  1.00145  0.99997  2001  0.99972  1.00195  0.99896  1.00149  1.00038  0.99876  0.99719  2002  0.99992  0.99954  1.00260  1.00215  1.00014  1.00062  0.99884  2003  1.00205  0.99792  1.00927  1.02232  1.01032  1.00874  1.00854  2004  1.00128  0.99604  0.99772  1.01434  1.00536  1.00609  1.00356  2005  1.00133  0.99177  1.00051  1.01253  1.00643  1.00413  1.00336  2006  1.00135  0.99560  0.99768  1.01012  1.00489  1.00403  1.00129  2007  0.99968  0.99995  1.00397  1.00927  1.00600  1.00418  1.00221  Average  1.00030  0.99871  1.00040  1.00490  1.00060  1.00060  0.99945  72  Chapter 1. Productivity Performance of Canada The detailed breakdown of the average contributions by changes in export and import sectoral prices shows that the real income growth generated by the business sector has benefited from lower machinery and equipment import prices. The negative effects of higher real import prices in energy and services on real income growth were offset by the large positive contribution by machinery and equipment. Other real import prices had only minor contributions. Although the cumulative effects of the real export price changes had been quite small, the cumulative effects of the real import price changes had been large and so the overall improvement in terms of trade 1961-2007 was relatively large. Note that the results obtained using the net output approach were slightly bigger than the results obtained using the gross output approach.  1.7  Conclusion  Statistics Canada has substantially revised their published KLEMS database and we use these revised data in our analysis. We have included data from 2007 and revised our estimations for other years. Furthermore, Statistics Canada made available to the authors some unpublished data from their KLEMS database on land, inventories and ICT capital which improved our earlier estimates and narrowed the differences of our estimates of TFP growth in Canada and the official Statistics Canada estimates. By incorporating these data, the precision of our analytical results was highly enhanced. We also provide a more detailed breakdown of exports and imports in the present paper so that the effects of changes in real export and import prices by commodity category on real income can be determined. There are six major conclusions that we can draw from the results. First, using new detailed data, we have shown that the productivity performance of the business sector of the Canadian economy has been reasonably satisfactory over the past 47 years. In particular, traditional gross income TFP growth averaged 1.01 percent per year over the period 196273  Chapter 1. Productivity Performance of Canada 200724 and when the net output framework was used, TFP growth averaged 1.04 percent per year. However, there was a long period (1974-1991) where the productivity performance of the Canadian business sector was decidedly unsatisfactory. We also compared Canada’s productivity performance with those of Australia and Japan. The three countries had very similar overall average real income growth. However, the estimated Canadian TFP growth rate is relatively lower than Australia and much lower than Japan. Thus the overall welfare in Canada is comparatively good on an international level, but there is still need to encourage productivity improvement. Second, we have shown that the role of explanatory factors for growth in the real income generated by the business sector of the Canadian economy changes substantially when we shift from the standard gross product growth accounting framework to a theoretically more appropriate net product growth accounting framework. In general, the main positive drivers of real income growth (growth in labour input, TFP growth and declining real import prices) are magnified but the effects of capital services input growth are greatly diminished when we switch to the net output framework as compared to the gross output framework.25 An important implication of this result is that improvements in TFP probably become the most important factor for explaining improvements in per capita living standards in the long run and the favourable effects of capital deepening are not as big as they appear to be in the traditional gross income growth accounting methodology. Third, we have shown that the “top down” approach performs well when estimating the productivity functions using the final demand deliveries. Thus, the large difference between the estimates from this approach and the official KLEMS estimates using the “bottom up” approach indicates there is a need to review the calculation of the official estimates. More 24 The corresponding Statistics Canada average Multifactor Productivity growth rate over 1962-2006 was only 0.43 percent per year. 25 Diewert, Mizobuchi and Nomura (2005) and Diewert and Lawrence (2006) found similar results for Japan and Australia using a similar net output framework.  74  Chapter 1. Productivity Performance of Canada importantly, this unveils the possibility that the wide ”productivity gap” between Canada and the U.S. may not have been as severe as portrayed. Fourth, the results presented here show that over short periods of time, changes in the external price environment facing an economy can have substantial effects on living standards. Thus during the years of the present decade, the real net income generated by the Canadian business sector grew at an average rate of 4.07 percent per year and declines in real import prices (the China effect) contributed 1.91 percentage points to this increase, which was greater than the effects of quality adjusted labour input growth (1.64 percentage points per year), increases in waiting services (0.53 percentage points per year).26 Fifth, in the decomposition of year to year real income growth, we found that changes in export and import prices can have substantial effects on real income growth. By applying detailed trade statistics, we constructed export and import price indexes for the 15 commodity classes to calculate their individual contribution to the real income growth. We found that the increases in real energy products export prices and the lower real import prices of machinery and equipment (except auto) had contributed most to real income growth. The lower export prices in machinery and equipment and autos, and higher real energy products import prices contributed negatively to the real income growth. Overall, the cumulative effects of real import price changes were much larger than the cumulative effects of real export price changes. Finally, the study uncovered many data problems which should be addressed in future work on Canadian productivity performance. A discussion of the data problems is presented in Chapter 2. More generally, it is evident that statistical agencies are able to provide reasonably accurate data on the prices and quantities of the outputs produced and intermediate inputs used 26 The Canadian experience with improvement in the terms of trade during the past decade is similar to the Australian experience; see Diewert and Lawrence (2006).  75  Chapter 1. Productivity Performance of Canada by the various industries in the economy. This is in large part due to the fact that the System of National Accounts 1993 used by most statistical agencies has developed an adequate methodology for the treatment of gross outputs and intermediate inputs. However, the corresponding methodology for the treatment of primary inputs was not well developed.27 In particular, the treatment of capital services was absent the System of National Accounts 1993 and will only be introduced in the next international version of the System of National Accounts. This absence of a standard methodology for the treatment of capital services means that national statistical agencies have not been able to deliver a generally accepted treatment of capital services in their productivity accounts. Thus detailed data on capital stocks and flows by industry is either not available from national statistical agencies or is not provided due to the lack of information on capital inputs. Given the importance of accurate information on productivity growth, it is important that international agencies provide guidance on acceptable methods for measuring primary input prices and volumes and that national statistical agencies provide more details on how they construct their estimates of primary inputs in their productivity accounts. National departments that have an interest in better productivity measurement (e.g., central banks, department of finance and industry departments) should support initiatives that will improve the measurement of primary input growth.  1.8  Explaining Real Income Growth with The Translog Approach  1.8.1  Introduction  This section will present in details the theoretical frameworks that are used in the main text. Subsection 1.8.2 looks at the production theory framework that is mainly drawn from Diewert and Morrison (1986). Sub27  The System of National Accounts 1993 (SNA) has a good chapter on wage indexes but does not provide a standard methodology for the treatment of self-employment labour input. The recent preliminary manual on the measurement of capital by Schreyer (2007) fills in an important methodological gap in the existing SNA.  76  Chapter 1. Productivity Performance of Canada section 1.8.3 explains the Translog GDP function approach. Subsection 1.8.4 shows how to decompose aggregate contribution factor due to changes in all market sector primary inputs into separate effects. Then in Subsection 1.8.5, we explain the deflated NDP Translog approach. Finally, Subsection 1.8.6 introduces how sectoral contributions to real income growth are calculated.  1.8.2  The Production Theory Framework  In this subsection, we present the production theory framework which will be used in the main text of the paper.28 The main reference is Diewert and Morrison (1986)29 but we also draw on the theory of the output price index, which was developed by Fisher and Shell (1972) and Archibald (1977). This theory is the producer theory counterpart to the theory of the cost of living index for a single consumer (or household) that was first developed by the Russian economist, A.A. Kon¨ us (1924). These economic approaches to price indexes rely on the assumption of (competitive) optimizing behaviour on the part of economic agents (consumers or producers). Thus we consider only the market sector of the economy in what follows; i.e., that part of the economy that is motivated by profit maximizing behaviour. In our empirical work, we define the market sector to be the Canadian business sector of the economy less the rental and owner occupied housing sectors.30 28 With the exception of the last subsection of this section, the material is drawn from Diewert (2005b), Diewert, Mizobuchi and Nomura (2005) and Diewert and Lawrence (2006). 29 The theory also draws on Samuelson (1953), Diewert (1974; 133-141)(1980)(1983;10771100), Fox and Kohli (1998), Kohli (1978)(1990)(1991)(2003)(2004a)(2004b)(2006)(2008), Morrison and Diewert (1990), Samuelson (1953) and Sato (1976). 30 The Canadian business sector excludes all of the general government sectors such as schools, hospitals, universities, defence and public administration where no independent measures of output can be obtained. For owner occupied housing, output is equal to input and hence no productivity improvements can be generated by this sector according to SNA conventions. Due to the difficulties involved in splitting up the residential housing stock into the rental and owner occupied portions, we omit the entire residential housing stock and the consumption of residential housing services in our data. However, we do include investment in residential housing, since that investment is part of the output of the market production sector.  77  Chapter 1. Productivity Performance of Canada Initially, we assume that the market sector of the economy produces quantities of M (net) outputs31 , y ≡ [y1 , . . . , yM ], which are sold at the positive producer prices P ≡ [P1 , . . . , PM ]. We further assume that the market sector of the economy uses positive quantities of N primary inputs, x ≡ [x1 , . . . , xN ] which are purchased at the positive primary input prices W ≡ [W1 , . . . , WN ]. In period t, we assume that there is a feasible set of output vectors y that can be produced by the market sector if the vector of primary inputs x is utilised by the market sector of the economy; denote this period t production possibilities set by S t . We assume that S t is a closed convex cone that exhibits a free disposal property.32 Given a vector of output prices P and a vector of available primary inputs x, we define the period t market sector GDP function, g t (P, x), as follows,33  g t (P, x) ≡ max{P · y : (y, x) belongs to S t }; y  t = 0, 1, 2 . . .  (1.9)  Thus market sector GDP depends on t (which represents the period t technology set S t ), on the vector of output prices P that the market sector faces 31 If the mth commodity is an import (or other produced input) into the market sector of the economy, then the corresponding quantity ym is indexed with a negative sign. We will follow Kohli (1978)(1991) and Woodland (1982) in assuming that imports flow through the domestic production sector and are “transformed” (perhaps only by adding transportation, wholesaling and retailing margins) by the domestic production sector. The recent textbook by Feenstra (2004; 76) also uses this approach. 32 For more explanation for the meaning of these properties, see Diewert (1973)(1974; 134) or Woodland(1982) or Kohli (1978)(1991). The assumption that S t is a cone means that the technology is subject to constant returns to scale. This is an important assumption since it implies that the value of outputs should equal the value of inputs in equilibrium. In our empirical work, we use an ex post rate of return in our user costs of capital, which forces the value of inputs to equal the value of outputs for each period. The function g t is known as the GDP function or the national product function in the international trade literature (see Kohli (1978)(1991), Woodland (1982) and Feenstra (2004; 76)). It was introduced into the economics literature by Samuelson (1953). Alternative terms for this function include: (i) the gross profit function; see Gorman (1968); (ii) the restricted profit function; see Lau (1976) and McFadden(1978); and (iii) the variable profit function; see Diewert (1973)(1974). 33 The function g t (P, x) will be linearly homogeneous and convex in the components of P and linearly homogeneous and concave in the components of x; see Diewert (1973)(1974; 136). Notation: P · y ≡ M m=1 Pm ym .  78  Chapter 1. Productivity Performance of Canada and on x, the vector of primary inputs that is available to the market sector. If P t is the period t output price vector and xt is the vector of inputs used by the market sector during period t and if the GDP function is differentiable with respect to the components of P at the point (P t , xt ), then the period t vector of market sector outputs y t will be equal to the vector of first order partial derivatives of g t (P t , xt ) with respect to the components of P ; i.e., we will have the following equations for each period t:34 y t = ∇P g t (P t , xt );  t = 0, 1, 2, . . .  (1.10)  Thus the period t market sector supply vector y t can be obtained by differentiating the period t market sector GDP function with respect to the components of the period t output price vector P t . If the GDP function is differentiable with respect to the components of x at the point (P t , xt ), then the period t vector of input prices W t will be equal to the vector of first order partial derivatives of g t (P t , xt ) with respect to the components of x; i.e., we will have the following equations for each period t:35 W t = ∇x g t (P t , xt );  t = 0, 1, 2, . . .  (1.11)  Thus the period t market sector input prices W t paid to primary inputs can be obtained by differentiating the period t market sector GDP function with respect to the components of the period t input quantity vector xt . The constant returns to scale assumption on the technology sets S t implies that the value of outputs will equal the value of inputs in period t; i.e., we have the following relationships: 34  These relationships are due to Hotelling (1932; 594). Note that ∇P g t (P t , xt ) ≡  t ∂g t (P t ,xt ) (P t ,xt ) , ..., ∂g ∂P ∂P1 M 35  .  These relationships are due to Samuelson (1953) and Diewert (1974; 140). Note that  ∇x g t (P t , xt ) ≡  t ∂g t (P t ,xt ) (P t ,xt ) , ..., ∂g ∂x ∂x1 N  .  79  Chapter 1. Productivity Performance of Canada  g t (P t , xt ) = P t · y t = W t · xt ;  t = 0, 1, 2, . . .  (1.12)  The above material will be useful in what follows but of course, our focus is not on GDP; instead our focus is on the income generated by the market sector or more precisely, on the real income generated by the market sector. However, since market sector GDP (the value of market sector production) is distributed to the factors of production used by the market sector, nominal market sector GDP will be equal to nominal market sector income; i.e., from Equation (1.12), we have g t (P t , xt ) = P t · y t = W t · xt . As an approximate welfare measure that can be associated with market sector production,36 we will choose to measure the real income generated by the market sector in period t, rt , in terms of the number of consumption bundles that the nominal income could purchase in period t; i.e., define ρt as follows,  ρt ≡  W t · xt PCt  t = 0, 1, 2, . . .  = w t · xt = pt · y t  = g t (pt , xt )  (1.13)  where PCt > 0 is the period t consumption expenditures deflator and the  market sector period t real output price pt and real input price wt vectors  are defined as the corresponding nominal price vectors deflated by the consumption expenditures price index; i.e., we have the following definitions:37 36  Since some of the primary inputs used by the market sector can be owned by foreigners, our measure of domestic welfare generated by the market production sector is only an approximate one. Moreover, our suggested welfare measure is not sensitive to the distribution of the income that is generated by the market sector. 37 Our approach is similar to the approach advocated by Kohli (2004b; 92), except he essentially deflates nominal GDP by the domestic expenditures deflator rather than just the domestic (household) expenditures deflator; i.e., he deflates by the deflator for C + G + I, whereas we suggest deflating by the deflator for C. Another difference in his approach compared to the present approach is that we restrict our analysis to the market  80  Chapter 1. Productivity Performance of Canada  pt ≡  Pt ; PCt  wt ≡  Wt ; PCt  t = 0, 1, 2, . . .  (1.14)  The first and last equality in (1.13) imply that period t real income, ρt ,  is equal to the period t GDP function, evaluated at the period t real  output price vector pt and the period t input vector xt , g t (P t , xt ). Thus the growth in real income over time can be explained by three main factors: τ (Technical Progress or TFP growth), growth in real output prices and the growth of primary inputs. We will shortly give formal definitions for these three growth factors. Using the linear homogeneity properties of the GDP functions g t (P, x) in P and x separately, we can show that the following counterparts to the relations (1.10) and (1.11) hold using the deflated prices p and w:38 y t = ∇p g t (pt , xt );  t = 0, 1, 2, . . .  (1.15)  wt = ∇x g t (pt , xt );  t = 0, 1, 2, . . .  (1.16)  Now we are ready to define a family of period t productivity growth factors or technical progress shift factors τ (p, x, t):39 τ (p, x, t) ≡  g t (p, x) ; g t−1 (p, x)  t = 1, 2, . . .  (1.17)  Thus τ (p, x, t) measures the proportional change in the real income produced by the market sector at the reference real output prices p and reference sector GDP, whereas Kohli deflates all of GDP (probably due to data limitations). Our treatment of the balance of trade surplus or deficit is also different. 38 If producers in the market sector of the economy are solving the profit maximization problem that is associated with g t (P, x), which uses the original output prices P , then they will also solve the profit maximization problem that uses the normalized output prices p ≡ PPC ; i.e., they will also solve the problem defined by g t (p, x). 39 This measure of technical progress is due to Diewert and Morrison(1986; 662). A special case of it was defined earlier by Diewert (1983; 1063).  81  Chapter 1. Productivity Performance of Canada input quantities used by the market sector x where the numerator in Equation (1.17) uses the period t technology and the denominator uses the period t − 1 technology. Thus each choice of reference vectors p and x will generate a possibly different measure of the shift in technology going from period t−1 to period t. Note that we are using the chain system to measure the shift in technology. It is natural to choose special reference vectors for the measure of technical progress defined by Equation (1.17): a Laspeyres type measure τLt that  chooses the period t − 1 reference vectors pt−1 and xt−1 and a Paasche type  measure τPt that chooses the period t reference vectors pt and xt :  τLt ≡ τ (pt−1 , xt−1 , t) =  τPt ≡ τ (pt , xt , t) =  g t (pt−1 , xt−1 ) ; g t−1 (pt−1 , xt−1 )  g t (pt , xt ) ; g t−1 (pt , xt )  t = 1, 2, . . .  t = 1, 2, . . .  (1.18)  (1.19)  Since both measures of technical progress are equally valid, it is natural to average them to obtain an overall measure of technical change. If we want to treat the two measures in a symmetric manner and we want the measure to satisfy the time reversal property from index number theory40 (so that the estimate going backwards is equal to the reciprocal of the estimate going forwards), then the geometric mean will be the best simple average to take.41 Thus we define the geometric mean of (1.18) and (1.19) as follows,42 τ t ≡ τLt τPt  1 2  t = 1, 2, . . .  (1.20)  40  See Fisher (1922; 64). See the discussion in Diewert (1997) on choosing the “best” symmetric average of Laspeyres and Paasche indexes that will lead to the satisfaction of the time reversal test by the resulting average index. 42 The theoretical productivity change indexes defined by (1.19)-(1.20) were first defined by Diewert and Morrison (1986; 662-663) in the nominal GDP context. See Diewert (1993) for properties of symmetric means. 41  82  Chapter 1. Productivity Performance of Canada At this point, it is not clear how we will obtain empirical estimates for the theoretical productivity growth indexes defined by (1.18)-(1.20). One obvious way would be to assume a functional form for the GDP function g t (p, x), collect data on output and input prices and quantities for the market sector for a number of years (and for the consumption expenditures deflator), add error terms to Equations (1.15) and (1.16) and use econometric techniques to estimate the unknown parameters in the assumed functional form. However, econometric techniques are generally not completely straightforward: different econometricians will make different stochastic specifications and will choose different functional forms.43 Moreover, as the number of outputs and inputs grows, it will be impossible to estimate a flexible functional form. Thus we will suggest methods for implementing measures like (1.20) in this section that are based on exact index number techniques. We turn now to the problem of defining theoretical indexes for the effects on real income due to changes in real output prices. Define a family of period t real output price growth factors α(pt−1 , pt , x, s):44  α(pt−1 , pt , x, s) ≡  g s (pt , x) ; g s (pt−1 , x)  s = 1, 2, . . .  (1.21)  Thus α(pt−1 , pt , x, s) measures the proportional change in the real income produced by the market sector that is induced by the change in real output prices going from period t − 1 to t, using the technology that is available during period s and using the reference input quantities x. Thus each choice 43  “The estimation of GDP functions such as (19) can be controversial, however, since it raises issues such as estimation technique and stochastic specification. ... We therefore prefer to opt for a more straightforward index number approach.” Kohli (2004a; 344). 44 This measure of real output price change was essentially defined by Fisher and Shell(1972; 56-58), Samuelson and Swamy (1974; 588-592), Archibald (1977; 60-61), Diewert (1980; 460-461)(1983; 1055) and Balk (1998; 83-89). Readers who are familiar with the theory of the true cost of living index will note that the real output price index defined by (1.21) is analogous to the Kon¨ us (1924) true cost of living index which is a ratio of C(u,pt ) where u is a reference utility level: g s replaces C and the cost functions, say C(u,p t−1 ) reference utility level u is replaced by the vector of reference variables x.  83  Chapter 1. Productivity Performance of Canada of the reference technology s and the reference input vector x will generate a possibly different measure of the effect on real income of a change in real output prices going from period t − 1 to period t. Again, it is natural to choose special reference vectors for the measures t that chooses the period t − 1 defined by (1.21): a Laspeyres type measure αL  reference technology and reference input vector xt−1 and a Paasche type  measure αPt that chooses the period t reference technology and reference  input vector xt :  t ≡ α(pt−1 , pt , xt−1 , t−1) = αL  αPt ≡ α(pt−1 , pt , xt , t) =  g t−1 (pt , xt−1 ) ; g t−1 (pt−1 , xt−1 )  g t (pt , xt ) ; g t (pt−1 , xt )  t = 1, 2, . . . (1.22)  t = 1, 2, . . .  (1.23)  Since both measures of real output price change are equally valid, it is natural to average them to obtain an overall measure of the effects on real income of the change in real output prices:45 t t αt ≡ αL αP  1 2  ;  t = 1, 2, . . .  (1.24)  Finally, we look at the problem of defining theoretical indexes for the effects on real income due to changes in input quantities. Define a family of period t real input quantity growth factors β(xt−1 , xt , p, s):46  β(xt−1 , xt , p, s) ≡  g s (p, xt ) ; g s (p, xt−1 )  s = 1, 2, . . .  (1.25)  45 The indexes defined by (1.21)-(1.24) were defined by Diewert and Morrison (1986; 664) in the nominal GDP function context. 46 This type of index was defined as a true index of value added by Sato (1976; 438) and as a real input index by Diewert (1980; 456).  84  Chapter 1. Productivity Performance of Canada Thus β(xt−1 , xt , p, s) measures the proportional change in the real income produced by the market sector that is induced by the change in input quantities used by the market sector going from period t − 1 to t, using the technology that is available during period s and using the reference real output prices p. Thus each choice of the reference technology s and the reference real output price vector p will generate a possibly different measure of the effect on real income of a change in input quantities going from period t − 1 to period t. Again, it is natural to choose special reference vectors for the measures defined by (1.25): a Laspeyres type measure βLt that chooses the period t − 1 reference technology and reference real output price vector pt−1 and  a Paasche type measure βPt that chooses the period t reference technology  and reference real output price vector pt :  βLt ≡ β(xt−1 , xt , pt−1 , t−1) =  βPt ≡ β(xt−1 , xt , pt , t) =  g t−1 (pt−1 , xt ) ; g t−1 (pt−1 , xt−1 )  t = 1, 2, . . . (1.26)  g t (pt , xt ) ; g t (pt , xt−1 )  t = 1, 2, . . . (1.27)  Since both measures of real input growth are equally valid, it is natural to average them to obtain an overall measure of the effects of input growth on real income:47 β t ≡ [βLt βPt ] 2 1  t = 1, 2, . . .  (1.28)  Recall that market sector real income for period t was defined by (1.13) as ρt equal to nominal period t factor payments W t · xt deflated by the household consumption price deflator PCt . It is convenient to define γ t as the period t chain rate of growth factor for real income: 47 The theoretical indexes defined by (1.25)-(1.28) were defined in Diewert and Morrison (1986; 665) in the nominal GDP context.  85  Chapter 1. Productivity Performance of Canada  γt ≡  ρt ; ρt−1  t = 1, 2, . . .  (1.29)  It turns out that the definitions for γ t and the technology, output price and input quantity growth factors τ (p, x, t), α(pt−1 , pt , x, s), β(xt−1 , xt , p, s) defined by (1.17), (1.21) and (1.25) respectively satisfy some interesting identities, which we will now develop. We have:  γt ≡ =  =  ρt ρt−1 g t (pt , xt ) g t−1 (pt−1 , xt−1 ) g t (pt , xt ) g t−1 (pt , xt )  g t−1 (pt , xt ) g t−1 (pt−1 , xt )  = τPt α(pt−1 , pt , xt , t − 1)βLt  t = 0, 1, 2, ... using definitions (1.12) and (1.13)  g t−1 (pt−1 , xt ) g t−1 (pt−1 , xt−1 ) using definitions (1.19), (1.21) and (1.26)  (1.30) In a similar fashion, we can establish the following companion identity:  γ t ≡ τLt α(pt−1 , pt , xt−1 , t)βPt ;  using definitions (1.18), (1.21) and (1.27)  (1.31) Thus multiplying (1.30) and (1.31) together and taking positive square roots of both sides of the resulting identity and using definitions (1.20) and (1.28), we obtain the following identity:  γ t ≡ τ t [α(pt−1 , pt , xt , t − 1)α(pt−1 , pt , xt−1 , t)] 2 β t ; 1  t = 1, 2, . . . (1.32)  86  Chapter 1. Productivity Performance of Canada We can derive the following alternative decomposition for γ t into growth factors in a similar way:  γ t ≡ τ t αt [β(xt−1 , xt , pt , t − 1)β(xt−1 , xt , pt−1 , t)] 2 ; 1  t = 1, 2, . . . (1.33)  It is quite likely that the real output price growth factor [α(pt−1 , pt , xt , t − 1)α(pt−1 , pt , xt−1 , t)]1/2 is fairly close to αt defined by (1.24) and it is quite likely that the input growth factor [β(xt−1 , xt , pt , t − 1)β(xt−1 , xt , pt−1 , t)]1/2 is quite close to β t defined by (1.28); i.e., we have the following approximate equalities:  [α(pt−1 , pt , xt , t − 1)α(pt−1 , pt , xt−1 , t)] 2 ≈ αt ;  t = 1, 2, . . . (1.34)  [β(xt−1 , xt , pt , t − 1)β(xt−1 , xt , pt−1 , t)] 2 ≈ β t ;  t = 1, 2, . . . (1.35)  1  1  Substituting (1.34) and (1.35) into (1.32) and (1.33) respectively leads to the following approximate decomposition for the growth of real income into explanatory factors: γ t ≈ τ t αt β t ;  t = 1, 2, . . .  (1.36)  where τ t is a technology growth factor, αt is a growth in real output prices factor and β t is a growth in primary inputs factor. Rather than look at explanatory factors for the growth in real market sector income, it is sometimes convenient to express the level of real income in period t in terms of an index of the technology level or of TFP in period t, T t , of the level of real output prices in period t, At , and of the level of  87  Chapter 1. Productivity Performance of Canada primary input quantities in period t, B t .48 Thus we use the growth factors τ t , αt and β t as follows to define the levels T t , At and B t :  T 0 ≡ 1;  T t ≡ T t−1 τ t ; t  t−1 t  t = 1, 2, . . .  (1.37)  A ≡ 1;  A ≡A  α;  t = 1, 2, . . .  (1.38)  B 0 ≡ 1;  B t ≡ B t−1 β t ;  t = 1, 2, . . .  (1.39)  0  Using the approximate equalities (1.36) for the chain links that appear in (1.37)-(1.39), we can establish the following approximate relationship for the level of real income in period t, ρt , and the period t levels for technology, real output prices and input quantities: ρt ≈ T t At B t ; ρ0  t = 0, 1, 2, . . .  (1.40)  In the following subsection, we note a set of assumptions on the technology sets that will ensure that the approximate real income growth decompositions (1.36) and (1.40) hold as exact equalities.  1.8.3  The Translog GDP Function Approach  We now follow the example of Diewert and Morrison (1986; 663) and assume that the log of the period t (deflated) GDP function, g t (p, x), has the following translog functional form:49  ln g t (p, x) ≡ at0 +  M m=1  atm ln ptm  48 This type of levels presentation of the data is quite instructive when presented in graphical form. It was suggested by Kohli (1990) and used extensively by him; see Kohli (1991)(2003)(2004a)(2004b) and Fox and Kohli (1998). 49 This functional form was first suggested by Diewert (1974; 139) as a generalization of the translog functional form introduced by Christensen, Jorgenson and Lau(1971). Diewert (1974; 139) indicated that this functional form was flexible.  88  Chapter 1. Productivity Performance of Canada  +  1 2  M  M  m=1 k=1  N  + n=1 M  amk ln ptm ln ptk  btn ln xtn + M  + m=1 n=1  1 2  N  N  n=1 j=1  bnj ln xtn ln xtj  cmn ln ptm ln xtn ;  t = 0, 1, 2, . . . (1.41)  Note that the coefficients for the quadratic terms are assumed to be constant over time. The coefficients must satisfy the following restrictions in order for g t to satisfy the linear homogeneity properties that we have assumed in Subsection 1.8.2:50  M  atm = 1  for  t = 0, 1, 2, . . .  (1.42)  btn = 1  for  t = 0, 1, 2, . . .  (1.43)  amk = akm  for  all k, m  (1.44)  bnj = bjn  for  all n, j  (1.45)  amk = 0  for  m = 1, . . . , M  (1.46)  bnj = 0  for  n = 1, . . . , N  (1.47)  m=1 N  n=1  M k=1 N j=1 50 There are additional restrictions on the parameters which are necessary to ensure that g t (p, x) is convex in p and concave in x. Note that when we divide the original prices by one of the prices, then one of the scaled prices will be identically equal to one and hence its logarithm will be identically equal to zero.  89  Chapter 1. Productivity Performance of Canada N  cmn = 0  for  m = 1, . . . , M  (1.48)  cmn = 0  for  n = 1, . . . , N  (1.49)  n=1 M m=1  Recall the approximate decomposition of real income growth going from period t − 1 to t given by (1.36), γ t ≈ τ t αt β t . Diewert and Morrison (1986; 663) showed that51 if g t−1 and g t are defined by (1.41)-(1.49) above and there is competitive profit maximizing behaviour on the part of all market sector producers for all periods t, then (1.36) holds as an exact equality; i.e., we have γ t = τ t αt β t  t = 1, 2, . . .  (1.50)  In addition, Diewert and Morrison (1986; 663-665) showed that τ t , αt and β t could be calculated using empirically observable price and quantity data for periods t − 1 and t as follows,  ln α  t  M  = m=1  1 2  t−1 pt−1 m ym pt−1 · y t−1  +  t ptm ym pt · y t  ln  ptm pt−1 m  = ln PT (pt−1 , pt , y t−1 , y t )  ln β t =  N n=1  1 2  wnt−1 xt−1 n wt−1 · xt−1  +  = ln QT (wt−1 , wt , xt−1 , xt )  (1.51)  wnt xtn w t · xt  ln  xtn xt−1 n (1.52)  51 Diewert and Morrison established their proof using the nominal GDP function g t (P, x). However, it is easy to rework their proof using the deflated GDP function g t (p, x) using the fact that g t (p, x) = g t (P/PC , x) = g t (P, x)/PC and the linear homogeneity property of g t (P, x) in P .  90  Chapter 1. Productivity Performance of Canada  τt =  γt αt β t  (1.53)  ornqvist where PT (pt−1 , pt , y t−1 , y t ) is the T¨ornqvist (T¨ornqvist (1936) and T¨ and T¨ornqvist (1937)) output price index and QT (wt−1 , wt , xt−1 , xt ) is the T¨ornqvist input quantity index. Since Equations (1.50) now hold as exact identities under our present assumptions, Equations (1.40), the cumulated counterparts to Equations (1.36), will also hold as exact decompositions; i.e., under our present assumptions, we have ρt = T t At B t ρ0  t = 1, 2, . . .  (1.54)  We will implement the real income decompositions (1.50) and (1.54) in Sections 1.3 and 1.4.  1.8.4  The Translog GDP Function Approach and Changes in the Terms of Trade  For some purposes, it is convenient to decompose the aggregate period t contribution factor due to changes in all deflated output prices αt into separate effects for each change in each output price. Similarly, it can sometimes be useful to decompose the aggregate period t contribution factor due to changes in all market sector primary input quantities β t into separate effects for each change in each input quantity. In this subsection, we show how this can be done, making the same assumptions on the technology that were made in the previous subsection. We first model the effects of a change in a single (deflated) output price, say pm , going from period t − 1 to t.  Counterparts to the theoretical  Laspeyres and Paasche type price indexes defined by (1.22) and (1.23) in  91  Chapter 1. Productivity Performance of Canada Subsection 1.8.2 for changes in all (deflated) output prices are the following t that chooses the period t − 1 reference techLaspeyres type measure αLm  nology and holds constant other output prices at their period t − 1 levels and holds inputs constant at their period t − 1 levels xt−1 and a Paasche type measure αPt m that chooses the period t reference technology and refer-  ence input vector xt and holds constant other output prices at their period t levels:  t ≡ αLm  t−1 t−1 t−1 t−1 t ) g t−1 (pt−1 1 , . . . , pm−1 , pm , pm+1 , . . . , pM , x ; t−1 t−1 t−1 g (p , x )  m = 1, . . . , M ; t = 1, 2, . . . (1.55) αPt m ≡  g t (pt , xt ) ; t t t g t (pt1 , . . . , ptm−1 , pt−1 m , pm+1 , . . . , pM , x ) m = 1, . . . , M ; t = 1, 2, . . . (1.56)  Since both measures of real output price change are equally valid, it is natural to average them to obtain an overall measure of the effects on real income of the change in the real price of output m:52  t t αm ≡ [αLm αPt m ] 2 ; 1  m = 1, . . . , M ; t = 1, 2, . . .  (1.57)  Under the assumption that the deflated GDP functions g t (p, x) have the translog functional forms as defined by (1.41)-(1.49) in the previous subsection, the arguments of Diewert and Morrison (1986; 666) can be adapted to give us the following result: 52 The indexes defined by (1.55)-(1.57) were defined by Diewert and Morrison (1986; 666) in the nominal GDP function context.  92  Chapter 1. Productivity Performance of Canada  t ln αm =  1 2  t−1 pt−1 m ym t−1 p · y t−1  +  t ptm ym t p · yt  ln  ptm pt−1 m  ;  m = 1, . . . , M ; t = 1, 2, . . . (1.58) t is equal to the mth term in the summation of the terms Note that ln αm  on the right hand side of (1.51). This observation means that we have the following exact decomposition of the period t aggregate real output price contribution factor αt into a product of separate price contribution factors; i.e., we have under present assumptions: t ; αt = α1t α2t . . . αM  t = 1, 2, . . .  (1.59)  The above decomposition is useful for analyzing how real changes in the price of exports (i.e., a change in the price of exports relative to the price of domestic consumption) and in the price of imports impact on the real income generated by the market sector. In the empirical illustration which follows later, we let M equal three. The three net outputs are:  1. Domestic sales (C + I + G); 2. Exports (X) and 3. Imports (M ). Commodities 1 and 2 are outputs, so y1 and y2 will be positive but commodity 3 is an input into the market sector, so y3 will be negative. Hence an increase in the real price of exports will increase real income but an increase in the real price of imports will decrease the real income generated by the market sector, as is evident by looking at the contribution terms defined by t > 0) and for m = 3 (where y t < 0). (1.58) for m = 2 (where ym m  93  Chapter 1. Productivity Performance of Canada As mentioned above, it is also useful to have a decomposition of the aggregate contribution of input growth to the growth of real income into separate contributions for each important class of primary input that is used by the market sector. We now model the effects of a change in a single input quantity, say xn , going from period t − 1 to t. Counterparts to the theoretical Laspeyres and Paasche type quantity indexes defined by (1.26) and (1.27) t above for changes in input n are the following Laspeyres type measure βLn  that chooses the period t − 1 reference technology and holds constant other input quantities at their period t − 1 levels and holds real output prices at their period t − 1 levels pt−1 and a Paasche type measure βPt n that chooses the period t reference technology and reference real output price vector pt  and holds constant other input quantities at their period t levels:  t βLn ≡  t−1 t−1 t−1 t g t−1 (pt−1 , xt−1 1 , . . . , xn−1 , xn , xn+1 , . . . , xN ) ; g t−1 (pt−1 , xt−1 )  n = 1, . . . , N ; t = 1, 2, . . .  βPt n ≡  (1.60)  g t (pt , xt ) ; t t g t (pt , xt1 , . . . , xtn−1 , xt−1 n , xn+1 , . . . , pN ) n = 1, . . . , N ; t = 1, 2, . . .  (1.61)  Since both measures of input change are equally valid, as usual, we average them to obtain an overall measure of the effects on real income of the change in the quantity of input n:53  t βnt ≡ [βLn βPt n ] 2 ; 1  n = 1, . . . , N ; t = 1, 2, . . .  (1.62)  53  The indexes defined by (1.60)-(1.62) were defined by Diewert and Morrison (1986; 667) in the nominal GDP function context.  94  Chapter 1. Productivity Performance of Canada Under the assumption that the deflated GDP functions g t (p, x) have the translog functional forms as defined by (1.41)-(1.49) in the previous subsection, the arguments of Diewert and Morrison (1986; 667) can be adapted to give us the following result:  ln βnt =  1 2  wnt−1 xt−1 n wt−1 · xt−1  +  wnt xtn wt · xt  ln  xtn xt−1 n  ;  n = 1, . . . , N ; t = 1, 2, . . . (1.63) Note that ln βnt is equal to the nth term in the summation of the terms on the right hand side of (1.52). This observation means that we have the following exact decomposition of the period t aggregate input growth contribution factor β t into a product of separate input quantity contribution factors; i.e., we have under present assumptions: t ; β t = β1t β2t . . . βN  1.8.5  t = 1, 2 . . .  (1.64)  The Deflated NDP Translog Approach  There is a severe flaw with all of the analysis presented in the previous subsections. The problem is that depreciation payments are part of the user cost of capital for each asset but depreciation does not provide households with any sustainable purchasing power. Hence our real income measure defined by (1.13) is overstated. To see why GDP overstates income, consider the model of production that is described by the following quotations: “We must look at the production process during a period of time, with a beginning and an end. It starts, at the commencement of the Period, with an Initial Capital Stock; to this there is applied a Flow Input of labour, and from it there emerges a Flow Output called Consumption; then there is a Closing Stock of Capital left 95  Chapter 1. Productivity Performance of Canada over at the end. If Inputs are the things that are put in, the Outputs are the things that are got out, and the production of the Period is considered in isolation, then the Initial Capital Stock is an Input. A Stock Input to the Flow Input of labour; and further (what is less well recognized in the tradition, but is equally clear when we are strict with translation), the Closing Capital Stock is an Output, a Stock Output to match the Flow Output of Consumption Goods. Both input and output have stock and flow components; capital appears both as input and as output” John R. Hicks (1961; 23). “The business firm can be viewed as a receptacle into which factors of production, or inputs, flow and out of which outputs flow...The total of the inputs with which the firm can work within the time period specified includes those inherited from the previous period and those acquired during the current period. The total of the outputs of the business firm in the same period includes the amounts of outputs currently sold and the amounts of inputs which are bequeathed to the firm in its succeeding period of activity.” Edgar O. Edwards and Philip W. Bell (1961; 71-72). Hicks and Edwards and Bell obviously had the same model of production in mind: in each accounting period, the business unit combines the capital stocks and goods in process that it has inherited from the previous period with “flow” inputs purchased in the current period (such as labour, materials, services and additional durable inputs) to produce current period “flow” outputs as well as end of the period depreciated capital stock components which are regarded as outputs from the perspective of the current period (but will be regarded as inputs from the perspective of the next period).54 54  For more on this model of production and additional references to the literature, see the Appendices in Diewert (1977)(1980). The usual user cost of capital can be derived from this framework if depreciation is independent of use.  96  Chapter 1. Productivity Performance of Canada All of the “flow” inputs that are purchased during the period and all of the “flow” outputs that are sold during the period are the inputs and outputs that appear in the usual definition of cash flow. These are the flow inputs and outputs that are very familiar to national income accountants. But this is not the end of the story: the firm inherits an endowment of assets at the beginning of the production period and at the end of the period, the firm will have the net profit or loss that has occurred due to its sales of outputs and its purchases of inputs during the period. As well, it will have a stock of assets that it can use when it starts production in the following period. Just focusing on the flow transactions that occur within the production period will not give a complete picture of the firm’s productive activities. Hence, to get a complete picture of the firm’s production activities over the course of a period, it is necessary to add the value of the closing stock of assets less the beginning of the period stock of assets to the cash flow that accrued to the firm from its sales and purchases of market goods and services during the accounting period. We illustrate the above theory by considering a very simple two output, two input model of the market sector. One of the outputs is output in period t, Y t and the other output is an investment good, I t . One of the inputs is the flow of non-capital primary input X t and the other input is K t , capital services. Suppose that the average prices during period t of a t and P t respectively. Suppose further that unit of Y t , X t and I t are PYt , PX I  the interest rate prevailing at the beginning of period t is rt . The value of  the beginning of period t capital stock is assumed to be PIt , the investment price for period t. In order to induce households to let the business sector use the initial stock of capital, firms have to pay households interest equal to rt PIt K t . Then neglecting balance sheet items, the market sector’s period t cash flow is:55 t CF t ≡ PYt Y t + PIt I t − PX X t − rt PIt K t 55  (1.65)  For equity financed firms, we need to include an imputed return for equity capital.  97  Chapter 1. Productivity Performance of Canada K t is interpreted as the firm’s beginning of period t stock of capital it has at its disposal and its end of period stock of capital is defined to be K t+1 . These capital stocks are valued at the balance sheet prices prevailing at the beginning and end of period t, PIt and PIt+1 respectively. The market sector period t pure profit is defined as its cash flow plus the value of its end of period t capital stock less the value of its beginning of period t capital stock: Πt ≡ CF t + PIt+1 K t+1 − PIt K t  (1.66)  Now the end of period depreciated stock of capital is related to the beginning of the period stock by the following equation: K t+1 = (1 − δ)K t  (1.67)  where 0 < δ < 1 denotes the depreciation rate. Now substitute (1.65) and (1.67) into the definition of pure profits (1.66) and we obtain the following expression:  t X t − rt PIt K t Πt ≡ PYt Y t + PIt I t − PX  +PIt+1 (1 − δ)K t − PIt K t t = PYt Y t + PIt I t − PX Xt  −{rt PIt + δPIt+1 − (PIt+1 − PIt )}K t  (1.68)  The expression that precedes the capital stock K t , {rt PIt +δPIt+1 −(PIt+1 −  PIt )}, can be recognized as the user cost of capital;56 it is the gross rental price that must be paid to a capitalist in order to induce him or her to loan the services of a unit of the capital stock to the production sector. 56  See Christensen and Jorgenson (1969) for a derivation in continuous time and Diewert (1980; 471) for a derivation in discrete time.  98  Chapter 1. Productivity Performance of Canada Some simplifications for (1.68) occur if we make two additional assumptions: • Assume that producers and households expect price level stability so that the end of the period price for a new unit of capital PIt+1 is expected to be equal to the beginning of the period price for a new unit of capital PIt ; in this case, we can interpret rt as the period t real interest rate;57 • Assume that pure profits are zero so that Πt equals zero. Substituting these two assumptions into Equation (1.68) leads to the following expression: Πt = PYt Y t + PIt I t − PXt X t − {rt PIt + δPIt }K t = 0  (1.69)  Equation (1.69) can be rearranged to yield the following value of output equals value of input equation: t X t + {rt PIt + δPIt }K t PYt Y t + PIt I t = PX  (1.70)  Equation (1.70) is essentially the closed economy counterpart to the (gross) value of outputs equals (gross) value of primary inputs equation (1.12), P t · y t = W t · xt , that we have been using thus far in this section. We now come to the point of this rather long digression: the (gross) payments to primary inputs that is defined by the right hand side of (1.70) is not income, in the sense of Hicks.58 The owner of a unit of capital cannot spend the entire period t gross rental income {rt PIt + δPIt } on consumption during period t  because the depreciation portion of the rental, δPIt , is required in order to  keep his or her capital intact. Thus the owner of a new unit of capital at 57 This assumption can be relaxed somewhat and we can still end up with much the same model; see Diewert (2006a). 58 We will use Hicks’ third concept of income here: “Income No. 3 must be defined as the maximum amount of money which the individual can spend this week, and still be able to expect to spend this week, and still be able to expect to spend the same amount in real terms in each ensuing week.” J.R. Hicks (1946; 174).  99  Chapter 1. Productivity Performance of Canada the beginning of period t loans the unit to the market sector and gets the gross return {rt PIt + δPIt } at the end of the period plus the depreciated unit of the initial capital stock, which is worth only (1 − δ)PIt . Thus δPIt of this  gross return must be set aside in order to restore the lender of the capital services to his or her original wealth position at the beginning of period t. This means that period t Hicksian market sector income is not the value t X t + {r t P t + δP t }K t ; instead it is the of payments to primary inputs, PX I I  value of payments to labour PXt X t plus the reward for waiting, rt PIt K t . Us-  ing this definition of market sector (net) Hicksian income, we can rearrange Equation (1.70) as follows:  Hicksian market sector income ≡ PXt X t + rt PIt K t = PYt Y t + PIt I t − δPIt K t = Value of consumption + value of gross investment − value of depreciation  (1.71)  Thus in this Hicksian net income framework, our new output concept is equal to our old output concept less the value of depreciation. We take the price of depreciation to be the corresponding investment price PIt and the quantity of depreciation is taken to be the depreciation rate times the beginning of the period stock, δK t . Hence the overstatement of income problem that is implicit in the approaches used in previous subsections can readily be remedied: all we need to do is to take the user cost formula for an asset and decompose it into two parts: • One part that represents depreciation and foreseen obsolescence, δPIt K t , and • The remaining part that is the reward for postponing consumption, rt PIt K t . 100  Chapter 1. Productivity Performance of Canada In our empirical work, our user costs in the gross output approach took the following form: ut = (rt + δ t + τ t )PIt  (1.72)  where rt is the balancing period t real rate of interest, δ t is a geometric depreciation rate for period t, τ t is an average capital taxation rate on the asset and PIt is the period t investment price for the asset. However, when we used the net output approach, we split up each (gross product) user cost times the beginning of the period stock K t into the depreciation component δ t PIt K t and the remaining term (rt + τ t )PIt K t and we regarded the second term as a genuine income component but the first term was treated as an intermediate input cost for the market sector and was an offset to gross investment made by the market sector during the period under consideration. In the paper, when the net approach was used, the investment aggregate I was a net investment aggregate (gross investment components were indexed with a positive sign in the aggregate and depreciation components were indexed with a negative sign in the aggregate). The capital services aggregate in the net approach was a “reward for waiting” capital services aggregate rather than the gross return aggregate that was used in the gross output approach.59  1.8.6  Sectoral Contributions to Real Income Growth  The above theory applied to the market sector as a whole. However, it is of considerable interest to determine which separate industries contributed the most to the overall growth of real income generated by the market sector of the economy. Hence, in this subsection, we outline how this can be done if industry data on outputs, inputs and the corresponding prices 59 This approach seems to be broadly consistent with an approach advocated by Rymes (1968) (1983), who stressed the role of waiting services: “Second, one can consider the ‘waiting’ or ‘abstinence’ associated with the net returns to capital as the non-labour primary input.” T.K. Rymes (1968; 362). Denison (1974) also advocated a net output approach to productivity measurement.  101  Chapter 1. Productivity Performance of Canada are available.60 However, at the outset, it should be noted that in general, we will not be able to single out the effects of changes in real international prices as we were able to do when the entire business sector is treated as a single industry.61 We assume that there are I industries in the market sector of the economy. As in Subsection 1.8.2, we assume that there is a common list of M (net) outputs which each industry produces or uses as intermediate inputs. The it ], which net output vector for industry i in period t is y it ≡ [y1it , . . . , yM  are sold at the positive producer prices for industry i in period t, P it ≡  it ], for i = 1, . . . , I. There is also a common list of N primary [P1it , . . . , PM  inputs used by each industry. In period t, we assume that industry i uses it non-negative quantities of N primary inputs, xit ≡ [xit 1 , . . . , xN ] which are  purchased at the positive primary input prices W it ≡ [W1it , . . . , WNit ] for  i = 1, . . . , I. In each period t, we assume that there is a feasible set of net output vectors y i that can be produced by industry i if the vector of primary inputs xi is utilised by that industry; denote this period t production possibilities set by S it . We assume that S it is a closed convex cone that exhibits a free disposal property. We shall take the net product point of view developed in the previous subsection for each industry in what follows. 60  In Canada, such data are available from the Input-Output and Productivity Divisions of Statistics Canada. However, these data for the past five years are not available at present. 61 The problem is not methodological; it is a data problem. In order to determine the effects of changing real import and export prices on the real income generated by an industry, we require information on the value and price of exports produced by the industry and on the value and price of imports used by the industry. However, the System of National Accounts 1993 does not set up the production accounts so that the exports produced and imports used by an industry are recorded in the recommended system of production accounts. In theory, this problem can be remedied simply by distinguishing industry outputs as being either exported or delivered to domestic users and by distinguishing industry inputs as being either imports or supplied by a domestic producer; see Diewert (2007b)(2007c) for the details of the resulting modified industry accounts. In practice, it will be extremely difficult to collect the required information. For further discussion of these issues, see Section 2.1 in Chapter 2.  102  Chapter 1. Productivity Performance of Canada Given a vector of industry i net output prices P it and a vector of available primary inputs xit for that industry, we define the industry i period t net product function, g it (P it , xit ), as follows,  g it (P it , xit ) ≡ max{P it · y : (y, xit ) belongs to S it } y it it  = P y ; i = 1, . . . , I; t = 0, 1, 2, . . . (1.73) Since we have assumed constant returns to scale for each industry, it is natural to assume that the income generated by industry i in period t, W it · xit , is equal to the corresponding value of net product, P it · y it ; i.e., we have:  P it · y it = W it · xit  i = 1, . . . , I; t = 0, 1, 2, . . .  (1.74)  Define the period t, industry i real input and output price vectors, wit and pit respectively, as follows,  wit ≡  W it ; PCt  pit ≡  P it ; PCt  i = 1, . . . , I; t = 0, 1, 2, . . .  (1.75)  As in Subsection 1.8.2, we can define the real income generated by industry i in period t, ρit , as the nominal income generated by industry i in period t, W it · xit , divided by the consumption price deflator for period t, PCt . Using (1.73)-(1.75), we have:  ρit ≡  W it · xit PCt  i = 1, . . . , I; t = 0, 1, 2, . . .  = wit · xit  103  Chapter 1. Productivity Performance of Canada  =  P it · y it PCt  = pit · y it  = g it (pit , xit )  (1.76)  where the last equality follows using (1.73)-(1.75) and the linear homogeneity of g it (P it , xit ) in P it . We now rework the theoretical analysis presented in Subsections 1.8.21.8.4, except we apply it at the industry level instead of the economy-wide market sector level. Thus define γ it as the period t chain link rate of growth factor for the real income generated by industry i:  γ it ≡  ρit ; ρit−1  i = 1, . . . , I; t = 1, 2, . . .  (1.77)  Now assume that the industry i, period t (deflated) GDP function, g it (p, x), has a translog functional form analogous to that defined above by (1.41)(1.49). Repeat the analysis at the national level that led up to Equation (1.50), except now apply it at the industry level. We can derive the following industry counterparts to the national equation (1.50): ρit pit · y it = = γ it = τ it αit β it ; pit−1 · y it−1 ρit−1  i = 1, . . . , I; t = 0, 1, 2, . . . (1.78)  where the period t, industry i chain link technical progress growth rate τ it , output price growth rate αit and input quantity growth rate β it can be calculated using the period t and t − 1 price and quantity data for industry i as follows, for i = 1, . . . , I; t = 0, 1, 2, . . .:  ln α  it  M  ≡ m=1  1 2  it−1 pit−1 m ym pit−1 · y it−1  +  it pit m ym pit · y it  ln  pit m it−1 pm  104  Chapter 1. Productivity Performance of Canada = ln PT (pit−1 , pit , y it−1 , y it )  ln β it ≡  N n=1  1 2  wnit−1 xit−1 n wit−1 · xit−1  +  (1.79)  wnit xit n wit · xit  ln  xit n xit−1 m  = ln QT (wit−1 , wit , xit−1 , xit )  τ it ≡  (1.80)  γ it αit β it  (1.81)  where PT (pit−1 , pit , y it−1 , y it ) is the period t, industry i T¨ornqvist output price index and QT (wit−1 , wit , xit−1 , xit ) is the period t, industry i T¨ornqvist input quantity index. Recall that in Subsection 1.8.3, we defined cumulated counterparts to the chain link Equations (1.50). We can do the same type of operation for the industry data. Thus define the industry i level of total factor productivity in period t relative to period 0 as T it , the industry i level of real output prices in period t relative to period 0 as Ait and the industry i level of primary input in period t relative to period 0 as B it . These industry levels can be defined in terms of the corresponding industry chain link factors, τ it , αit and β it as follows, T i0 ≡ 1;  T it ≡ T it−1 τ it ;  t = 1, 2, . . .  (1.82)  Ai0 ≡ 1;  Ait ≡ Ait−1 αit ;  t = 1, 2, . . .  (1.83)  B i0 ≡ 1;  B it ≡ B it−1 β it ;  t = 1, 2, . . .  (1.84)  105  Chapter 1. Productivity Performance of Canada Since Equations (1.78) hold as exact identities under our present assumptions, the following cumulated counterparts to these equations will also hold as exact decompositions: ρit pit · y it = = T it Ait B it ; pi0 · y i0 ρi0  i = 1, . . . , I; t = 1, 2, . . .  (1.85)  Thus three factors contribute to the period t level of real income generated by industry i relative to the period 0 level: the level of period t total factor productivity of industry i in period t (relative to period 0), T it , the growth in real output prices for industry i going from period 0 to t, Ait , and the growth in primary inputs utilised by industry i going from period 0 to t, B it . The nominal value of market sector output in period t is the corresponding sum of industry nominal values,  I it i=1 P  · y it , which can be converted into  the period t real income generated by the market sector, ρt , by dividing this sum by the period t consumption price deflator, PCt :  ρt ≡  I i=1  P it · y it = PCt  I  pit · y it =  i=1  I  ρit ;  t = 0, 1, . . .  (1.86)  i=1  where the last equality follows using (1.76). Define industry i’s share of market sector nominal (or real) net output in period 0 as, s0i ≡  ρi0 ; ρ0  i = 1, . . . , I  (1.87)  Using the above definitions, we can decompose the growth in market sector real income, going from period 0 to t, as follows,  106  Chapter 1. Productivity Performance of Canada  t  0  ρ /ρ  I it i=1 ρ  =  using (1.86)  ρ0 I  = i=1 I  ρit ρi0 s0i  = i=1 I  =  ρi0 ρ0  ρit ρi0  s0i T it Ait B it  using (1.87)  using (1.85)  (1.88)  i=1  Equation (1.88) shows the factors that determine the evolution of market sector real income growth over time. There are four sets of factors at work: • The industrial structure of net product in the base period; i.e., the base period industry shares of market sector net output, s0i ; • The total factor productivity performance of industry i cumulated from the base period to the current period; i.e., the industry productivity factors, T it ; • The growth in industry output prices (deflated by the price of the consumption aggregate) going from period 0 to t; i.e., the industry real output price factors, Ait and • The growth in primary inputs utilised by industry i going from period 0 to t; i.e., the industry primary input growth factors, B it . Note that if an industry i experiences growth in its (net) output prices relative to the price of consumption, then the corresponding real output price factor Ait will be greater than one and this effect will contribute to overall real income growth. Traditional Total Factor Productivity decompositions does not include this type of factor; i.e., the traditional analysis ignores  107  Chapter 1. Productivity Performance of Canada favourable (or unfavourable) output price effects.62  62  Improvements in the country’s terms of trade are also ignored by the traditional methodology. This does not mean that the traditional emphasis on pure efficiency improvements is “wrong”; it just does not answer the question that we are focusing on, which is: what is the rate of growth in consumption equivalents that the market sector of the economy is generating?  108  Chapter 2  Business Sector Data on Outputs and Inputs for Canada 1961-2007 2.1  Introduction  The basic approach to measuring productivity growth is to use recently released information on business sector outputs and inputs from Statistics Canada’s KLEMS database along with information on aggregate final demand expenditures in order to construct “top down” measures of the productivity performance of the Canadian business sector.63 We also make extensive use of Statistics Canada’s National Balance Sheet estimates for information on various capital inputs used by the business sector. Thus the present approach to productivity measurement is an aggregate “top down” approach as opposed to the usual industry “bottom up” approach which makes use of detailed data on inputs used and outputs produced by industrial sectors and aggregates up sectoral productivity growth rates in order to obtain national business sector estimates.64 With reliable data, the two approaches should give very similar answers.65 Unfortunately, data on in63  The database used in Chapter 1.8 was constructed in October, 2008. The “bottom up” approach is used by the Statistics Canada KLEMS program; see Baldwin, Wu and Yan (2007) for an overview and Baldwin and Gu (2007) for additional information on the construction of the Statistics Canada KLEMS capital services aggregates. 65 In fact, if indirect tax effects could be ignored and if nominal and real input output tables were perfectly consistent, the two approaches should give exactly the same answer; see Diewert (2006b)(2007c). 64  109  Chapter 2. Business Sector Data on Outputs and Inputs for Canada 1961-2007 dustry inputs and outputs are not likely to be as reliable as the corresponding national data for a variety of reasons66 so it is useful to provide a check on the industry approach to productivity measurement by using the national aggregate approach. There is another reason for undertaking a productivity study using final demand data and this reason is that the effects of changes in a country’s terms of trade can be measured in this framework whereas these effects cannot be measured in the industry accounts framework using the existing System of National Accounts 1993 (SNA 1993); see Chapter 15 in Eurostat, IMF, OECD, UN and World Bank (1993). In particular, the Input Output accounts as outlined in Table 15.1 in the SNA 1993 do not show the role of international trade in goods and services by industry. Exports and imports enter the main supply and use tables (Table 15.1) as additions (or subtractions) to total net supply or to total domestic final demand in the familiar C +I +G+X −M setup. This means that Table 15.1 in the main production accounts of SNA 1993 does not elaborate on which industries are actually using the imports or on which industries are actually doing the exporting by commodity.67 Thus at present, data difficulties prevent us from looking at the effects of changes in the terms of trade using the “bottom up” industry aggregation approach. Diewert and Lawrence (2000) undertook a study of Canada’s business sector productivity using the national approach for the years 1962-1996 and Diewert (2002) extended their data to cover the years 1962-1998. The study in Chapter 1 is an extension of these previous studies but there are some differences: • Statistics Canada has provided new data on national expenditure aggregates back to 1961 using annual chained index numbers and so it is 66  For a detailed discussion of these reasons, see Diewert (2001). It should be noted that SNA 1993 does have a recommended optional Table 15.5 which is exactly suited to our present needs; i.e., this table provides the detail for imports by commodity and by industry. However, SNA 1993 does not provide a recommendation for a corresponding commodity by industry table for exports. 67  110  Chapter 2. Business Sector Data on Outputs and Inputs for Canada 1961-2007 no longer necessary to work with the old fixed base data on the most disaggregated level possible and then use chain indexes to aggregate up these data. • Statistics Canada has also provided new data on the outputs produced and inputs used by the Canadian business sector back to 1961 using chained Fisher indexes as part of their KLEMS productivity measurement program. In particular, we will use the KLEMS estimates of labour input, which are a big improvement over the estimates of labour input used by Diewert and Lawrence. • Diewert and Lawrence (2000) worked with a rather narrow definition of the government sector; their definition included only the public administration industry. In this study, we adopt the Statistics Canada definition of the non-business sector (except that we add to it the residential rental housing industry) and include the general government sector and the publicly funded defence, hospital and education sectors in the non-business sector.68 Since output in the non-business sector is measured by input, the use of the broader definition of the government sector should lead to higher estimates of productivity growth in the business sector compared to the estimates tabled in Diewert and Lawrence (2000) and Diewert (2002). • Statistics Canada has reorganized its information on indirect taxes (less subsidies) into two categories: taxes that fall primarily on outputs and taxes that fall primarily on inputs. This new information is very useful in making adjustments to output prices for indirect tax effects.69 68  The non-business sector consists of the following industries: (1) Government funding of hospitals; (2) Government funding of residential care; (3) Government funding of universities; (4) Government funding of other education; (5) Defence services; (6) Other municipal government services; (7) Other provincial government services and (8) Other federal government services. 69 In early studies of the Total Factor Productivity of an economy like those done by Solow (1957) and Jorgenson and Griliches (1967), outputs were priced at final demand prices, which include indirect taxes. However, Jorgenson and Griliches (1972; 85) noted that this treatment was not consistent with competitive price taking behaviour on the part of producers, since producers do not derive any benefit from indirect taxes that fall  111  Chapter 2. Business Sector Data on Outputs and Inputs for Canada 1961-2007 There had been a substantial revision of the Statistics Canada data used in Diewert (2008), therefore we use in this paper more recent published Statistics Canada data70 as well as some unpublished disaggregated capital data made available to us from the Statistics Canada KLEMS database. Note that the business sector used here differs from the Statistics Canada business sector in that we have excluded all residential housing services (Owner Occupied Housing services plus Rental Housing services) from our business aggregate whereas Statistics Canada includes the services of rental housing in its business aggregate.71 The main conceptual changes in our present database from the data tabled in Diewert (2008) are as follows: • The trade data were disaggregated; • Machinery and equipment investment in Diewert (2008) has been disaggregated into ICT machinery and equipment and non-ICT machinery and equipment;72 • We used the Statistics Canada KLEMS data on the price of inventory stocks in the present study whereas before, we used another Statistics Canada price series to value inventory stocks; • The depreciation rates for non-residential structures, ICT machinery and equipment and non-ICT machinery and equipment were reestimated using balance sheet and KLEMS program information along with revised investment information. on their outputs and thus these taxes should be removed. 70 These data were obtained in October 2008. 71 Our reason for excluding the services of rental housing from our business sector aggregate is due to the lack of accurate data on residential structures investment on rental housing and the lack of information on the quantity and value of land that is occupied by rental housing. Our measure of business sector labour input is exactly the same as that used by the Statistics Canada KLEMS program so only our output measure and capital services input measures differ from the corresponding KLEMS estimates. 72 Our thanks to Wulong Gu for making the disaggregated data available to us.  112  Chapter 2. Business Sector Data on Outputs and Inputs for Canada 1961-2007 In Section 2.2, we will list the basic final demand expenditure series that were used. Section 2.3 simply lists the three published business sector measures of quality adjusted labour input for the Canadian business sector that are available on CANSIM as part of the Statistics Canada KLEMS program. Section 2.4 studies the problems associated with forming estimates for capital inputs. Section 2.5 forms estimates of tax rates on primary inputs. This information is used to calculate estimates of balancing after tax real rates of return. Then this information is used along with the information developed in previous subsections in order to calculate user costs for five classes of capital input: machinery and equipment, non-residential structures, agricultural land, non-agricultural and non-residential business land and inventories. Section 2.6 concludes with some observations on the weak points in the data and recommendations for further work on developing a set of productivity accounts for Canada.  2.2  Estimates of Canadian Final Demand Expenditures  Much of the information tabled in this section is updated information that can be found in the Canadian Economic Observer, Statistics Canada (2007), Table 1: Gross Domestic Product (GDP) by Income and Expenditure (millions of dollars and in chained 2002 dollars). The October 2008 version of these data were used, using the Statistics Canada online data service CANSIM II, which were listed as quarterly data. If the quarterly data were seasonally adjusted, then the data for a year were summed and divided by four in order to obtain annual data. If the quarterly data were not seasonally adjusted, then they were simply summed in order to obtain annual data. In what follows, we will use the CANSIM individual series label to identify the exact series used. The first two series are Personal Expenditures on Goods and Services in current and constant chained 2002 dollars, CANSIM II series V498087 and V1992044 respectively. Dividing the current dollar series VCT by the 113  Chapter 2. Business Sector Data on Outputs and Inputs for Canada 1961-2007 constant dollar series QCT gives us an implicit price series PCT for personal consumption. We would like to exclude the imputed expenditures on Owner Occupied Housing (OOH) from the above series since there is no possibility of productivity gains occurring in this sector. However, if we exclude imputed rent from the business sector output series, we also need to exclude the services of the owner occupied housing capital stock as an input into the business sector. Unfortunately, we are not able to construct a reliable measure of the Owner Occupied Housing capital stock from available data; we can only construct a more reliable residential housing capital stock which includes the housing capital stock that is rented. We also were not able to split residential land input into reliable owner occupied and rental components.73 Hence we excluded both imputed and paid rents from our list of business sector outputs and we excluded the entire residential housing stock and the associated land as inputs into the business sector.74 Information on current dollar expenditures on imputed rents and paid rents (this is the series VP R in Table 2.1) for the years 1961-2007 is available from CANSIM II series V498532 and V498533 respectively. The corresponding information on chained 1997 constant dollar expenditures on imputed rents and paid rents (QP R ) is available from CANSIM II series V1992078 and V1992079 only for the years 1981-2006.75 We divide VP R by QP R in order to form 73 The determination of the structures and land inputs into the production of rented residential housing is a difficult task since the investment data on residential housing is not decomposed into owned and rented investments. This lack of information was also a problem for the Statistics Canada KLEMS program: “Data on investment in rental residential buildings are not available. For the annual MFP programs, we divide the total investment in residential building into rental building and owner-occupied dwelling using paid rents for rental buildings and imputed rents for owner occupied dwelling as the split ratios. The investment in residential buildings and paid and imputed rents are available from the Income and Expenditure Accounts. On average, we find that about 30% of total rents are paid rents and the remaining 70% are imputed rents.” Baldwin, Gu and Yan (2007; 43). 74 This means our productivity estimates will be biased downward slightly since the inputs that are used in the rental housing market are included in our estimates but the corresponding outputs are not included. 75 We did construct the corresponding expenditure based price series for imputed rents for the period 1981-2006 and compared this price index with the corresponding industry  114  Chapter 2. Business Sector Data on Outputs and Inputs for Canada 1961-2007 a price index for paid rents, PP R . We could follow the same strategy to form a price index for imputed rents for the years 1981-2007.76 However, an alternative series on the imputed value of OOH services for the years 1961-2004 is available from the industry accounts. This series is CANSIM II series V3859926, Business Sector: Owner Occupied Dwellings, from Table 370023: Gross Domestic Product (GDP) at Basic Prices in Current Dollars, System of National Accounts, Benchmark Values, by North American Industry Classification System (NAICS) and is listed as VIM R in Table 2.1.77 The final demand value series for imputed rents (not listed) is about 13% higher than its industry counterpart, VIM R . We use the industry series for imputed rents rather than the final demand series because we want our business sector value added to closely approximate the Statistics Canada KLEMS program business sector value added, except that our aggregate will not include paid residential rents.78  based price index for imputed rents described below for the years 1981-2004 and found that the movements were similar. We used the expenditure based price index for the years 2004-2007 to extend the industry based price index from 2004 to 2007. 76 We explain below how this industry based value series for imputed rents was extended from 2004 to 2007. 77 We explain below how this industry based value series for imputed rents was extended from 2004 to 2007. 78 The KLEMS business sector value added aggregate excludes imputed rents whereas our business sector value added aggregate will exclude both imputed and paid rents. Our treatment of inventory change is also different.  115  Chapter 2. Business Sector Data on Outputs and Inputs for Canada 1961-2007 Table 2.1: Housing Value, Quantity and Price Series for Imputed and Paid Rents  Year t  t VIM R  QtIM R  VPt R  QtP R  t PIM R  t PP R  1961  2292  2292  1107  1107  1.00000  1.00000  1962  2436  2380  1176  1149  1.02350  1.02350  1963  2660  2412  1290  1170  1.10275  1.10275  1964  2832  2477  1396  1221  1.14316  1.14316  1965  2976  2531  1503  1278  1.17565  1.17565  1966  3249  2620  1658  1337  1.23992  1.23992  1967  3585  2678  1860  1390  1.33856  1.33856  1968  3985  2707  2091  1420  1.47212  1.47212  1969  4416  2784  2342  1476  1.58633  1.58633  1970  4897  2833  2645  1530  1.72855  1.72855  1971  5388  2864  2918  1551  1.88118  1.88118  1972  5757  2866  3183  1584  2.00889  2.00889  1973  6307  2862  3451  1566  2.20366  2.20366  1974  7107  2923  3787  1558  2.43126  2.43126  1975  8313  2992  4290  1544  2.77854  2.77854  1976  10038  3072  4842  1482  3.26746  3.26746  1977  12126  3084  5443  1384  3.93199  3.93199  1978  14090  3051  6106  1322  4.61807  4.61807  1979  15797  2996  6829  1295  5.27283  5.27283  1980  17869  3053  7686  1313  5.85278  5.85278  1981  20512  3159  8822  1359  6.49322  6.49322  1982  23489  3213  10082  1410  7.31046  7.15154  1983  26285  3256  11295  1444  8.07270  7.82159  1984  28446  3294  12181  1471  8.63567  8.28079  1985  30694  3360  12967  1500  9.13517  8.64482  1986  33386  3463  13955  1539  9.64089  9.06928  1987  36117  3573  15090  1599  10.10837  9.43653  Continued on Next Page. . .  116  Chapter 2. Business Sector Data on Outputs and Inputs for Canada 1961-2007 Table 2.1 – Continued Year t  t VIM R  QtIM R  VPt R  QtP R  t PIM R  t PP R  1988  39587  3801  16419  1662  10.41493  9.87670  1989  44078  4011  18201  1726  10.98935  10.54481  1990  48016  4221  19786  1798  11.37552  11.00446  1991  51779  4469  21133  1853  11.58636  11.40566  1992  54872  4627  22269  1899  11.85900  11.72872  1993  57263  4770  23108  1943  12.00486  11.89235  1994  60557  4887  24056  1982  12.39142  12.13540  1995  63613  5001  24869  2016  12.72013  12.33820  1996  65418  5116  25632  2049  12.78691  12.51068  1997  67405  5245  26425  2097  12.85127  12.59838  1998  69835  5389  27223  2139  12.95872  12.72809  1999  72144  5557  28173  2187  12.98263  12.87911  2000  74582  5704  29059  2231  13.07545  13.02515  2001  77093  5843  30092  2279  13.19410  13.20509  2002  80895  6074  31491  2341  13.31831  13.44940  2003  83916  6250  32829  2413  13.42651  13.60407  2004  87614  6482  34133  2487  13.51648  13.72279  2005  91546  6730  35435  2560  13.60266  13.83987  2006  96714  6985  37137  2638  13.84590  14.07851  2007  103152  7246  39262  2714  14.23568  14.46634  117  Chapter 2. Business Sector Data on Outputs and Inputs for Canada 1961-2007 We now describe how we estimated a price index for the paid rents series for the years 1961-1981 and how we formed a price index for the industry value added series for the imputed rents for OOH for the years 1961-2007. An old series for the industry value added generated by OOH, CANSIM II series V334072, Canada: Current Prices; Business Sector; Owner Occupied Dwellings, from Table 3790001, Gross Domestic Product (GDP) at Factor Cost, System of National Accounts Benchmark Values, by Industry, is available for the years 1961-1997. The corresponding series in constant 1992 dollars is available for the years 1961-2000 as CANSIM II Series V328857 in Table 3790004. We use these two series to form a price index for imt puted rent for the years 1961-1997, PIM R in Table 2.1. A constant dollar  industry series for the services of OOH for the years 1997-2007 can be obtained from CANSIM II Series V14183160, Canada; Seasonally Adjusted at Annual Rates; Chained 1997 Dollars; Owner Occupied Dwellings in Table 3790018, Gross Domestic Product (GDP) at Basic Prices by NAICS.79 t Dividing VIM R by this constant dollar series will give us a price index for  imputed rents running from 1997 to 2007 and we link this series to the earlier t PIM R series that ran from 1961 to 1997. We then normalized the price series  t to equal 1 in 1961 and formed the quantity series QtIM R as VIM R divided  t t t t by PIM R . VIM R , QIM R and PIM R are listed in Table 2.1. Recall that we  have a value series for paid rents, VPt R , that covers the years 1961-2007 but  the corresponding price index series, PPt R , covers only the years 1981-2007. t We extend PPt R back to 1961 using the movements in PIM R . The resulting  price series is normalized to equal 1 in 1961 and a quantity series for paid rents, QtP R , is obtained by dividing VPt R by PPt R . These three series are also listed in Table 2.1.80 79 Somewhat mysteriously, this constant dollar series extends all the way to 2007 whereas the corresponding current dollar series ends at 2004. As noted above, we extended the industry price index for imputed rents from 2004 to 2007 using the movements in the corresponding expenditure based price index for imputed rents over the years 2004-2007. Given this extended price index plus the industry based constant dollar series for imputed rents, the industry based value series for imputed rents was extended to 2007. 80 The units for all value and quantity series are millions of current dollars for the V series and millions of 1961 dollars for the Q series.  118  Chapter 2. Business Sector Data on Outputs and Inputs for Canada 1961-2007 Recall the price and quantity series for a consumption aggregate (which includes all rents, paid and imputed), PCT and QCT , along with the two price and quantity series for imputed and paid rents in Table 2.1. We changed the sign of the rent quantity series from plus to minus and then calculated a chained Fisher net consumption aggregate by aggregating all consumption (plus sign on the quantities) and rents (negative sign on the quantities). The resulting price and quantity series should closely approximate the price and quantity of consumption excluding housing services. However, the price series includes indirect taxes (less subsidies) on outputs but for productivity measurement purposes, as mentioned earlier, these tax wedges should be excluded. Statistics Canada has a series for indirect taxes less subsidies on t , CANSIM II series V1997473, for the years 1961-2007. We products VIT  subtracted two other tax series from this indirect tax series because these other tax series will be taken into account separately in the price of exports of goods (this is the Oil Export Tax series, CANSIM series V499746) and in the price of imports of goods (this is the Customs Import Duties series, CANSIM series V499741). The resulting indirect taxes less subsidies on products (less trade taxes) series was used to remove the tax wedges on the price of consumption series. The resulting price and quantity of consumption series, PCt and QtC , are listed in Tables 2.2 and 2.3.81  81  We renormalize all price and quantity series so that the normalized price is 1 in 1961. The units for quantity and value series are in millions of current and 1961 dollars respectively.  119  Chapter 2. Business Sector Data on Outputs and Inputs for Canada 1961-2007 Table 2.2: Prices Indexes for Business Sector Outputs: Consumption and Investment  Year  t PC  t PIG  t PIR  t PICT  t PIM E  t PIN R  t PII  1961  1.00000  1.00000  1.00000  1.00000  1.00000  1.00000  1.00000  1962  1.00538  1.00855  1.00504  0.99939  1.01477  1.00592  1.00240  1963  1.02055  1.03939  1.02769  0.99650  1.06598  1.03251  1.01346  1964  1.02437  1.06231  1.07312  1.00502  1.06773  1.06158  1.03169  1965  1.03690  1.13926  1.13368  1.02393  1.10198  1.12281  1.05486  1966  1.07553  1.21007  1.20765  1.02944  1.12251  1.19323  1.07231  1967  1.11050  1.23160  1.28518  1.07473  1.12792  1.24188  1.08449  1968  1.15168  1.23844  1.31431  1.10902  1.13973  1.25227  1.10971  1969  1.18980  1.28464  1.38118  1.14444  1.17197  1.32495  1.13490  1970  1.22208  1.33877  1.42615  1.19472  1.23029  1.39058  1.15046  1971  1.24828  1.40873  1.53179  1.22722  1.27216  1.46812  1.20264  1972  1.29847  1.48528  1.67349  1.26956  1.30444  1.55098  1.31611  1973  1.38744  1.64838  1.97123  1.30782  1.34535  1.71873  1.42683  1974  1.58382  2.01078  2.36134  1.35698  1.51430  2.03419  1.54691  1975  1.82198  2.23421  2.56072  1.45867  1.73112  2.27337  1.65112  1976  1.90726  2.34499  2.76853  1.45190  1.82918  2.40093  1.77007  1977  2.03175  2.48990  2.87768  1.44467  1.98862  2.52980  1.92666  1978  2.19264  2.66206  3.04069  1.40748  2.19453  2.71145  2.13561  1979  2.40645  2.88522  3.28046  1.38366  2.44348  2.96312  2.34120  1980  2.69497  3.19858  3.55455  1.21888  2.69736  3.32520  2.55883  1981  2.95335  3.68313  3.99273  1.12880  2.98813  3.68676  2.75849  1982  3.22860  3.92658  4.08226  1.16812  3.19883  3.96113  2.91040  1983  3.46323  4.01498  4.25350  1.02222  3.26756  3.93090  3.04428  1984  3.61506  4.17063  4.41785  0.96342  3.40812  4.08142  3.12590  1985  3.72257  4.20827  4.55564  0.89167  3.57237  4.21351  3.15838  1986  3.80422  4.20267  4.90827  0.82129  3.70845  4.27520  3.20668  1987  3.89726  4.22375  5.40819  0.75277  3.69804  4.47320  3.26271  Continued on Next Page. . .  120  Chapter 2. Business Sector Data on Outputs and Inputs for Canada 1961-2007 Table 2.2 – Continued Year  t PC  t PIG  t PIR  t PICT  t PIM E  t PIN R  t PII  1988  4.00205  4.33769  5.78293  0.71401  3.69075  4.72840  3.31521  1989  4.11690  4.43728  6.13195  0.64691  3.79390  4.92520  3.36284  1990  4.35206  4.53066  6.11231  0.61403  3.87130  5.08853  3.36458  1991  4.59099  4.31837  6.32257  0.54586  3.74170  5.00311  3.51938  1992  4.65258  4.31896  6.39710  0.51043  3.88288  4.97541  3.52557  1993  4.74252  4.34342  6.58445  0.50164  4.04523  5.03758  3.70024  1994  4.77089  4.42033  6.76485  0.48119  4.24754  5.20497  3.83041  1995  4.79147  4.51572  6.76717  0.44755  4.44740  5.27332  3.91857  1996  4.88952  4.53812  6.75581  0.41150  4.57050  5.43035  3.78206  1997  4.96547  4.57906  6.87512  0.39399  4.69221  5.56694  3.80677  1998  5.03224  4.59706  6.95993  0.36919  4.90374  5.71450  3.86651  1999  5.12045  4.57201  7.13210  0.33873  4.94358  5.82995  3.95640  2000  5.25425  4.68967  7.29782  0.32384  5.01831  6.02775  4.04039  2001  5.40970  4.68012  7.48766  0.31733  5.17017  6.07934  4.10861  2002  5.47743  4.72977  7.81242  0.30560  5.25354  6.18175  3.87128  2003  5.61543  4.72659  8.21290  0.27567  4.94953  6.30506  3.91048  2004  5.69551  4.79882  8.71618  0.24913  4.85253  6.70389  3.95736  2005  5.81654  4.91308  9.11452  0.23077  4.79667  7.12403  4.02157  2006  5.92386  5.09878  9.78750  0.21439  4.71044  7.64020  4.14724  2007  6.02712  5.28795  10.49192  0.20857  4.58255  8.04719  4.15442  121  Chapter 2. Business Sector Data on Outputs and Inputs for Canada 1961-2007 Table 2.3: Quantity Indexes for Business Sector Outputs: Consumption and Investment  QtC  QtIG  QtIM E  QtIN R  1961  20265  1887  2211  312  1925  2618  1  1962  21331  2094  2271  358  2067  2545  560  1963  22290  2101  2354  399  2199  2637  936  1964  23529  2141  2715  412  2744  3050  532  1965  24974  2426  2825  472  3257  3320  1228  1966  26240  2668  2699  580  3899  3802  1313  1967  27228  2718  2754  617  3866  3613  454  1968  28525  2758  3132  592  3537  3593  619  1969  29923  2700  3551  673  3817  3592  1580  1970  30450  2645  3254  677  3868  3946  180  1971  32321  2985  3728  744  3924  4089  −232  1972  34891  2938  4066  806  4338  4074  148  1973  37676  2781  4371  932  5336  4396  2542  1974  39789  2845  4464  1167  5925  4675  4839  1975  41468  2962  4386  1310  6089  5286  −334  1976  43911  2855  5172  1568  6219  5168  250  1977  45480  2916  5242  1647  6057  5479  1319  1978  47255  2875  5291  1895  6202  5626  1300  1979  48694  2803  5251  2225  6988  6337  3691  1980  49521  2869  4977  3074  7165  7055  167  1981  49914  2967  5279  4423  8211  7620  891  1982  48210  3095  4340  4205  6867  6929  −4603  1983  49567  2991  5079  5184  6414  6361  −1049  1984  51955  3124  5131  6297  6397  6288  1744  1985  54842  3497  5578  7438  6909  6590  1467  1986  56973  3485  6267  8984  7454  6210  1029  1987  59433  3621  7190  12586  8084  6454  1701  Year t  QtIR  QtICT  QtII  Continued on Next Page. . .  122  Chapter 2. Business Sector Data on Outputs and Inputs for Canada 1961-2007 Table 2.3 – Continued QtC  QtIG  1988  61835  3789  7340  1989  63760  4184  7640  Year t  QtIR  QtICT  QtIM E  QtIN R  QtII  14111  9704  7110  2078  17019  10096  7345  1406  1990  64001  4463  6835  18392  9331  7346  −583  1991  62103  4692  5824  19836  8856  7075  −3899  1992  62821  4621  6238  24332  7995  5960  −551  1993  63783  4560  6024  26001  7449  5993  −878  1994  65822  4894  6271  30044  8001  6533  709  1995  67161  4740  5340  33319  8373  6574  3164  1996  68967  4536  5852  39474  8500  6696  2417  1997  72495  4390  6330  49575  10434  7881  1780  1998  74536  4361  6106  60330  10647  7906  2329  1999  77521  5039  6324  72466  11106  8101  1249  2000  80901  5229  6656  85220  11270  8266  2915  2001  82720  5830  7363  84440  10798  8712  −1856  2002  85721  6044  8403  83843  10625  8195  3367  2003  88311  6370  8854  90145  11417  8651  −1722  2004  91157  6773  9519  105113  12185  9257  1058  2005  94522  7521  9845  115589  13473  10184  1981  2006  98708  8021  10061  130883  14066  11093  1104  2007  103366  8644  10363  140217  15069  11047  1757  123  Chapter 2. Business Sector Data on Outputs and Inputs for Canada 1961-2007 We turn our attention to the investment components of final demand. Current dollar government gross fixed capital formation is available as CANSIM II series V498093 for the years 1961-2007. The corresponding chained 2002 dollar series is CANSIM II series V1992050 and we use these two series to form price and quantity series for general government sector investment, t and Qt , which are listed in Tables 2.2 and 2.3.82 PIG IG  The current and constant chained dollar series for the years 1961-2006 for residential structures investment can be obtained as CANSIM II series V498096 and V1992053 respectively, the current and constant chained dollar series for non-residential structures investment can be obtained as CANSIM II series V498098 and V1992053 respectively and the resulting price and t , Pt t t quantity series are denoted by PIR IN R , QIR and QIN R and are listed in  Tables 2.2 and 2.3. Statistics Canada provided us with unpublished series on the price and value of ICT and non-ICT machinery and equipment investments for the years 1961-2006. The resulting price and quantity series for ICT investment and non-ICT machinery and equipment are denoted by t t t 83 as listed in Tables 2.2 and 2.3. , QtICT , PIM PICT E and QIM E respectively,  t These tables also include the price and quantity of inventory change, PII  and QtII change but the description of how they were constructed is deferred until we discuss how we formed estimates of the beginning of the year stocks of inventories. 82  The price series for investment should be adjusted for indirect taxes that fall on investment outputs. Since these taxes are relatively small and it is difficult to collect consistent information on these taxes over our sample period, we neglect these indirect tax wedges on investment components of final expenditure. 83 We used the rates of increase in the price and quantity of all investment in machinery and equipment going from 2006 to 2007 from the published Statistics Canada expenditure accounts (see the current and constant chained dollar series for machinery and equipment investment, CANSIM II series V498099 and V1992056 respectively) to extend the ICT and non-ICT machinery and equipment price and quantity series to 2007. When we aggregated up the two unpublished ICT and other machinery and equipment investment series using chained Fisher indexes and compared the resulting aggregate KLEMS based machinery and equipment investment series with the corresponding national accounts based series, we found that the value series were very close but the KLEMS based price series grew 2.32 fold over the years 1961-2006 whereas the national accounts based price series grew 2.60 fold, which is 12% higher. We chose to use the KLEMS based data series for investment in machinery and equipment.  124  Chapter 2. Business Sector Data on Outputs and Inputs for Canada 1961-2007 All of the outputs described above can be regarded as outputs produced by the business sector and sold to final demanders. However, the business sector also sells goods and services to the non-business sector and it also purchases smaller amounts of goods and services from the non-business sector. We now describe how we formed price and quantity estimates for the net sales of the business sector to the non-business sector. For the years 1961-2007 from the National Income and Expenditure Accounts, CANSIM II series V498092; Government Current Expenditure on Goods and Services, Table 3800002, we have estimates of total government gross current expenditure on goods and services (less sales of goods and services to the business sector) in current dollars. From the same table and for the same years, CANSIM II series V1992049; Government Current Expenditure on Goods and Services, Table 3800002, we have estimates of total government gross current expenditure on goods and services (less sales of goods and services to the business sector) in chained 2002 dollars. We use these two series to form price and quantity series for final demand government sector expenditures, PGt and QtG , which are listed in Table 2.4.  125  Chapter 2. Business Sector Data on Outputs and Inputs for Canada 1961-2007 Table 2.4: Business Sector, Non-business Sector, Government Final Demand and KLEMS Business Sector Price and Quantity Aggregates  QtB  QtN  QtG  1961  33097  5204  6624  30805  1.00000  1.00000  1.00000  1.00000  1962  35338  5480  6928  33059  1.00919  1.03863  1.02916  1.00509  1963  37217  5713  7164  35013  1.02992  1.08205  1.05990  1.01881  1964  39810  5952  7542  37567  1.04877  1.14227  1.09761  1.03600  1965  42658  6120  7883  40122  1.07554  1.21527  1.15160  1.06938  1966  45529  6409  8581  42827  1.12248  1.34490  1.24333  1.11745  1967  46616  6870  9334  43728  1.16053  1.45671  1.32836  1.15516  1968  49335  7263  9944  46133  1.19304  1.54780  1.41575  1.18948  1969  51965  7585  10376  48537  1.23452  1.71317  1.53846  1.23074  1970  52968  7962  11287  49889  1.29437  1.84146  1.64279  1.27609  1971  55844  8255  11631  51843  1.33615  1.95964  1.75921  1.33535  1972  59086  8549  11995  54998  1.40300  2.12810  1.89268  1.40260  1973  63467  8887  12559  59206  1.54872  2.31234  2.04912  1.55366  1974  65346  9295  13357  61310  1.79635  2.65779  2.33927  1.79872  1975  65545  9790  14251  62061  2.03754  3.05325  2.65962  2.01795  1976  70082  10097  14525  66118  2.17572  3.45337  2.99471  2.15434  1977  72425  10348  15205  68823  2.32657  3.73913  3.24567  2.27212  1978  74875  10644  15473  71979  2.51505  3.96850  3.45708  2.42047  1979  77878  10805  15635  75134  2.80145  4.29236  3.77635  2.69350  1980  79169  11138  16169  76938  3.12982  4.66836  4.14895  2.98837  1981  81847  11496  16441  80244  3.40152  5.22919  4.64970  3.21385  1982  78970  11693  16767  77088  3.65996  5.83630  5.18504  3.44460  1983  81077  11952  17045  79192  3.89493  6.12278  5.48059  3.65571  1984  86041  12198  17222  84752  4.03944  6.37296  5.69544  3.76522  1985  90944  12471  17959  89260  4.13895  6.57967  5.90593  3.87314  1986  93580  12708  18283  91514  4.19906  6.81314  6.09362  3.92907  Year t  QtBKLEM S  t PB  t PN  t PG  t PBKLEM S  Continued on Next Page. . .  126  Chapter 2. Business Sector Data on Outputs and Inputs for Canada 1961-2007 Table 2.4 – Continued Year t  QtB  QtN  QtG  QtBKLEM S  t PB  t PN  t PG  t PBKLEM S  1987  97824  12840  18525  96022  4.38354  7.17180  6.36273  4.08969  1988  102723  13057  19370  100981  4.58086  7.53056  6.60377  4.26785  1989  105427  13224  19903  103685  4.75619  8.03351  6.95695  4.41096  1990  106128  13541  20605  103235  4.85194  8.60166  7.34874  4.52286  1991  104194  13849  21208  99027  4.90597  9.01919  7.64969  4.63908  1992  105171  14045  21414  99628  4.92162  9.36188  7.88200  4.64471  1993  108151  14150  21422  102483  4.97722  9.50865  7.98997  4.69377  1994  113766  14218  21156  108795  5.08269  9.55910  8.11082  4.75837  1995  117124  14279  21034  112401  5.23638  9.61980  8.19895  4.89045  1996  119744  14025  20786  114956  5.33957  9.72826  8.23447  4.99301  1997  125797  13787  20579  121417  5.40168  9.95401  8.34626  5.04144  1998  131475  13890  21240  127127  5.37429  10.07510  8.44225  5.00882  1999  139515  14320  21687  135542  5.47529  10.18237  8.57899  5.10349  2000  147808  14614  22356  144108  5.71191  10.65177  8.94989  5.34163  2001  149733  14926  23229  146512  5.80825  10.88586  9.11375  5.41622  2002  153895  15241  23802  150269  5.82601  11.29659  9.42889  5.42824  2003  156933  15608  24551  153124  6.02555  11.73665  9.71085  5.62745  2004  162130  15921  25044  158534  6.23215  11.96951  9.87845  5.82088  2005  167081  16154  25429  163342  6.45631  12.39824  10.23229  6.04364  2006  171718  16519  26385  168301  6.65413  12.80917  10.57143  6.21454  2007  175972  16945  27359  172358  6.89723  13.13041  10.83655  6.44335  127  Chapter 2. Business Sector Data on Outputs and Inputs for Canada 1961-2007 Recall that the Statistics Canada KLEMS productivity program business sector value added aggregate includes rental residential housing but excludes the services of owned residential housing (whereas our business sector value added aggregate excludes all forms of residential rents). The Industry Division of Statistics Canada produces yet another business sector estimate of nominal and real value added (at factor cost) which includes all residential rents, both imputed and paid. We will denote this value added aggregate by VBt in year t. Statistics Canada also produces a companion non-business sector value added aggregate (at factor cost) which we will denote by VNt in year t. If the value of indirect taxes less subsidies on products for year t, t , is added to the sum of these two industry value added aggregates, we VIT  get an estimate of the value of GDP at final demand prices in year t; i.e., we have the following identity: t t = VGDP VBt + VNt + VIT  (2.1)  We will now describe how we formed estimates for VBt and VNt along with the corresponding price and quantity decompositions. From Table 3790024, Gross Domestic Product (GDP) at Basic Prices in Current Dollars, SNA, Benchmark Values, Special Industry Aggregations Based on the North American Industry Classification System (NAICS), we can obtain the VBt series (title is Canada: Business Sector Industries) for the years 1961-2004 from CANSIM II Series V3860037. From the same Table 3790024, we can obtain the VNt series (title is Canada: Non-Business Sector Industries) for the years 1961-2004 from CANSIM II Series V3860040. We can obtain price indexes PBt , PNt and quantity indexes QtB , QtN for VBt and VNt for the years 1961-1997 by using the series V334562, V335071, V334565 and V335074 from CANSIM Table 3790002, Gross Domestic Product (GDP) at Factor Cost, System of National Accounts Benchmark Values by Industry (Special Aggregations). These series give business and non-business sector value added at basic prices in current dollars and in constant 1992 dollars. Using CANSIM Table 3790020, we can find estimates for QtB (Series V14182646)  and for QtN (Series V14182651) in chained 1997 dollars for the years 1997128  Chapter 2. Business Sector Data on Outputs and Inputs for Canada 1961-2007 2007. Hence using these series in conjunction with our earlier value series VBt  and VNt which run from 1961 to 2004, we can obtain price series for business and non-business sector value added at basic prices for the years 1997-2004. These price series can be linked to our earlier price series PBt and PNt which extended to 1997 so that the resulting price series will run from 1961 to 2004. However, we still do not have price or value series for the B and N sectors for 2004-2007, although we do have quantity series for these years. We extended the price series PNt from 2004 using an implicit price index for government goods and services, which was constructed using CANSIM Table 3800002, series V498092 in current dollars and series V1992049 for chained 2002 dollars. It turns out that the total of VBt and VNt is available in another CANSIM II series V3860274. Canada, Gross Domestic Product (GDP) at Basic Prices in Table 3800030: GDP and GNP at Market Prices and Net National Income at Basic Prices. Thus we have enough information to deduce the price PBt and the value of business sector output VBt for the years 2004-2007. The business and non-business sector price and quantity series, PBt , PNt and QtB , QtN for real value added at basic prices are listed in Table 2.4. It is also of some interest to compare the price and quantity of the above Industry Division business sector prices and quantities PBt and QtB  t with the corresponding business sector prices and quantities PBKLEM S and  QtBKLEM S that originate with the Statistics Canada productivity program.84  These series are also listed in Table 2.4. The source for QtBKLEM S for the  years 1961-2007 is CANSIM II series V41712932: Canada, Real Gross Domestic Product (GDP), Business Sector from Table 3830021: Multifactor Productivity, Value Added, Capital Input and Labour Input in the Aggregate Business Sector and Major Sub-sectors by the North American Industry Classification (NAICS). The corresponding nominal value added series t VBKLEM S is available in the same table for the years 1961-2004 as CANSIM 84 t Recall that the Productivity Program business sector value added aggregate VKLEM S should be equal to the Industry Division value added aggregate VBt less the value of t imputed rents from the Industry Division, VIM R.  129  Chapter 2. Business Sector Data on Outputs and Inputs for Canada 1961-2007 II series V41713153: Canada: Gross Domestic Product (GDP), Business t Sector. The values VBKLEM S for the missing years 2005-2007 can be obt tained by adding the value of imputed rents, VIM R , to the Industry Division t value added for the Business Sector, VBt . Finally, PBKLEM S can be obtained  t t by dividing VBKLEM S by QBKLEM S . As usual, we normalized the resulting  t price and quantity series so that PBKLEM S equals 1 when t equals 1961.  t t The resulting PBKLEM S and QBKLEM S are listed in Table 2.4.  Recall the GDP identity defined by (2.1), which expressed the nominal t , at final demand prices as being equal to the value value of GPD, VGDP  added of the Industry Division business sector value added at basic prices, VBt , plus non-business sector value added, VNt , plus the value of indirect  t . We can also express the value of taxes less subsidies on products, VIT  GDP at final demand prices as the familiar sum of final demand values; i.e., as the following sum of final demand expenditures on consumption plus investment plus government expenditures on goods and services plus exports less imports: t t t = VCT + VIt + VGt + VXt − VM VGDP  (2.2)  t as the value We define a new consumption aggregate at basic prices VCN  t , less indirect taxes less subsidies of consumption at final demand prices, VCT  t : on products, VIT  t t t VCN ≡ VCT − VIT  (2.3)  Now equate the two expressions for the value of GDP given by (2.1) and (2.2) and use the resulting equation to express business sector value added VBt in terms of final demand components and the value of non-business sector  value added VNt . Making use of (2.3), the resulting equation is the following one:85 85 The identity (2.4) is not quite consistent with our treatment of indirect taxes less subsidies since we also made some indirect tax adjustments to the prices of exports and imports as explained above; i.e., since we used a slight modification of (2.3) to adjust  130  Chapter 2. Business Sector Data on Outputs and Inputs for Canada 1961-2007  t t VBt = VCN + VIt + VXt − VM + (VGt − VNt )  (2.4)  Conceptually, the aggregate VGt − VNt should be equal to the sales of the business sector of goods and services to the non-business sector less the purchases of intermediate inputs of the business sector from the non-business sector. Put another way, the business sector’s net sales of goods and services t ) should equal its net deliveries to final demand sectors (VCt + VIt + VXt − VM  plus its net deliveries to the non-business sector (VGt − VNt ).  Recall that we did not use the Industry Division’s concept of Business Sector value added; we subtracted the value of imputed and paid residential rent from our business sector aggregate. Let VRt be equal to the sum of  t t imputed residential rent VIM R and paid residential rent VP R (see Table 2.1  t , we should for these series). Conceptually, if we subtract rents VRt from VCN  get VCt , the consumption aggregate whose price and quantity is listed in  Tables 2.2 and 2.3. Thus subtracting VRt from both sides of (2.4) leads to the following identity: t + (VGt − VNt ) VBt − VRt = VCt + VIt + VXt − VM  (2.5)  Thus our business sector value added aggregate can be formed using either the left or right hand sides of the identity (2.5). We will use the right hand side of (2.5) to form our value measure of business sector net output since we want to focus on the effects of changing international prices on the performance of the business sector. How should the corresponding real quantities that correspond to the value aggregates on either side of (2.5) be calculated? Obviously, each cell in the supply and use tables that correspond to the value aggregate on the left hand side of (2.5) could be aggregated up using a chained superlative index final demand consumption prices for indirect tax wedges, we used a corresponding slight modification of the identity (2.4).  131  Chapter 2. Business Sector Data on Outputs and Inputs for Canada 1961-2007 number formula provided that an appropriate price deflator were available for each cell.86 On the other hand, the value cells that are components on the right hand side of (2.5) that correspond to final demand components (at basic prices) could be aggregated up using a chained superlative index number formula. We can then ask: under what conditions would the corresponding quantity aggregates be equal? This question is addressed by Moyer, Reinsdorf and Yuskavage (2006) and in more detail by Diewert (2006b)(2007b)(2007c). The answer to this question is that if the detailed data are constructed in an appropriate manner and the Fisher formula is used, then the direct industry aggregation and the aggregation of final demand component approaches are perfectly consistent.87 In addition, if two stage aggregation procedures are used and a superlative index number formula is used at each stage of aggregation, then the theoretical and empirical results in Diewert (1978) show that the commonly used single stage superlative indexes will approximate their two or more stage counterparts to a high degree of approximation if the chain principle is used.88 Using the above results, we will construct our measure of business sector real value added by aggregating up the value components on the right hand side of (2.5). Rather than work with both VGt and VNt as final demand components, we will aggregate over these two components to form the value t equal to (VGt − VNt ), and conceptually, this value aggregate aggregate VGN  should be equal to the net deliveries of goods and services of our business sector to the non-business sector less the purchases of intermediate inputs by our business sector from the non-business sector. The year t price and t and QtGN , that correspond to these value aggrequantity aggregates, PGN  t are calculated using chained Fisher indexes with QtN getting a gates VGN  t and QtGN are listed in negative weight in the index number formula. PGN 86  Quantities in the Make matrix would have a positive sign while quantities in the Use matrix would have a negative sign. 87 See Diewert (2006b)(2007b) and the numerical examples in Diewert (2007c) in particular. 88 The results of Hill (2006) show that these approximation results will not necessarily hold for mean of order r superlative indexes if r is large in magnitude.  132  Chapter 2. Business Sector Data on Outputs and Inputs for Canada 1961-2007 Table 2.5.  133  Chapter 2. Business Sector Data on Outputs and Inputs for Canada 1961-2007 Table 2.5: Price Indexes for Business Sector Outputs: Net Sales to the Non-business Sector  Year t  t PGN  QtGN  1961  1.00000  1420  1962  0.99388  1447  1963  0.97566  1446  1964  0.92683  1596  1965  0.91239  1798  1966  0.88328  2320  1967  0.89087  2684  1968  0.96139  2950  1969  0.96769  3068  1970  1.00582  3858  1971  1.10291  3885  1972  1.14412  3942  1973  1.22092  4247  1974  1.35752  4819  1975  1.48446  5397  1976  1.64265  5254  1977  1.78697  5963  1978  1.92861  5833  1979  2.18540  5796  1980  2.48896  6062  1981  2.79431  5845  1982  3.10614  6019  1983  3.37412  5998  1984  3.48567  5838  1985  3.67204  6539  1986  3.74218  6635  1987  3.79747  6789  Continued on Next Page. . .  134  Chapter 2. Business Sector Data on Outputs and Inputs for Canada 1961-2007 Table 2.5 – Continued Year t  t PGN  QtGN  1988  3.78607  7814  1989  3.82606  8423  1990  3.86395  9043  1991  3.92576  9508  1992  3.94383  9458  1993  3.96988  9223  1994  4.17941  8537  1995  4.29853  8165  1996  4.20205  8263  1997  4.10280  8414  1998  4.13974  9511  1999  4.29033  9380  2000  4.43225  10022  2001  4.45850  11040  2002  4.56662  11443  2003  4.56905  12088  2004  4.60878  12331  2005  4.77387  12551  2006  4.93210  13651  2007  5.05578  14632  135  Chapter 2. Business Sector Data on Outputs and Inputs for Canada 1961-2007 We now turn our attention to the export and import components of final demand. Current dollar exports of goods are available as CANSIM II series V498104 for the years 1961-2007. The corresponding chained 2002 dollar series is CANSIM II series V1992061 and we use these two series to form price and quantity series for the exports of goods. However, in this study, we will form series for more detailed components of the exports and imports of goods. Current dollar exports of services are available as CANSIM II series V498105 for the years 1961-2007. The corresponding chained 2002 dollar series is CANSIM II series V1992062 and we use these two series to form t and Qt , which are price and quantity series for the exports of services, P16 16  listed in Tables 2.6 and 2.7. Our starting point for obtaining disaggregated data on the exports and imports of goods is CANSIM Table 3800012, Exports and Imports of Goods and Services, Canada, Current Prices. It is possible to obtain disaggregated information on the value of exports for the following seven classes for the years 1971-2007: • Q9 , Exports of agricultural and fish products; • Q10 , Exports of energy products; • Q11 , Exports of forest products; • Q12 , Exports of industrial goods and materials (excluding energy and forest product exports); • Q13 , Exports of machinery and equipment (excluding automotive products); • Q14 , Exports of automotive products and • Q15 , Exports of other consumer goods (excluding automotive products).  136  Chapter 2. Business Sector Data on Outputs and Inputs for Canada 1961-2007 The CANSIM series numbers for the first seven classes of exports are V498730-V498736. It is also possible to find corresponding constant dollar series in 1992 constant dollars over the period 1971-1997 in CANSIM Table 3800012 and the CANSIM series numbers are V498767-V498773. Finally, constant dollar chained estimates for these export categories (in 1997 chained dollars) can be found for the years 1981-2007 in CANSIM Table 3800012 and the series numbers are V1992162- V1992168. We used these series to form chained price and quantity series for these seven export categories for the years 1981-2007. The constant dollar price series were linked to each chained price series at the year 1981 in order to extend the chained series back to 1971.89 There remains the problem of obtaining price series for the above classes of exports to cover the years 1961-1971. From Leacy (1983), Series G415-428 Foreign trade, domestic exports, excluding coin and bullion, by main commodity sections, current values, we can obtain value series covering exports for the years 1946-1975 for the following five commodity classes: • Live animals (G415); • Food, feed, beverages and tobacco (G417); • Crude materials (inedible) (G419); • Fabricated materials (inedible) (G421); • End products (inedible) (G423). From the same source, price indexes for each of the above five classes of exports are available as Series K57-K61 in the Table with the title: Export price indexes, trade of Canada commodity classification, 1926-1975. Thus we can find price and quantity series for these five classes of exports that cover the years 1961-1971. Unfortunately, these price indexes are of the fixed base 89  There were two other categories in the export and import classifications: Special transactions and Other balance of payments adjustments. These categories were small and were omitted in our analysis.  137  Chapter 2. Business Sector Data on Outputs and Inputs for Canada 1961-2007 variety with a base year of 1948 so they are likely to differ substantially from the corresponding chain indexes (which are not publicly available). However, Leacy (1983) also lists as part of export price Series K57-K61 (Panel A) for the above 5 classes of exports some indexes that have a 1971 base year but these price indexes cover only the years 1968-1975. We use these latter price indexes to construct export price indexes for the years 1968-1971 and then we use the 1948 based indexes to further extend these five series back to 1961. The above operations give us five disaggregated export price and quantity series for the period 1961-1971 but we have seven classes of exports of goods for the years 1971-2007. We generated Fisher chained price and quantity indexes for exports of Live animals and for exports of Food, feed, beverages and tobacco for the years 1961-1971 and linked these series to our earlier series, P9t and Qt9 , exports of agricultural and fish products. But we need some additional series so that we can match the export and import series for the 1960s to the series that cover the post 1971 period. We will create separate export series for energy, forest products, automotive products and other consumer goods. Our sources for these extra series are the input output tables for the Canadian economy that cover the years 1961-1981 (see Statistics Canada (1987a)(1987b)). In order to create a price and quantity series for aggregate Energy exports for the years 1961-1971, we aggregated data for six classes of energy exports using the M level of aggregation: Coal, Crude mineral oils, Natural gas, Gasoline and fuel oil, Other petroleum and coal products and Electric power. These components were aggregated using Fisher (1922) chained indexes. The resulting price and quantity series were linked to our earlier price and t and Qt , for energy at the year 1971. quantity series, P10 10  In order to create aggregate Forestry exports for the years 1961-1971, we aggregated data for seven classes of forest product exports using the M level of aggregation: Lumber and timber, Veneer and plywood, Other wood fab138  Chapter 2. Business Sector Data on Outputs and Inputs for Canada 1961-2007 ricated materials, Furniture and fixtures, Pulp, Newsprint and other paper stock, and Paper products. These components were aggregated using Fisher (1922) chained indexes. The resulting price and quantity series were linked t and Qt , for forest product to our earlier price and quantity series, P11 11  exports at the year 1971. We aggregated the input output data for two classes of automotive product exports using the M level of aggregation: Motor vehicles and Motor vehicle parts. These components were aggregated using Fisher (1922) chained indexes. The resulting price and quantity series were linked to our earlier t and Qt , for automotive product exports at price and quantity series, P14 14  the year 1971. In order to create an aggregate for Exports of other consumer goods (excluding automotive products) for the years 1961-1971, we aggregated data for eight classes of consumer goods type exports using the M level of aggregation: Leather and leather products, Other textile products, Hosiery and knitted wear, clothing and accessories, appliances and receivers (households), Pharmaceuticals, Other chemical products and Other manufactured products. These components were aggregated using Fisher (1922) chained indexes. The resulting price and quantity series were linked to our earlier t and Qt , for exports of other consumer goods price and quantity series, P15 15  at the year 1971. We generated price and quantity series over the years 1961-1971 for Exports of industrial goods and materials (excluding energy and forest product t and Qt , as a chained Fisher aggregate of our price and quanexports), P12 12  tity series for Crude materials (inedible) (G419) and Fabricated materials t and Qt ) (inedible) (G421) less our series for exports of energy products P10 10 t and Qt ).90 The resulting export price and exports of forest products (P11 11 90  All four prices are entered as positive numbers in the index number formula while the first two quantities are entered positively and the last two quantities are entered negatively.  139  Chapter 2. Business Sector Data on Outputs and Inputs for Canada 1961-2007 and quantity series for the years 1961-1971 are linked to our earlier series t and Qt at the year 1971. for P12 12  Finally, we generated price and quantity series over the years 1961-1971 for Exports of machinery and equipment (excluding automotive products), t and Qt , as a chained Fisher aggregate of our price and quantity series P13 13  for exports of end products (inedible)(G423) less our series for exports of t and Qt ) and less exports of other consumer goods automotive products P14 14 t and Qt ).91 The resulting export price and quantity series for the years (P15 15 t and Qt at the year 1971. 1961-1971 are linked to our earlier series for P13 13  There is one additional adjustment which affects the price of energy exports. During the years 1974-1985, Canada imposed an export tax on its energy exports, which is included in the price of exports. However, producers do not receive this export tax revenue and so it must be subtracted from the export price. This adjustment of the export price index for exports of goods can be accomplished using the Oil Export Tax series, CANSIM series V499746 from the National Income and Expenditure Accounts. After t and making this adjustment, the resulting price and quantity series are P10  Qt10 , which are listed in Tables 2.6 and 2.7 along with the other price and quantity series for the eight classes of exports.  91  All three prices are entered as positive numbers in the index number formula while the first quantity is indexed with a positive sign and the last two quantities are indexed with negative signs.  140  Chapter 2. Business Sector Data on Outputs and Inputs for Canada 1961-2007 Table 2.6: Price Indexes for Eight Commodity Classes of Exports, 1961-2007  Year t  P9t  t P10  t P11  t P12  t P13  t P14  t P15  t P16  1961  1.00000  1.00000  1.00000  1.00000  1.00000  1.00000  1.00000  1.00000  1962  1.05659  0.98534  1.03706  1.01977  1.03488  1.01167  0.99880  1.01858  1963  1.05292  1.01104  1.04975  1.02746  1.04642  1.02781  1.00853  1.04402  1964  1.06146  0.99854  1.06857  1.04333  1.06452  1.03785  1.02909  1.07736  1965  1.07552  1.01886  1.08666  1.06777  1.09035  1.03617  1.03692  1.12159  1966  1.13905  1.02441  1.10752  1.12360  1.12856  1.05032  1.05855  1.18364  1967  1.15369  0.96541  1.12890  1.15308  1.17596  1.06607  1.07630  1.26762  1968  1.14873  1.00454  1.15804  1.22731  1.25191  1.08830  1.10360  1.34554  1969  1.11043  1.05284  1.21130  1.26468  1.28144  1.09898  1.14735  1.41327  1970  1.07460  1.06679  1.20448  1.33958  1.36481  1.12160  1.14015  1.50120  1971  1.10272  1.12301  1.24675  1.28936  1.37181  1.15193  1.15590  1.57903  1972  1.16574  1.15550  1.34600  1.30236  1.41712  1.17423  1.18547  1.66505  1973  1.65562  1.34434  1.61655  1.49521  1.46833  1.18656  1.25988  1.79018  1974  2.47603  2.37211  1.97968  1.94140  1.65047  1.27902  1.42184  2.02338  1975  2.44172  3.47049  2.31660  2.10541  1.86186  1.40313  1.56638  2.29526  1976  2.27544  4.36340  2.39386  2.22651  1.90587  1.48840  1.66746  2.51896  1977  2.13627  5.37242  2.62455  2.50906  2.02758  1.61582  1.76846  2.72893  1978  2.37658  5.96944  2.90165  2.77108  2.11566  1.78223  1.88273  2.90258  1979  2.88982  7.40098  3.47704  3.53655  2.27395  1.96622  2.00068  3.14587  1980  3.28365  10.39436  3.73797  4.45064  2.41362  2.17714  2.24946  3.49696  1981  3.56932  11.17124  3.96546  4.55036  2.55522  2.41373  2.49659  3.94719  1982  3.46707  11.69844  3.85284  4.31903  2.72280  2.63304  2.66992  4.34103  1983  3.39310  11.53824  3.83642  4.35230  2.75291  2.73815  2.80369  4.66353  1984  3.53005  11.27486  4.26336  4.45480  2.69323  2.90539  2.88654  4.87299  1985  3.51503  10.86479  4.32498  4.41174  2.68327  3.12972  3.00305  5.13320  1986  3.43645  7.75428  4.78008  4.59745  2.97757  3.01503  3.28669  5.40392  1987  3.32350  7.55588  5.20316  4.73208  3.02031  3.02347  3.40733  5.59379  Continued on Next Page. . .  141  Chapter 2. Business Sector Data on Outputs and Inputs for Canada 1961-2007 Table 2.6 – Continued Year t  P9t  t P10  t P11  t P12  t P13  t P14  t P15  t P16  1988  3.49449  6.35623  5.36339  5.05742  3.04258  2.91102  3.52049  5.72724  1989  3.73084  6.87199  5.59564  5.03255  3.06798  2.86952  3.71759  5.93246  1990  3.47655  7.84685  5.24427  4.74911  3.11615  2.87959  3.73982  6.08151  1991  3.14681  6.82156  4.69849  4.45427  3.07176  2.96769  3.82601  6.29162  1992  3.48817  7.07296  4.84492  4.47367  3.06279  3.15213  3.82959  6.33223  1993  3.71935  7.47530  5.37177  4.53039  3.09777  3.35449  3.89234  6.51684  1994  3.96389  7.25372  6.18573  5.11913  3.17302  3.52359  3.99910  6.66734  1995  4.40988  7.11026  7.36498  5.72144  3.21973  3.65787  4.09757  6.88429  1996  4.71556  8.77046  6.80562  5.32837  3.18323  3.74681  4.16933  7.03680  1997  4.46079  9.00830  6.77946  5.28875  3.12343  3.83816  4.19780  7.22032  1998  4.38396  7.35245  7.02909  5.13339  3.13173  4.05650  4.25986  7.34563  1999  4.31935  9.26009  7.12821  5.03383  3.08837  4.03743  4.31663  7.46192  2000  4.37477  15.28916  7.16500  5.39152  3.06944  4.02821  4.36260  7.72519  2001  4.62634  15.71996  7.30172  5.31976  3.09180  4.17233  4.43899  7.74816  2002  4.63446  13.37888  6.82281  5.26906  3.10623  4.21091  4.46478  7.87210  2003  4.53195  16.64222  6.33572  5.24802  2.97965  3.84484  4.47389  7.93621  2004  4.46913  18.36274  6.82304  5.79875  2.92001  3.66998  4.48713  8.09962  2005  4.14794  23.46718  6.44732  6.11917  2.87531  3.46196  4.51188  8.29947  2006  4.09923  22.90367  6.08392  6.87211  2.82144  3.31328  4.54101  8.42538  2007  4.38863  23.04150  5.71715  7.39951  2.78882  3.16568  4.54902  8.58781  142  Chapter 2. Business Sector Data on Outputs and Inputs for Canada 1961-2007 Table 2.7: Quantity Indexes for Eight Commodity Classes of Exports, 1961-2007  Year t  Qt9  Qt10  Qt11  Qt12  Qt13  Qt14  Qt15  Qt16  1961  1432.1  252.3  1573.6  2057.2  402.1  50.1  63.7  1036.0  1962  1328.6  363.9  1593.1  2130.2  511.5  62.5  77.7  1132.0  1963  1571.9  364.1  1684.6  2242.7  581.5  93.7  90.9  1204.0  1964  1963.5  408.9  1834.5  2531.7  792.8  187.7  102.2  1285.6  1965  1798.9  425.8  1869.5  2721.2  791.3  347.5  112.1  1359.7  1966  1954.0  484.2  1964.9  2809.4  910.9  985.6  135.7  1492.0  1967  1613.2  578.7  1927.6  2997.6  1147.4  1620.4  146.5  1864.9  1968  1589.7  657.2  2127.4  3282.7  1239.5  2533.5  183.3  1499.0  1969  1492.5  757.0  2290.6  3091.0  1378.2  3181.6  214.8  1657.1  1970  1967.9  951.2  2339.2  3675.8  1437.1  3137.5  236.2  1767.3  1971  2168.3  1154.0  2374.2  3611.1  1436.8  3613.9  243.1  1747.3  1972  2268.9  1479.9  2662.0  3707.9  1662.5  4001.8  270.8  1720.7  1973  2217.2  1789.7  2825.7  4085.7  1950.5  4539.2  319.8  1893.1  1974  1778.2  1508.4  2809.5  3986.3  2088.5  4430.7  323.6  2094.5  1975  1879.8  1216.3  2185.5  3513.3  2165.0  4554.8  289.8  1997.6  1976  2047.5  977.6  2718.2  3824.4  2325.5  5499.2  311.3  2048.5  1977  2445.0  919.8  3001.2  3955.3  2345.1  6388.1  338.2  2052.1  1978  2524.7  937.6  3313.3  4281.0  2914.9  6954.2  404.7  2272.8  1979  2538.2  1101.5  3369.6  4292.6  3880.9  6004.4  506.8  2556.4  1980  2785.9  951.8  3286.8  4633.9  4469.2  5002.0  570.8  2658.0  1981  2923.8  950.1  3126.5  4529.3  4810.1  5585.9  548.3  2738.1  1982  3142.7  1009.7  2962.3  4133.5  4586.9  6387.7  525.1  2508.2  1983  3256.3  1079.1  3284.3  4169.5  4413.1  7748.3  546.8  2534.8  1984  3299.9  1210.3  3501.3  4965.2  5761.5  10115.0  652.0  2685.2  1985  2979.2  1461.0  3544.7  5178.9  6360.5  10539.3  666.0  2839.4  1986  3178.0  1416.6  3708.8  5606.6  6826.0  10494.1  767.4  3254.1  1987  3565.5  1701.0  4031.4  5790.7  6884.4  10540.5  775.7  3294.5  Continued on Next Page. . .  143  Chapter 2. Business Sector Data on Outputs and Inputs for Canada 1961-2007 Table 2.7 – Continued Year t  Qt9  Qt10  Qt11  Qt12  Qt13  1988  3527.3  2009.2  1989  3101.7  1990 1991  Qt14  Qt15  Qt16  4025.0  6316.1  7120.6  11928.4  798.8  3546.0  1997.4  3836.1  6412.9  7810.3  11838.6  709.3  3704.0  3830.8  1779.1  3877.8  6765.1  9259.5  12042.3  895.2  3857.1  4169.0  2068.3  3958.3  7016.2  9536.5  10949.5  908.0  3892.6  1992  4397.4  2184.7  4131.5  7237.9  10413.1  12087.4  1167.0  4156.5  1993  4342.7  2374.6  4352.4  7773.7  11895.0  14490.7  1440.8  4519.2  1994  4746.4  2646.9  4708.9  8301.8  14402.7  16349.5  1775.9  5093.3  1995  4754.3  2868.3  4989.3  8896.4  17402.0  17200.2  2029.5  5395.8  1996  4913.1  2970.5  5073.4  9821.5  19457.0  16912.7  2278.7  5850.5  1997  5553.7  3016.9  5178.1  10708.6  22070.0  18099.8  2555.4  6263.6  1998  5711.7  3238.6  5042.1  11525.7  25770.1  19342.2  2949.8  7084.9  1999  5929.8  3226.3  5623.2  11889.3  28713.2  24097.5  3239.8  7400.4  2000  6309.4  3476.8  5969.8  12608.7  35853.8  24300.2  3484.0  7936.8  2001  6717.8  3547.7  5517.3  12744.0  33169.5  22176.3  3673.6  7967.1  2002  6661.8  3687.1  5458.9  13317.5  31256.9  22958.2  3959.7  8276.2  2003  6450.4  3636.6  5448.2  12730.0  29760.9  22727.9  3841.6  7982.5  2004  6863.6  3708.9  5777.2  13443.3  31200.7  24629.3  3848.1  8268.8  2005  7256.1  3705.8  5653.2  13721.1  32346.3  25417.7  3800.7  8284.5  2006  7613.6  3789.4  5478.7  13664.5  33058.2  24838.7  3922.2  8185.4  2007  7831.6  3977.5  5118.7  14111.7  33500.8  24419.5  4118.9  8042.4  144  Chapter 2. Business Sector Data on Outputs and Inputs for Canada 1961-2007 We now turn our attention to imports. Current dollar information on imports of services can be found as CANSIM II series V498108 for the years 1961-2007 and the corresponding constant 2002 chained dollar series is CANSIM II series V1992065. We use these two series to form price and t and Qt , which are listed in quantity series for the imports of services, P23 23  Tables 2.8 and 2.9. Note that since imported goods and services are inputs into the business sector, when we form a value added aggregate, a minus sign is appended to any quantity series pertaining to imports. As was the case for our treatment of exports, the starting point for obtaining disaggregated data on imports of goods is CANSIM Table 3800012, Exports and Imports of Goods and Services, Canada, Current Prices. Using this table, it is possible to obtain disaggregated information on the value of imports for the same seven classes of imported good that was used for exports for the years 1971-2007. However, imports of forest products was small throughout the sample period and so this import component was aggregated with imports of industrial goods and materials (excluding forest and energy imports).92 Thus we used CANSIM Table 38000012 in order to generate prices and quantities for the following six classes of imports for the years 1971-2007:93 • Q17 , Imports of agricultural and fish products; • Q18 , Imports of energy products; • Q19 , Imports of industrial goods and materials (including imports of forest products but excluding imports of energy products); • Q20 , Imports of machinery and equipment (excluding automotive products); • Q21 , Imports of automotive products and • Q22 , Imports of other consumer goods. 92 93  Chained Fisher indexes were used in order to do the aggregation. As in the case of export indexes, we used chained indexes whenever they were available.  145  Chapter 2. Business Sector Data on Outputs and Inputs for Canada 1961-2007 There remains the problem of obtaining price series for the above six classes of imports to cover the years 1961-1971. From Leacy (1983), Series G429-442: Foreign trade, imports, excluding coin and bullion, by main commodity sections, current values, 1946-1975, millions of dollars (all countries), we can obtain value series covering imports for the years 1946-1975 for the following five commodity classes: • Live animals (G429); • Food, feed, beverages and tobacco (G431); • Crude materials (inedible) (G433); • Fabricated materials (inedible) (G435); • End products (inedible) (G437). From the same source, price indexes for each of the above five classes of imports are available as Series K62-K67 in the Table with the title: Import price indexes, trade of Canada commodity classification, 1926-1975. Thus we can find price and quantity series for these five classes of exports that cover the years 1961-1971. Unfortunately, these price indexes are of the fixed base variety with a base year of 1948 so they are likely to differ substantially from the corresponding chain indexes. However, as was the case for export price indexes, Leacy (1983) also lists as part of import price Series K62-K67 (Panel A) for the above five classes of imports counterpart indexes that have a 1971 base year but these price indexes cover only the years 1968-1975. We used these latter price indexes to construct import price indexes for the years 1968-1971 and then we use the 1948 based indexes to further extend these 5 series back to 1961. The above operations gave us five disaggregated export price and quantity series for the period 1961-1971 but we have six classes of imports of goods for the years 1971-2007. We generated Fisher chained price and quantity indexes for imports of Live animals and for exports of Food, feed, beverages 146  Chapter 2. Business Sector Data on Outputs and Inputs for Canada 1961-2007 and tobacco for the years 1961-1971 and linked these series to our earlier t and Qt , imports of agricultural and fish products. As was the series, P17 17  case for extending our export series back to the 1960s, we need some additional series so that we can match the import series for the 1960s to the series that cover the post 1971 period. We will create separate import series for energy, automotive products and other consumer goods using the input output tables for the Canadian economy that cover the years 19611981 (see Statistics Canada (1987a)(1987b)). The rest of our import series computations parallel our export series computations, except that we did not generate a separate series for forest product imports due to their small size throughout the sample period. The price of imports does not include import duties that are added to the international cost of these imported goods. Hence we must add these import duties to the price of imports. We assumed that energy, automotive and service imports were exempt from import duties and we assumed a uniform rate for the remaining import categories.94 The series on customs import duties is CANSIM II series V499741 and after adjusting the price of imports using this series, the resulting price and quantity series for the imports of goods and services are listed in Tables 2.8 and 2.9.  94  This is only a very rough approximation to the truth.  147  Chapter 2. Business Sector Data on Outputs and Inputs for Canada 1961-2007 Table 2.8: Price Indexes for Seven Commodity Classes of Imports, 1961-2007  Year t  t P17  t P18  t P19  t P20  t P21  t P22  t P23  1961  1.00000  1.00000  1.00000  1.00000  1.00000  1.00000  1.00000  1962  1.04883  1.02856  1.05578  1.09816  1.05195  1.01253  1.05228  1963  1.30508  1.00981  1.08637  1.09743  1.06958  1.01429  1.07921  1964  1.28197  1.00471  1.10651  1.09682  1.09414  1.02308  1.09777  1965  1.06938  1.02524  1.12598  1.12082  1.07648  1.02077  1.13619  1966  1.05080  1.05430  1.13068  1.15070  1.09270  1.03586  1.17024  1967  1.02344  0.98958  1.16189  1.18684  1.12302  1.05402  1.22610  1968  1.05513  1.04752  1.14800  1.19862  1.16016  1.06806  1.28761  1969  1.06398  1.00624  1.17757  1.21963  1.18860  1.08873  1.36902  1970  1.13052  1.01555  1.19414  1.22480  1.20108  1.09806  1.42900  1971  1.16283  1.11732  1.16445  1.25747  1.24031  1.10649  1.49900  1972  1.24651  1.20329  1.15700  1.26668  1.26989  1.14591  1.53952  1973  1.52015  1.43514  1.27659  1.29789  1.29640  1.18635  1.63282  1974  1.87300  4.22172  1.65984  1.42128  1.40659  1.29665  1.75102  1975  1.97450  5.43394  1.83511  1.67222  1.63529  1.44968  1.98130  1976  1.85375  5.63358  1.85932  1.67883  1.71458  1.52194  2.04366  1977  2.24663  6.42113  2.09118  1.81185  1.93225  1.71419  2.34119  1978  2.49806  7.14772  2.42204  1.83462  2.20965  1.95044  2.69834  1979  2.79922  9.22856  2.93293  1.94108  2.44387  2.14927  3.04229  1980  3.05858  14.04487  3.40659  1.71574  2.71841  2.44883  3.39335  1981  3.35205  16.89476  3.70305  1.67637  3.26363  2.79785  3.81611  1982  3.27980  16.65354  3.79588  1.82550  3.52543  2.94081  4.10352  1983  3.19245  15.35401  3.78529  1.74893  3.62370  2.95617  4.30819  1984  3.39093  15.57686  3.91220  1.78888  3.82410  3.17305  4.63802  1985  3.29961  16.05971  3.85316  1.81429  4.02632  3.27951  4.98855  1986  3.52718  10.45556  3.94881  1.83440  4.20229  3.51229  5.29276  1987  3.44257  10.91648  3.94832  1.74520  4.15738  3.53064  5.28864  Continued on Next Page. . .  148  Chapter 2. Business Sector Data on Outputs and Inputs for Canada 1961-2007 Table 2.8 – Continued Year t  t P17  t P18  t P19  t P20  t P21  t P22  t P23  1988  3.50154  9.09050  4.05930  1.67407  4.03524  3.50861  5.09254  1989  3.45272  9.69611  4.06260  1.63403  4.08857  3.52386  5.09401  1990  3.45595  11.94481  4.01098  1.61746  4.16440  3.57930  5.24914  1991  3.43419  10.19284  3.86391  1.58265  4.12731  3.59254  5.31731  1992  3.43112  10.16216  3.93946  1.63533  4.39452  3.79083  5.63381  1993  3.44632  9.84044  4.05394  1.70897  4.68083  4.02323  6.19826  1994  3.70890  9.83576  4.32449  1.78245  4.99133  4.27046  6.68231  1995  3.92677  10.17736  4.70222  1.74664  5.15679  4.39186  6.90224  1996  3.88275  11.94335  4.52874  1.66405  5.19805  4.36419  7.01253  1997  3.99625  11.84732  4.52516  1.62932  5.26105  4.38392  7.26074  1998  3.97357  9.76539  4.64624  1.67064  5.51748  4.66991  7.80221  1999  3.85703  11.43514  4.55850  1.62996  5.52100  4.67035  7.97505  2000  3.84946  16.95310  4.72363  1.59815  5.52428  4.69368  8.18543  2001  3.99707  16.32263  4.90529  1.63313  5.67016  4.91888  8.64576  2002  4.05634  16.80830  4.85791  1.63385  5.73431  4.93873  8.84642  2003  3.91718  18.58307  4.58281  1.46138  5.39577  4.49645  8.39644  2004  3.81803  21.35868  4.74735  1.34910  5.21353  4.18264  8.24087  2005  3.72978  27.03564  4.82559  1.26664  5.00314  4.01666  8.14870  2006  3.62982  30.53146  5.01981  1.20443  4.84982  3.84873  8.19437  2007  3.73318  31.10128  4.92307  1.14851  4.63116  3.67501  8.15599  149  Chapter 2. Business Sector Data on Outputs and Inputs for Canada 1961-2007 Table 2.9: Quantity Indexes for Seven Commodity Classes of Imports, 1961-2007  Qt17  Qt18  Qt19  Qt20  1961  824.1  478.0  1850.2  1840.7  615.1  754.0  1535.0  1962  840.3  480.7  1926.5  1797.9  699.1  811.6  1479.6  1963  786.0  541.8  1952.1  1788.1  707.1  799.6  1466.8  1964  810.8  547.7  2235.8  2092.0  857.4  842.2  1618.7  1965  934.5  583.4  2482.1  2400.7  1157.2  948.7  1692.5  1966  1026.7  592.8  2666.5  2826.7  1573.4  1066.8  1850.0  1967  1086.6  638.9  2528.6  3077.0  1963.9  1109.3  1934.6  1968  1127.9  685.5  2726.8  3135.8  2649.7  1240.3  2013.8  1969  1290.9  737.9  3052.2  3522.5  3030.9  1461.0  2336.7  1970  1279.6  766.3  3041.7  3548.7  2758.7  1459.4  2471.7  1971  1293.5  817.1  3359.3  3712.9  3249.2  1650.7  2476.3  1972  1424.2  895.9  3855.8  4450.5  3819.2  2010.7  2585.9  1973  1625.4  928.1  4182.5  5370.1  4619.0  2325.5  2888.3  1974  1674.4  788.6  4742.9  6357.6  4931.1  2642.7  3277.5  1975  1675.3  766.1  4071.1  6031.1  4954.5  2583.3  3550.2  1976  1940.0  719.6  4177.8  6189.4  5417.1  2998.7  3971.3  1977  1846.2  654.2  4160.4  6312.2  5864.7  2943.3  3883.1  1978  1872.6  625.7  4499.4  7651.1  5918.6  2988.3  3824.2  1979  1855.1  624.7  5082.4  9325.4  6096.9  3148.4  3637.1  1980  1921.5  598.8  4917.6  12269.7  4900.3  3016.1  3760.0  1981  1994.6  573.7  5163.8  15342.7  4799.6  3113.5  3860.0  1982  1914.8  404.5  4146.0  11976.5  4128.6  2890.3  3594.5  1983  1985.9  336.2  4779.6  13552.5  5142.3  3228.1  3689.7  1984  2169.3  393.7  5469.4  16671.9  6668.0  3613.1  3774.9  1985  2190.2  395.2  6052.7  17490.1  7684.4  3557.1  3904.9  1986  2287.9  486.3  6369.6  19038.1  7845.2  3806.2  4267.7  1987  2388.1  541.7  6424.7  21231.5  7844.3  3991.8  4536.3  Year t  Qt21  Qt22  Qt23  Continued on Next Page. . .  150  Chapter 2. Business Sector Data on Outputs and Inputs for Canada 1961-2007 Table 2.9 – Continued Year t  Qt17  Qt18  Qt19  1988  2381.6  569.4  7298.0  1989  2625.7  641.6  1990  2779.6  1991  2871.4  1992  Qt20  Qt21  Qt22  Qt23  26838.5  8225.8  4266.7  5184.6  7645.9  29100.6  7812.7  4681.5  5792.5  686.3  7577.4  29167.1  7319.2  4868.6  6405.6  650.4  7342.9  29676.0  7501.5  5065.0  6664.1  3100.3  637.4  7950.1  31183.3  7664.1  5459.7  6738.8  1993  3437.0  708.1  8947.7  33412.5  8533.4  5711.6  6865.2  1994  3617.1  707.6  10111.7  39325.5  9583.4  5854.9  6755.0  1995  3597.8  711.1  10694.6  45780.5  9712.7  6144.3  6763.0  1996  3839.3  804.2  11268.5  48400.6  9832.0  6242.8  7110.7  1997  4101.5  897.0  13177.8  58701.8  11561.6  7109.8  7374.5  1998  4531.5  884.1  14104.5  63169.9  12105.0  7726.6  7366.6  1999  4760.9  936.4  14810.3  69069.7  13753.7  8239.7  7683.6  2000  4998.2  1053.1  15872.4  79742.9  14016.8  8861.7  8114.0  2001  5301.6  1087.2  15123.8  71310.1  12799.3  9072.2  7954.1  2002  5594.2  985.7  15449.5  67567.7  14207.5  9805.4  8090.6  2003  5735.7  1066.2  15563.8  70541.0  14176.3  10757.9  8830.5  2004  5842.3  1160.3  16838.7  80434.0  14839.8  11893.8  9363.0  2005  6159.0  1245.4  17648.1  91268.2  15667.0  12840.8  9823.2  2006  6731.2  1134.1  18076.1  99174.3  16464.5  14081.0  10104.3  2007  7118.8  1175.8  18659.2  105849.0  17274.7  15541.1  10697.4  151  Chapter 2. Business Sector Data on Outputs and Inputs for Canada 1961-2007 We now turn our attention to forming estimates of business sector labour input.  2.3  Business Sector Labour Input Estimates  Quality adjusted measures of the quantity of three types of labour for the years 1961-2007 are available from the Statistics Canada KLEMS productivity program; see CANSIM Table 3830021 which has the title: Multifactor Productivity, Value Added, Capital Input and Labour Input in the Aggregate Business Sector and Major Sub-Sectors, by the North American Industry Classification System (NAICS). The three series are V41713000 (the title is Canada: Labour Input of Workers with Primary or Secondary Education; Business Sector), V41713017 (Labour Input of workers with Some or Completed Post-Secondary Certificate or Diploma; Business Sector) and V41713034 (Labour Input of Workers with University Degree or Above, Business Sector). The corresponding value of labour input or labour compensation series are found in the same table and their CANSIM series numbers are V41713187, V41713204 and V41713221 respectively. These value series however only cover the years 1961-2004.95 These KLEMS labour series allowed us to construct the three business sector labour input series QtL1 , QtL2 and QtL3 for the years 1961-2007 (see Table 2.10 for a listing of  t , P t and P t for these data) and the corresponding wage index series PL1 L2 L3  the years 1961-2003 (see Table 2.10).  95 This is very puzzling: the quantity series run from 1961 to 2007 but the corresponding value series stops at 2004.  152  Chapter 2. Business Sector Data on Outputs and Inputs for Canada 1961-2007 Table 2.10: Price and Quantity Indexes for Three Types of Business Sector Labour  Year t  t PL1  t PL2  t PL3  QtL1  QtL2  QtL3  1961  1.00000  1.00000  1.00000  17122  710  1370  1962  1.02980  1.18632  1.01126  17345  1216  1448  1963  1.05959  1.21347  1.04509  17259  1723  1487  1964  1.10555  1.24560  1.09234  17482  2230  1566  1965  1.18267  1.29309  1.15814  17774  2762  1657  1966  1.26544  1.33357  1.24479  18152  3345  1788  1967  1.34988  1.36529  1.31145  18066  3852  1853  1968  1.44254  1.40528  1.40547  17740  4282  1853  1969  1.55946  1.46444  1.53399  17740  4789  1918  1970  1.66484  1.50846  1.62547  17379  5195  1983  1971  1.77957  1.67298  1.59351  17190  5777  2166  1972  1.93443  1.81847  1.60097  17190  6411  2336  1973  2.13905  1.97009  1.63324  17705  7247  2505  1974  2.49530  2.23902  1.78277  17826  7931  2740  1975  2.90404  2.52097  1.99402  17328  8362  2870  1976  3.39630  2.82284  2.16028  16916  8793  2897  1977  3.75369  3.03206  2.22664  16675  9300  3066  1978  3.90502  3.15557  2.39954  16967  10110  3314  1979  4.17384  3.37618  2.61573  17482  11149  3640  1980  4.51590  3.71514  2.83118  17654  11960  3901  1981  5.00927  4.09363  3.43078  17843  12492  4188  1982  5.50893  4.47191  3.61769  16761  11985  4149  1983  5.58062  4.79842  3.92881  16727  12188  4254  1984  5.95405  4.90205  4.14250  17053  12771  4697  1985  6.20453  5.19867  4.38987  17499  13455  5062  1986  6.33749  5.30440  4.65051  17946  14241  5480  1987  6.64512  5.43184  4.69809  18530  15153  5911  Continued on Next Page. . .  153  Chapter 2. Business Sector Data on Outputs and Inputs for Canada 1961-2007 Table 2.10 – Continued Year t  t PL1  t PL2  t PL3  QtL1  QtL2  QtL3  1988  7.12085  5.79003  4.88684  19045  16040  6432  1989  7.26900  5.96671  5.70897  19268  16648  6798  1990  7.26005  6.49660  5.97755  18908  16927  7007  1991  7.42364  6.73848  6.66739  17843  16547  7137  1992  7.57474  6.90164  6.68101  17242  16471  7450  1993  7.69212  6.83616  6.46124  16864  17079  8246  1994  7.69473  6.84046  6.22327  16795  18371  8768  1995  7.81823  7.01644  6.20187  16692  19537  9055  1996  7.93652  6.96834  6.54102  16812  20322  9538  1997  8.12992  7.18146  7.03193  16332  21918  10073  1998  8.34882  7.36378  7.31313  16383  22679  10856  1999  8.51971  7.52497  7.58425  16984  23414  11325  2000  8.92053  7.90809  7.95222  17311  24199  12082  2001  9.06503  8.10605  8.30514  16967  24706  12695  2002  9.14718  8.18965  8.50473  17173  25339  13048  2003  9.33522  8.37832  8.56340  16778  26226  13517  2004  9.56758  8.56262  8.80245  17156  27113  14313  2005  9.92027  8.90341  9.10008  17122  27062  15292  2006  10.37190  9.30873  9.51436  17396  27189  15983  2007  10.74529  9.64385  9.85688  17242  28101  16818  154  Chapter 2. Business Sector Data on Outputs and Inputs for Canada 1961-2007 The Statistics Canada productivity program aggregate labour input measure is described as follows: “The labour input is an aggregate of the hours worked of all persons classified by their education, work experience and class of employment (paid versus self-employed workers). This aggregate labour input measure is constructed by aggregating hours at work data for each of 56 types of workers classified by their educational attainment (4), work experience (7) and class of workers (2) using an annual chained-Fisher index. The effect of Fisher aggregation is to produce a measure of labour input that reflects both changes in total hours of work and changes in the composition of workers.” John R. Baldwin, Wulong Gu and Beiling Yan 2007; 37. Baldwin, Gu and Yan (2007; 26) describe their more disaggregated measures of labour input as follows: “Labour input for MFP measures reflects the compositional shifts of workers by education, experience and class of workers (paid versus self-employed). The growth of labour input (labour services) is an aggregate of the growth of hours worked by different classes of workers, weighted by the hourly wages of each class.” Thus each of the three types of labour classified by educational attainment QtL1 ,  QtL2 and QtL3 is a Fisher quantity aggregate over the other characteris-  tics, holding constant the relevant educational levels. Baldwin, Gu and Yan (2007; 26) also comment on the difficulties associated with breaking up the net operating surplus generated by the self-employed into labour and capital compensation components: “We have modified the assumptions about the share of labour going to the self-employed workers to reflect changes that occurred during the 1990s. In the past, it had been assumed that 155  Chapter 2. Business Sector Data on Outputs and Inputs for Canada 1961-2007 the self employed essentially earned incomes similar to the employed. The Census of Population up to 1990 showed that this was a reasonable assumption; however, during the 1990s, selfemployed income fell behind that of production workers. The new measure of self-employed for calculating labour input assumes that the hourly earning of self-employed workers is proportional to that of paid workers with the same level of education and experience. The proportional or scaling factor for each level of education and experience is based on the relative hourly earnings of paid versus self-employed workers derived from the Census of Population.” Overall, we believe that Statistics Canada has done an excellent job in constructing their new measures of labour input and we will use these measures in the present study.96 The effect of using the Statistics Canada measures of quality adjusted labour input is to increase the growth of labour input by about 37% over the sample period compared to using hours worked as the measure of labour input.97 Basically, there was a big shift in labour inputs from less skilled and less educated workers to more educated workers over this period which served to greatly increase quality adjusted labour input compared to unweighted hours worked by all types of labour. As noted above, the KLEMS estimates of real labour input for the three types of labour run from 1961-2007 but the corresponding value series stop 96  The labour input that is used in the residential rental of housing industry should be deducted from our measure of labour input (since we exclude all residential housing outputs from our definition of the business sector while the KLEMS program business sector excludes only the services of Owner Occupied Housing). However, the KLEMS database that is available in CANSIM does not include information on the three types of labour input that is used in the residential housing rental industry so we were not able to deduct these labour inputs from total business sector labour input. Thus our productivity estimates will have a tiny downward bias due to this factor. 97 Estimates of total hours worked in the KLEMS business sector for the years 1961-2007 are available from CANSIM II series V41712966, (Canada, Hours Worked, Business Sector) in Table 3830021 (Multifactor Productivity, Value Added, Capital Input and Labour Input in the Aggregate Business Sector and Major Sub-Sectors, by North American Industry Classification System (NAICS)).  156  Chapter 2. Business Sector Data on Outputs and Inputs for Canada 1961-2007 at 2004. Hence we need to estimate either wages or values for the three types of labour for the years 2004-2007. In order to accomplish this task, we formed our own estimates of the total value of labour input over the years 1961-2007. Estimates of wages, salaries and supplementary labour income for the business sector are available from CANSIM II series V498167 for the years 1961-2007. However, this measure of business sector payments for labour services neglects the labour input of the self-employed (and unpaid family workers); i.e., it includes only the gross wages of employees. The value of the labour services rendered by the self employed are part of the gross operating surplus of the household sector, which includes also the returns to the capital and land used by the self employed. An upper bound to the value of self employed labour services is the sum of unincorporated business net income which is available for 1961-2007 as CANSIM II series V498170. We assumed that two thirds of unincorporated net income is a return to labour and one third is the return to capital. We added this imputed labour income of the self employed to the labour income of employees in the business sector and compared this measure of total business sector labour compensation to the corresponding total labour compensation from the KLEMS database98 and found that these two series were very close until about 1995 and then they gradually diverged to end up about 4% apart in 2003. We used the rates of growth of our imperfect measure of business sector labour income growth to extend the official KLEMS business sector labour compensation series from 2004 to 2007. We then divided this extended measure of total labour compensation by the KLEMS business sector measure of aggregate labour input99 in order to obtain an implicit wage rate for aggregate business sector labour for the years 2004-2007. We used the movements in this implicit t , P t and P t from 2004 wage rate to extend the KLEMS wage indexes PL1 L2 L3  to 2007; see Table 2.10 for the results of these manipulations. 98  See the CANSIM II series V41713170, Canada, Labour Compensation, Business Sector, in Table 3830021, Multifactor Productivity, Value Added, Capital Input and Labour Input in the Aggregate Business Sector and Major Sub-Sectors, by NAICS. 99 See the CANSIM II series V41712949 with the title Canada, Labour Input, Business Sector.  157  Chapter 2. Business Sector Data on Outputs and Inputs for Canada 1961-2007 We now turn our attention to the problems associated with the estimation of beginning of the year capital stocks for the business sector.  2.4  Business Sector Capital Stock Estimates  Our general strategy in this section will be to use estimates from the National Balance Sheets to obtain estimates of inventory and land stocks used by the business sector (see Statistics Canada (1997)). This balance sheet information is also used to calibrate estimates of depreciation for reproducible capital stocks used by the business sector. For the years 1962-2008, beginning of the year estimates of various national wealth components can be obtained from the CANSIM II database. National totals for the value of various assets can be obtained from CANSIM Table 3780004 (National Balance Sheet Accounts, by Sectors) for residential structures (see series V34675), non-residential structures (V34676), machinery and equipment (V34677), inventories (V34679) and land (V34680). The same table has the corresponding asset values for the persons and unincorporated business sector; for residential structures (see series V33464), non-residential structures (V33465), machinery and equipment (V33466), inventories (V33468) and land (V33469). Table 3780004 also has the corresponding asset values for corporations and government business enterprises; for residential structures (see series V31693), non-residential structures (V31694), machinery and equipment (V31695), inventories (V31696) and land (V31697). Finally, Table 3780004 has the corresponding asset values for the government sector; for residential structures (see series V32575), non-residential structures (V32576), machinery and equipment (V32577), inventories (V32578) and land (V32579). We subtracted the government sector value of non-residential structures, machinery and equipment and inventories from the corresponding total economy asset values in order to obtain business sector estimates of the value of beginning of the year t busit , business machinery and ness sector non-residential structure stocks V KN R t t equipment stocks, V KM E , and business inventory stocks V KBI ; see Table  158  Chapter 2. Business Sector Data on Outputs and Inputs for Canada 1961-2007 2.11 for a listing of these business sector stock values. Although residential structures are not part of our domain of definition for business sector output, it will prove useful to have some information on the value of residential structures and residential land for comparison purposes. Thus the total value of residential structures from the national balance sheets for Canada, t , is also listed in Table 2.11. V KRS  We ended up not using the balance sheet information on the value of business sector machinery and equipment. Instead, we used the investment t and QtIICT , and on non-ICT series on ICT machinery and equipment, PIICT t t machinery and equipment, PIM E and QIM E , listed in Tables 2.2 and 2.3  that were provided to us by Statistics Canada from their KLEMS database. Statistics Canada also provided us with the companion business sector price and quantity series for the beginning of the year capital stocks for ICT and non-ICT machinery and equipment. This allowed us to compare the two price series for ICT. The KLEMS ICT investment price decreased from 1 in 1961 to 0.214 in 2006 whereas the KLEMS ICT capital stock price decreased from 1 in 1961 to 0.462 in 2006, which is a considerable difference.100 If there were no asset heterogeneity in the class of ICT investments (which there certainly is), then using the geometric model of depreciation (which is used by the KLEMS program), we would expect these two price series to be very close to each other. The KLEMS non-ICT machinery and equipment investment price increased from 1 in 1961 to 4.71 in 2006 whereas the KLEMS non-ICT capital stock price increased from 1 in 1961 to 5.23 in 2006, which again indicates some asset heterogeneity, but the divergence between these two series is not nearly as large as the divergence in the two ICT investment and capital stock price series. The question now arises: which of these two price series should we use for a geometric model of depreciation? 100 There are even bigger differences in the rates of growth of the KLEMS ICT capital stocks versus the corresponding KLEMS measures of ICT services growth: the ICT stock grew 110 fold over the period 1961-2006 while the flow of ICT services grew 591 fold over the same period. This seems implausible.  159  Chapter 2. Business Sector Data on Outputs and Inputs for Canada 1961-2007 Using the geometric or declining balance depreciation model of depreciation, the starting capital stock of a generic asset in period t + 1, QK t+1 , is equal to one minus the depreciation rate in period t, δ t , times the previous period’s starting stock, QK t , plus the new investment in the previous period, QtI ; i.e., we have: QK t+1 = (1 − δ t )QK t + QtI  (2.6)  Given information on beginning of the year capital stocks and investment during each year, the above equation can be solved for a balancing depreciation rate, δ t , that reconciles the investment information with the balance sheet information: δ t = [QK t − QK t+1 + QtI ]/QK t  (2.7)  Using the Statistics Canada KLEMS database for ICT investment and capital, we tried deflating the value data by either the ICT investment price deflator or the ICT capital stock price deflator. Using the latter deflator, the implied ICT depreciation rate trended up from 0.238 in 1961 to 0.837 in 2005. These depreciation rates seem to be too large. However, when we implemented Equation (2.7) using the ICT investment price deflator, the implied ICT depreciation rate trended up from 0.239 in 1961 to 0.329 in 2005. These depreciation rates seem to be very reasonable so we decided to use the ICT investment price deflator as the price of both ICT investments and capital stocks. Since the implied depreciation rates for ICT capital using Equation (2.7) had a pronounced upward trend, we regressed these rates on a constant and a time trend. The estimated constant was 0.19909 with a standard error of 0.01644 and the estimated trend parameter was 0.00263 with a standard error of 0.0006437 so that both estimated parameters were significant. Thus we decided to assume a starting depreciation rate of 0.200 in 1961 and we increased this depreciation rate by 0.00263 each subsequent year. We then used Equation (2.6) above along with the Statistics Canada 1961 value of the ICT capital stock as our starting value 160  Chapter 2. Business Sector Data on Outputs and Inputs for Canada 1961-2007 for the 1961 quantity of ICT capital in order to generate a new series for t (see the business sector beginning of the year ICT capital stock, QKICT  Table 2.12). An entirely analogous procedure was used to generate a new series for the business sector beginning of the year non-ICT machinery and t equipment capital stock, QKM E . Again, we used the KLEMS investment  t price deflator for non-ICT machinery and equipment, PIM E listed above  in Table 2.2 as the deflator for the value of the non-ICT machinery and equipment capital stocks and the associated values of investments from the KLEMS database and we calculated the implied depreciation rates using Equation (2.7). The resulting implied depreciation rates had a small upward trend, starting at 0.160 in 1961 and ending up at 0.175 in 2005. Since this implied depreciation rates for ICT capital had an upward trend, we regressed these rates on a constant and a time trend. The estimated constant was 0.15034 with a standard error of 0.005459 and the estimated trend parameter was 0.00067351 with a standard error of 0.0002137 so that both estimated parameters were significant. Thus we decided to assume a starting depreciation rate of 0.150 in 1961 and we increased this depreciation rate by 0.00067 each subsequent year. We then used Equation (2.6) along with the Statistics Canada KLEMS program 1961 value of the non-ICT machinery and equipment capital stock as our starting value for the 1961 quantity of non-ICT machinery and equipment capital in order to generate a new series t for the business sector beginning of the year ICT capital stock, QKM E (see  Table 2.12).101  101  We used the rate of price inflation in Machinery and Equipment investment (using t t national accounts data) over the years 2006-2007 in order to extend PIICT and PIM E to 2007.  161  Chapter 2. Business Sector Data on Outputs and Inputs for Canada 1961-2007 Table 2.11: Beginning of Year Asset Values for Residential Structures and Land and Six Business Sector Capital Stocks  t V KICT  t V KM E  t V KN R  t V KBI  t V KAL  t V KBL  t V KRL  1961  1236  11994  27850  13594  5954  6376  10680  28710  1962  1297  12291  29388  13698  6203  6820  11423  29923  1963  1384  13161  31414  14292  6573  7281  12083  31707  1964  1506  13526  33599  15398  7303  7840  12548  34310  1965  1633  14853  37288  16224  8156  8537  13682  37875  1966  1778  16465  41694  17884  9132  9621  14974  42144  1967  2079  18394  46012  19588  10281  10971  16365  46525  1968  2361  20118  48638  20303  11298  12138  17719  49296  1969  2575  21618  53654  21462  11476  13161  19830  54058  1970  2892  23849  58457  23742  11534  14717  22130  58649  1971  3129  25718  64343  24275  11709  16421  24498  65459  1972  3440  27339  70782  25097  12542  18614  28056  74892  1973  3777  29576  81305  27660  15088  21533  32837  92703  1974  4266  36087  100032  33614  19533  26117  40030  116929  1975  5202  44936  116373  43928  24583  32759  47559  133160  1976  5841  51020  128858  46336  28876  38954  53411  150350  1977  6670  58919  141312  50117  33354  43763  59187  164910  1978  7226  67818  157802  57091  39666  49318  66442  183222  1979  7968  78428  179306  66060  49189  55774  73636  207120  1980  7976  91337  210699  81062  62176  64557  83608  234113  1981  8990  106059  245999  89024  70108  75224  101065  272325  1982  12096  121176  278898  98428  70197  86636  118825  288844  1983  12154  125826  287481  90451  68838  93532  117239  307383  1984  13465  131389  306422  91417  65974  96021  127037  328929  1985  14798  137700  323505  99318  62107  101438  139331  348994  1986  16116  144732  336131  104983  57733  105796  142788  388346  1987  17570  147727  357496  109889  54270  113337  165711  444678  Year t  t V KRS  Continued on Next Page. . .  162  Chapter 2. Business Sector Data on Outputs and Inputs for Canada 1961-2007 Table 2.11 – Continued t V KRS  t V KICT  t V KM E  t V KN R  t V KBI  t V KAL  t V KBL  t V KRL  1988  21135  152490  384524  117358  53316  122547  195773  498048  1989  23038  167116  409950  126135  56789  134213  223325  551991  1990  26275  180720  433558  132675  61388  145366  262598  574912  1991  26883  179871  435196  130781  63028  155133  258677  614114  1992  28186  189167  439422  123077  61853  159092  287087  633754  1993  32035  195633  445261  121352  62227  162090  307046  667294  1994  34428  201705  460244  124117  64707  169691  330460  698905  1995  36200  210289  468982  131198  69745  179044  352720  713616  1996  37274  216893  485422  146615  75539  185095  343616  719997  1997  40724  223783  500321  150648  81546  193888  352946  743640  1998  45118  244159  522564  158409  86376  202313  374636  766757  1999  49415  255591  541413  169901  89497  211188  394371  797843  2000  56416  269491  568533  178794  92140  221506  425256  829875  2001  65453  286828  582195  194366  95020  234213  452800  867239  2002  69434  296458  602645  190023  97613  243534  502343  926184  2003  66302  282138  621034  192080  99905  254448  573841  1003732  2004  63616  282551  669186  187291  102197  270107  633362  1099801  2005  64580  287617  723667  193723  104627  293163  733545  1190816  2006  65678  295028  795080  204832  107323  313706  834713  1323948  2007  70684  299576  862641  215812  111010  334379  949764  1468026  Year t  163  Chapter 2. Business Sector Data on Outputs and Inputs for Canada 1961-2007 Determining the value of business sector land is difficult. The problem is that the household sector owns a considerable amount of land that is used for business purposes; i.e., unincorporated persons own farm land and rental business properties and the land used in these enterprises should appear as inputs into the business sector. The corporate business sector also owns some land associated with residential rental properties and we are trying to exclude these inputs from our measure of business sector input. We will make some rough approximations in an attempt to solve these difficulties. t We first find estimates for the price and quantity of agricultural land, PAL  t . Estimates of the area of agricultural land are available for the and QKAL  Census years 1981, 1986, 1991, 1996, 2001 and 2006 from CANSIM II series V32166910 and we interpolated the quantity of land in use in agriculture between these years using constant rates of growth (geometric interpolation). From Leacy (1983), series M-23, Area of Land in Farm Holdings, Census Data in thousands of acres, we can obtain estimates of the area of farm land for 1961 and 1971. After converting from acres to hectares, these data can be appended to the previous data and again geometric interpolation between t ; the various census years can be used to complete our estimates for QKAL  see Table 2.12 for a listing.102 CANSIM Table 20020 (Balance Sheet of the Agricultural Sector at December 31) has asset value data for the end of the year for 1981-2007, which is beginning of the year values for the years 19822008. The two series that are of interest to us from this table are V157698 (the value of farm real estate) and V157699 (the value of farm land), which t for year t. Thus for the years 1982-2008, the price we denote by V KAL t can be obtained by of agricultural land, the price of agricultural land, PAL t by QK t . For the years 1961-1980, we link P t to CANSIM dividing V KAL AL AL  series V381831 (the title is Canada, Value per Acre) in Table 20003, Value per Acre of Farm Land and Buildings. This last series runs from 1961 to t for the overlap years 19812008 and we found that it was quite close to PAL  2008. With estimates for the price and quantity of agricultural land for the 102 As usual, the listed data are normalised so that the corresponding price is unity in 1961.  164  Chapter 2. Business Sector Data on Outputs and Inputs for Canada 1961-2007 t ; years 1961-1980, we can form estimates for the corresponding values, V KAL t and QK t , see Table 2.11. The price and quantity of agricultural land, PAL AL  are listed in Tables 2.13 and 2.14. We assumed that agricultural land is an input into our business sector. t , is equal to the We also assumed that the value of residential land, V KRL  total value of household and unincorporated business land less the value of agricultural land. Finally, we assumed that the value of non-agricultural business land is equal to equal to the value of corporate enterprise land, t ; see Table 2.11. V KBL t and the quanWe also assumed that the quantity of residential land QKRL  t are constant over the sample tity of business non-agricultural land QKBL t t and PBL are proporperiod and hence the corresponding price series PRL  t t and V KBL for the years tional to the corresponding value series V KRL  1962-2007. We extended these two price series back to 1961 using the movement from 1961 to 1962 in another land price series; namely series S319 in Leacy (1983): Average Land Cost per Dwelling Unit, NHA, Single Det and P t are listed in Table 2.13 and the tached. These land price series, PRL BL  t t and QKBL are listed in Table 2.14.103 corresponding quantity series, QKRL  From Table 2.2, we have price deflators for non-residential structures for t t t year t, PIN R , and we use these deflators to divide V KN R by PIN R in order to t obtain preliminary beginning of the year capital stock quantity series QKM E  t .104 Recall that a series for the annual quantity of investment in and QKN R 103  The Statistics Canada KLEMS program made available to us their aggregate price and quantity series for land used in the business sector. These series cover our agricultural land and our nonagricultural and nonresidential land series and also cover the part of residential land that applies to rental housing. The KLEMS capital stock of land grew 2.01 fold over the period 1961-2006 and the corresponding price series grew 18.7 fold. On the other hand, our estimate of the growth in the quantity of agricultural land and nonresidential and nonagricultural business land is essentially zero, with corresponding 18.6 fold and 48.5 fold increases in the price of these two business land components. We estimate even higher growth rates in the price of residential land but more research in this area is needed. 104 The use of these prices (which are average prices over the year) for stock deflation purposes is not quite appropriate because conceptually, we should be using the prices that  165  Chapter 2. Business Sector Data on Outputs and Inputs for Canada 1961-2007 non-residential structures, QtIN R , is available from Table 2.3. Now we will apply Equation (2.7) again and generate a series of geometric depreciation t rates δN R for the non-residential stock of structures for the years 1962-  2006. The mean of these depreciation rates turned out to be 0.0616 with a standard deviation of 0.03253, so there was a considerable amount of variability in these rates. There appeared to be a slight upward trend in these depreciation rates so we regressed them on a time trend. The estimated constant was 0.045439 with a standard error of 0.009565 and the estimated trend parameter was 0.0007034 with a standard error of 0.0003621. Thus we decided to assume a starting depreciation rate of 0.045 in 1961 and we increased this depreciation rate by 0.0007 each subsequent year. We then used Equation (2.6) along with the Statistics Canada balance sheet value of the stock of non-residential structures in 1962 as our starting value in order to generate a new series for the business sector beginning of the year t non-residential capital stock, QKN R (see Table 2.14).  From Table 2.2, we have price deflators for residential structures for year t , and we use these deflators to divide V K t t t, PIR RS by PIR in order to obt . tain preliminary beginning of the year capital stock quantity series, QKRS  Recall that a series for the annual quantity of investment in residential structures, QtIR , is available from Table 2.3. Now we are in a position to apply Equation (2.7) again and generate a series of geometric depreciation rates t for the residential structures capital stock for the years 1962-2006. The δRS  mean of these depreciation rates turned out to be 0.040239 with a standard deviation of 0.01795. There appeared to be no trends in these depreciation rates so we decided to assume a constant geometric depreciation rate of 0.04 for each year. We then used Equation (2.6) along with the Statistics Canada balance sheet value of the stock of residential structures in 1962 as our starting value in order to generate a new series for the business sector t (see Table 2.14). beginning of the year residential capital stock, QKRS  prevail for these stock components at the beginning of the year rather than the average prices in the year which follows. However, for our purposes, the errors made here will not be material.  166  Chapter 2. Business Sector Data on Outputs and Inputs for Canada 1961-2007 t t t t The smoothed geometric depreciation rates δICT , δM E , δN R and δRS are  listed in Table 2.12. Table 2.12: Smoothed Geometric Depreciation Rates for ICT, Non-ICT Machinery and Equipment, Non-residential Structures and Residential Structures Capital Stocks Implied by the Balance Sheets and Investment Flow Data  Year t  t δICT  t δM E  t δN R  t δRS  1961  0.20000  0.15000  0.0450  0.04  1962  0.20263  0.15067  0.0457  0.04  1963  0.20526  0.15134  0.0464  0.04  1964  0.20789  0.15201  0.0471  0.04  1965  0.21052  0.15268  0.0478  0.04  1966  0.21315  0.15335  0.0485  0.04  1967  0.21578  0.15402  0.0492  0.04  1968  0.21841  0.15469  0.0499  0.04  1969  0.22104  0.15536  0.0506  0.04  1970  0.22367  0.15603  0.0513  0.04  1971  0.22630  0.15670  0.0500  0.04  1972  0.22893  0.15737  0.0527  0.04  1973  0.23156  0.15804  0.0534  0.04  1974  0.23419  0.15871  0.0541  0.04  1975  0.23682  0.15938  0.0548  0.04  1976  0.23945  0.16005  0.0555  0.04  1977  0.24208  0.16072  0.0562  0.04  1978  0.24471  0.16139  0.0569  0.04  1979  0.24734  0.16206  0.0576  0.04  1980  0.24997  0.16273  0.0583  0.04  1981  0.25260  0.16340  0.0590  0.04  1982  0.25523  0.16407  0.0597  0.04  1983  0.25786  0.16474  0.0604  0.04  Continued on Next Page. . .  167  Chapter 2. Business Sector Data on Outputs and Inputs for Canada 1961-2007 Table 2.12 – Continued Year t  t δICT  t δM E  t δN R  t δRS  1984  0.26049  0.16541  0.0611  0.04  1985  0.26312  0.16608  0.0618  0.04  1986  0.26575  0.16675  0.0625  0.04  1987  0.26838  0.16742  0.0632  0.04  1988  0.27101  0.16809  0.0639  0.04  1989  0.27364  0.16876  0.0646  0.04  1990  0.27627  0.16943  0.0653  0.04  1991  0.27890  0.17010  0.0660  0.04  1992  0.28153  0.17077  0.0667  0.04  1993  0.28416  0.17144  0.0674  0.04  1994  0.28679  0.17211  0.0681  0.04  1995  0.28942  0.17278  0.0688  0.04  1996  0.29205  0.17345  0.0695  0.04  1997  0.29468  0.17412  0.0702  0.04  1998  0.29731  0.17479  0.0709  0.04  1999  0.29994  0.17546  0.0716  0.04  2000  0.30257  0.17613  0.0723  0.04  2001  0.30520  0.17680  0.0730  0.04  2002  0.30783  0.17747  0.0737  0.04  2003  0.31046  0.17814  0.0744  0.04  2004  0.31309  0.17881  0.0751  0.04  2005  0.31572  0.17948  0.0758  0.04  2006  0.31835  0.18015  0.0765  0.04  2007  0.32098  0.18082  0.0772  0.04  168  Chapter 2. Business Sector Data on Outputs and Inputs for Canada 1961-2007 Table 2.13: Prices for Residential Structures and Land and Six Business Sector Capital Stocks  Year t  t P KICT  t P KM E  t P KN R  t P KBI  t PAL  t PBL  t PRL  t PRS  1961  1.00000  1.00000  1.00000  1.00000  1.00000  1.00000  1.00000  1.00000  1962  0.99939  1.01477  1.00592  1.00758  1.04000  1.06956  1.06956  1.00504  1963  0.99650  1.06598  1.03251  1.01000  1.10000  1.14186  1.13138  1.02769  1964  1.00502  1.06773  1.06158  1.02114  1.22000  1.22953  1.17489  1.07312  1965  1.02393  1.10198  1.12281  1.03951  1.36000  1.33883  1.28108  1.13368  1966  1.02944  1.12251  1.19323  1.06286  1.52000  1.50884  1.40204  1.20765  1967  1.07473  1.12792  1.24188  1.08044  1.72000  1.72055  1.53236  1.28518  1968  1.10902  1.13973  1.25227  1.09271  1.90000  1.90357  1.65913  1.31431  1969  1.14444  1.17197  1.32495  1.11812  1.94000  2.06400  1.85677  1.38118  1970  1.19472  1.23029  1.39058  1.14350  1.96000  2.30803  2.07210  1.42615  1971  1.22722  1.27216  1.46812  1.15918  2.00000  2.57526  2.29387  1.53179  1972  1.26956  1.30444  1.55098  1.21176  2.16000  2.91918  2.62704  1.67349  1973  1.30782  1.34535  1.71873  1.32609  2.62000  3.37696  3.07471  1.97123  1974  1.35698  1.51430  2.03419  1.43765  3.42000  4.09586  3.74822  2.36134  1975  1.45867  1.73112  2.27337  1.55864  4.34000  5.13750  4.45312  2.56072  1976  1.45190  1.82918  2.40093  1.66364  5.14000  6.10905  5.00114  2.76853  1977  1.44467  1.98862  2.52980  1.78349  5.92000  6.86323  5.54193  2.87768  1978  1.40748  2.19453  2.71145  1.94126  7.02000  7.73441  6.22122  3.04069  1979  1.38366  2.44348  2.96312  2.15180  8.68000  8.74688  6.89491  3.28046  1980  1.21888  2.69736  3.32520  2.35895  10.94000  10.12430  7.82863  3.55455  1981  1.12880  2.98813  3.68676  2.57823  12.30000  11.79717  9.46319  3.99273  1982  1.16812  3.19883  3.96113  2.77940  12.28000  13.58689  11.12613  4.08226  1983  1.02222  3.26756  3.93090  2.93246  12.00746  14.66837  10.97760  4.25350  1984  0.96342  3.40812  4.08142  3.06736  11.47452  15.05871  11.89508  4.41785  1985  0.89167  3.57237  4.21351  3.14959  10.77083  15.90824  13.04615  4.55564  1986  0.82129  3.70845  4.27520  3.18232  9.98330  16.59100  13.36985  4.90827  1987  0.75277  3.69804  4.47320  3.23099  9.38642  17.77433  15.51626  5.40819  Continued on Next Page. . .  169  Chapter 2. Business Sector Data on Outputs and Inputs for Canada 1961-2007 Table 2.13 – Continued Year t  t P KICT  t P KM E  t P KN R  t P KBI  1988  0.71401  3.69075  4.72840  3.28744  1989  0.64691  3.79390  4.92520  1990  0.61403  3.87130  1991  0.54586  3.74170  1992  0.51043  1993  t PAL  t PRS  t PBL  t PRL  9.22342  19.21871  18.33108  5.78293  3.34034  9.82631  21.04826  20.91090  6.13195  5.08853  3.38833  10.62426  22.79735  24.58825  6.11231  5.00311  3.39008  10.91042  24.32908  24.22111  6.32257  3.88288  4.97541  3.54606  10.69755  24.94996  26.88125  6.39710  0.50164  4.04523  5.03758  3.55229  10.75269  25.42013  28.75011  6.58445  1994  0.48119  4.24754  5.20497  3.72829  11.17139  26.61217  30.94243  6.76485  1995  0.44755  4.44740  5.27332  3.85944  12.03048  28.07898  33.02674  6.76717  1996  0.41150  4.57050  5.43035  3.94827  13.01832  29.02794  32.17431  6.75581  1997  0.39399  4.69221  5.56694  3.81073  14.07653  30.40692  33.04791  6.87512  1998  0.36919  4.90374  5.71450  3.83563  14.93456  31.72819  35.07889  6.95993  1999  0.33873  4.94358  5.82995  3.89582  15.49944  33.12003  36.92678  7.13210  2000  0.32384  5.01831  6.02775  3.98639  15.98327  34.73818  39.81866  7.29782  2001  0.31733  5.17017  6.07934  4.07102  16.50975  36.73098  42.39774  7.48766  2002  0.30560  5.25354  6.18175  4.13975  16.95600  38.19277  47.03668  7.81242  2003  0.27567  4.94953  6.30506  3.90062  17.34990  39.90438  53.73129  8.21290  2004  0.24913  4.85253  6.70389  3.94012  17.74345  42.36014  59.30452  8.71618  2005  0.23077  4.79667  7.12403  3.98736  18.16070  45.97595  68.68515  9.11452  2006  0.21439  4.71044  7.64020  4.05205  18.62402  49.19765  78.15794  9.78750  2007  0.20857  4.58255  8.04719  4.17868  19.25890  52.43974  88.93074  10.49192  170  Chapter 2. Business Sector Data on Outputs and Inputs for Canada 1961-2007 Table 2.14: Quantities of Residential Structures and Land and Six Business Sector Capital Stocks  Year t  t QKICT  t QKM E  t QKN R  t QKBI  t QKAL  t QKBL  t QKRL  t QKRS  1961  1236  11994  27850  13594  5954  6376  10680  28710  1962  1298  12112  29215  13595  5965  6376  10680  29773  1963  1389  12346  30425  14150  5976  6376  10680  30853  1964  1499  12668  31650  15079  5986  6376  10680  31972  1965  1595  13478  33210  15607  5997  6376  10680  33409  1966  1727  14668  34942  16826  6008  6376  10680  34898  1967  1934  16308  37050  18130  5977  6376  10680  36201  1968  2129  17651  38840  18580  5946  6376  10680  37507  1969  2250  18446  40495  19195  5915  6376  10680  39139  1970  2420  19385  42038  20763  5885  6376  10680  41124  1971  2550  20216  43827  20942  5854  6376  10680  42733  1972  2710  20958  45637  20711  5806  6376  10680  44752  1973  2888  21984  47305  20858  5759  6376  10680  47028  1974  3144  23831  49175  23381  5711  6376  10680  49518  1975  3566  25958  51190  28184  5664  6376  10680  52001  1976  4023  27892  53670  27852  5618  6376  10680  54307  1977  4617  29628  55859  28101  5634  6376  10680  57307  1978  5134  30903  58198  29409  5650  6376  10680  60257  1979  5759  32097  60513  30700  5667  6376  10680  63137  1980  6544  33862  63364  34364  5683  6376  10680  65863  1981  7965  35493  66725  34529  5700  6376  10680  68205  1982  10355  37881  70409  35413  5716  6376  10680  70756  1983  11890  38508  73134  30845  5733  6376  10680  72266  1984  13976  38552  75077  29803  5750  6376  10680  74455  1985  16596  38546  76778  31534  5766  6376  10680  76607  1986  19623  39028  78623  32989  5783  6376  10680  79121  1987  23340  39947  79920  34011  5782  6376  10680  82223  Continued on Next Page. . .  171  Chapter 2. Business Sector Data on Outputs and Inputs for Canada 1961-2007 Table 2.14 – Continued Year t  t QKICT  t QKM E  t QKN R  t QKBI  t QKAL  t QKBL  t QKRL  t QKRS  1988  29601  41317  81322  35699  5781  6376  10680  86124  1989  35612  44049  83235  37761  5779  6376  10680  90019  1990  42792  46682  85203  39156  5778  6376  10680  94058  1991  49249  48072  86985  38578  5777  6376  10680  97130  1992  55220  48718  88319  34708  5782  6376  10680  99069  1993  63861  48362  88388  34162  5787  6376  10680  101344  1994  71547  47487  88424  33291  5792  6376  10680  103314  1995  80884  47284  88935  33994  5797  6376  10680  105453  1996  90581  47455  89391  37134  5803  6376  10680  106574  1997  103363  47692  89874  39533  5793  6376  10680  108164  1998  122207  49790  91445  41299  5784  6376  10680  110167  1999  145882  51702  92868  43611  5774  6376  10680  111867  2000  174209  53702  94319  44851  5765  6376  10680  113715  2001  206260  55477  95766  47744  5755  6376  10680  115823  2002  227207  56430  97488  45902  5757  6376  10680  118553  2003  240511  57003  98498  49243  5758  6376  10680  122214  2004  255354  58228  99821  47534  5760  6376  10680  126179  2005  279847  59962  101581  48584  5761  6376  10680  130651  2006  306347  62633  104065  50550  5763  6376  10680  135269  2007  338899  65373  107198  51646  5764  6376  10680  139920  172  Chapter 2. Business Sector Data on Outputs and Inputs for Canada 1961-2007 End of the year current market value starting stocks of inventories for the entire economy and for the government sector are available from the National Balance Sheet Accounts; see CANSIM series V34679 and V32578 (Table 3780004) for the years 1961-2007. Subtracting the government inventory stocks from the total inventory stocks will give us estimates for the value of the business sector beginning of the year inventory stocks for the t . We can subtract the value of inventory change years 1962-2008, V KBI  for 1961 (see CANSIM II series V498100; Table 3800002; Canada, Current Prices, Business Investment in Inventories) from the starting stock of inventories in 1962 in order to extend the value of inventory stock series back to 1961. Diewert (2002), drawing on Diewert and Lawrence (2000), used older national balance sheet information to construct current and constant dollar estimates of beginning of the year stocks of inventories for the years 1962-1999. These series may be used to construct a price of inventory series t for the years 1962-1999. We extended this price series to the years PBI  1961 and 2000-2005 by using the Industrial Product Price Index for Canada and for All Commodities, CANSIM II series V1574377, table 3290039. The t can be divided by the inventory stock price inventory value series V KBI  t , in order to obtain a real beginning of the year business sector series PBI t . The resulting price and quantity series (after stock of inventories, QKBI  normalization so that the price is unity in 1961) are listed in Table 2.13 for t and Table 2.14 for QK t . PBI BI  It is possible to generate an alternative value of inventory stock series by cumulating information on the value of inventory change from the System of National Accounts. Thus the CANSIM II series V498100 estimates the current value of business investment in inventories, which conceptually, should equal the value of inventory change over the year. Using the balance sheet estimates of the starting stock of inventories for 1962 (which was $13,698 million) and the above series, we can cumulate inventory changes and obtain an alternative SNA based estimated value of inventory change, which ended up at $91,315 million at the start of 2007. However, using the balance sheet estimates for the beginning of 2008 value of business inventories, we 173  Chapter 2. Business Sector Data on Outputs and Inputs for Canada 1961-2007 obtained the estimate $215,812 million, which is 2.35 times as big as the implied SNA estimate. Thus the SNA based estimates basically give us an inventory to output ratio that is implausibly low at the end of the sample period. It is true that inventory to output ratios have been falling due to just in time delivery and other inventory management techniques but the number of goods that are being produced has also been growing, which implies an increasing need for inventories. In any case, we will take the balance sheet estimates of inventory stocks as the “truth”.105 Recalling Tables 2.2 and 2.3, a preliminary price series for inventory t in year t is set equal to P t+1 listed in Table 2.13.106 A prechange PII BI  liminary series for the quantity of inventory change in year t listed, QtII , is  t+1 , less the stock set equal to the stock at the beginning of year t + 1, QKBI t . These preliminary series, P t and Qt are at the beginning of year t, QKBI II II  t equals unity in 1961 and these re-normalized then re-normalized so that PII  series are the series which appear in Tables 2.2 and 2.3.  2.5  Primary Input Tax Rates, Balancing Real Rates of Return and User Costs  Non-residential structures (office buildings, factories, etc.) and business land have to pay property taxes on these inputs whereas machinery and equipment and inventory stocks are generally exempt from paying these taxes. Thus it is necessary to take into account property taxes when constructing user costs of capital for business non-residential structures and business land. Information on property taxes for the years 1961-2007 is available from Statistics Canada; see CANSIM II series V499942, Table 105  This choice will lead to an increase in measured Total Factor Productivity compared to estimates that rely on the SNA estimates of inventory change. See Diewert and Smith (1994) for a detailed accounting framework for inventories that is consistent with the Hicks (1961) and Edwards and Bell (1961) model of production and Diewert (2005b) for a critical review of SNA conventions for measuring inventory change. 106 Diewert (2005b) showed that in order to obtain a user cost of inventories that is consistent with other user costs and the measurement of output, inventory changes should be valued at end of year prices.  174  Chapter 2. Business Sector Data on Outputs and Inputs for Canada 1961-2007 3800035 (Real Property Taxes of Local Governments) and CANSIM II series V499841, Table 3800033 (Real Property Taxes of Provincial Governments). We approximate the asset base on which these taxes fall as the total beginning of the year national value of land, residential structures and non-residential structures. Data on these values are available for the years 1962-2008 from the National Balance Sheets and these data are described at the beginning of Section 2.4. These series were summed and the sum was used as the tax base for the sum of the two property tax series, V499942 plus V499841. The resulting property tax rates are reported as the series τPt in Table 2.15107 and it will be used in the construction of the user costs of business sector land and non-residential structures.108 It is of some interest to calculate the average business tax rate for taxes that apply to the use of financial capital in the business sector so we provided estimates for this tax rate by year. These business taxes that fall on the return to capital are defined to be the sum of the following taxes: • Taxes less subsidies on factors of production (CANSIM II series V1992216, Table 3800001) less local government and provincial government property taxes; • Total government taxes on income from corporations and government business enterprises (CANSIM II series V499131, Table 3800007 ) and • Total government taxes on income from non-residents (CANSIM II series V499132, Table 3800007). The sum of the above three sources of general business taxes that fall on capital stock components was divided by the corresponding sum of the beginning of the year value of assets for our six types of business sector asset; 107  The tax rate for 1961 was set equal to the corresponding rate for 1962. This is a very rough approximation to the actual property tax rates on business sector land and non-residential structures since actual property tax rates are different across different sectors and assets. For example, business sector property assets are generally taxed more heavily than household property assets. 108  175  Chapter 2. Business Sector Data on Outputs and Inputs for Canada 1961-2007 t t i.e., the above sum of taxes for year t was divided by P KICT ×QKICT (year t  t t starting value of ICT machinery and equipment) plus P KM E ×QKM E (year  t × QK t t starting value of non-ICT machinery and equipment) plus P KN R NR  t ×QK t (year (year t starting value of non-residential structures) plus P KBL BL  t × QK t t starting value of business sector land) plus P KAL AL (year t starting  t × QK t (year t value of starting stocks value of agricultural land) plus PBI BI  of inventories) and the resulting year t general business tax rate is denoted as τBt , which is listed in Table 2.15. Using the property tax rates τPt , the general business tax rates τBt , the ICT  t , the non-ICT machinery machinery and equipment depreciation rate δICT  t and equipment depreciation rates δM E and the non-residential structures  t , the user costs of ICT and non-ICT machinery and depreciation rates δN R  equipment, non-residential structures, business land, agricultural land and t t , Ut , Ut , Ut t , UM inventories, UICT E NR BL AL and UBI respectively, can be de-  fined as follows,109  t UICT  t t ≡ [rt + τBt + δICT ]P KICT  (2.8)  t t t t t UM E ≡ [r + τB + δM E ]P KM E  (2.9)  t t t t t t UN R ≡ [r + τB + τP + δN R ]P KN R  t UBI  (2.10)  t ≡ [rt + τBt ]P KBI  (2.11)  t t UAL ≡ [rt + τBt + τPt ]P KAL  (2.12)  t t UBL ≡ [rt + τBt + τPt ]P KBL  (2.13)  109  For additional material on user costs and many historical references, see Jorgenson (1989)(1996a) (1996b)and Diewert (2005a)(2006a) .  176  Chapter 2. Business Sector Data on Outputs and Inputs for Canada 1961-2007 where rt is suitable real rate of return that applies to the business sector in year t. In the present study, we will follow national income accounting conventions and will take rt to be the balancing real rate of return;110 i.e., it is the rate of return that is consistent with the year t value of business sector net output being equal to the value of primary inputs used by the business sector in year t, where the user costs (2.8)-(2.13) are used as prices for the beginning of the year capital inputs. Thus rt can be determined as the solution to the following linear in rt equation:  t t t t t t QtIG + PIR QtIR + PIN PCt QtC + PIG R QIN R + PIM E QIM E  t t t t t t t t +PII QtII + PGN QtGN + PXG QtXG + PXS QtXS + PM G QM G + PM S QM S  t t t t t t = PL1 QtL1 + PL2 QtL2 + PL3 QtL3 + [rt + τBt + δICT ]P KICT QKICT  t t t +[rt + τBt + δM E ]P KM E QKM E  110  For most purposes, it is probably preferable to use an exogenous real rate of return in the user costs (2.8)-(2.13) since the resulting prices will probably approximate market rental prices better. For discussion of this topic, see Diewert (2006a). However, in our study of productivity growth, there was little difference in the empirical results if the sample average real rate of return (4.827%) was used in place of the balancing real rate; i.e., in the gross output model, average TFP growth changed from 1.01% to1.25% per year and in the net output model, average TFP growth changed from 1.04% to 1.25% per year. This is similar to results obtained by Diewert and Lawrence (2005)(2006) for Australia. Their first study used the sample average balancing real rate for Australia whereas their second study used the year by year balancing real rates of return. However, Baldwin and Gu (2007; 27) found substantial differences for the Canadian business sector in their TFP growth rates for the period 1961-1981 where their estimated average TFP growth rates increased from the 0.90 to 1.01% per year range using balancing or endogenous interest rates to the 1.18 to 1.26% range using an exogenous interest rate. The differences that Baldwin and Gu (2007; 28) found for the 1981-2001 period were not nearly as large: an increase from the 0.30-0.38% range to the 0.32-0.43% range. Baldwin and Gu (2007; 18) mention that they used a constant real rate of interest equal to 5.1 % in their exogenous interest rate models, which is very close to the 4.827 % real rate that we used in our exogenous real rate computations.  177  Chapter 2. Business Sector Data on Outputs and Inputs for Canada 1961-2007 t t t t t t t t +[rt + τBt + τPt + δN R ]P KN R QKN R + [r + τB + τP ]P KBL QKBL  t t t t +[rt + τBt + τPt ]P KAL QKAL + [rt + τBt ]P KBI QKBI  (2.14)  where the various price and quantity series are defined in the tables of this Chapter.111 The resulting series of balancing real rates of return is listed in Table 2.15. Once rt has been determined, then the six series of user costs defined by (2.8)-(2.13) can also be calculated; these series are also listed in Table 2.15. Note that rt is a real after tax rate of return because we do not include a capital gains term in our user costs and all user costs are evaluated at the average prices for the corresponding investment good for year t.  111 t PXG and QtXG are chained Fisher aggregates of our seven classes of exports of goods, t t t PXS and QtXS are the price and quantity of exports of services, PM G and QM G are chained t t Fisher aggregates of our six classes of imports of goods and PM S and QM S are the price and quantity of imports of services.  178  Chapter 2. Business Sector Data on Outputs and Inputs for Canada 1961-2007 Table 2.15: Business Sector Tax Rates, Balancing Real Rates of Return and User Costs  Year t  t τP  t τB  rt  t UICT  t UM E  t UN R  t UBI  t UAL  t UBL  1961  0.01528  0.03225  0.03650  0.26875  0.21875  0.12903  0.06875  0.08403  0.08403  1962  0.01528  0.03281  0.03578  0.27106  0.22250  0.13034  0.06912  0.08723  0.08971  1963  0.01525  0.03318  0.04750  0.28494  0.24733  0.14696  0.08149  0.10552  0.10954  1964  0.01534  0.03468  0.05119  0.29524  0.25400  0.15745  0.08769  0.12348  0.12445  1965  0.01536  0.03353  0.04779  0.29882  0.25786  0.16222  0.08453  0.13148  0.12944  1966  0.01533  0.03249  0.05042  0.30478  0.26520  0.17510  0.08812  0.14933  0.14823  1967  0.01524  0.03054  0.03555  0.30294  0.24827  0.16210  0.07141  0.13989  0.13993  1968  0.01590  0.03299  0.03652  0.31930  0.25552  0.16944  0.07595  0.16227  0.16258  1969  0.01624  0.03428  0.03291  0.32986  0.26082  0.17758  0.07513  0.16185  0.17220  1970  0.01607  0.03153  0.03428  0.34584  0.27292  0.18519  0.07525  0.16047  0.18897  1971  0.01554  0.03206  0.02737  0.35065  0.27495  0.18640  0.06889  0.14993  0.19306  1972  0.01506  0.03382  0.02783  0.36891  0.28570  0.20072  0.07471  0.16570  0.22394  1973  0.01380  0.03757  0.05809  0.42794  0.34131  0.27990  0.12685  0.28677  0.36963  1974  0.01255  0.04094  0.06004  0.45483  0.39326  0.34100  0.14518  0.38829  0.46503  1975  0.01217  0.03639  0.02907  0.44093  0.38922  0.30106  0.10203  0.33691  0.39882  1976  0.01283  0.03314  0.03320  0.44398  0.41412  0.32335  0.11038  0.40696  0.48369  1977  0.01333  0.03120  0.04273  0.45653  0.46663  0.36292  0.13185  0.51658  0.59888  1978  0.01317  0.03105  0.04783  0.45546  0.52730  0.40389  0.15314  0.64625  0.71202  1979  0.01205  0.03217  0.06274  0.47357  0.62792  0.48763  0.20424  0.92848  0.93563  1980  0.01198  0.03248  0.05146  0.40699  0.66535  0.51280  0.19800  1.04933  0.97109  1981  0.01207  0.03143  0.04056  0.36639  0.70337  0.52742  0.18560  1.03391  0.99164  1982  0.01218  0.02773  0.02063  0.35463  0.67952  0.47628  0.13440  0.74340  0.82251  1983  0.01273  0.02880  0.04443  0.33845  0.77758  0.57532  0.21474  1.03215  1.26088  1984  0.01254  0.03198  0.05234  0.33220  0.85111  0.64470  0.25864  1.11142  1.45859  1985  0.01260  0.03173  0.05501  0.31196  0.90318  0.67898  0.27321  1.07001  1.58038  1986  0.01276  0.03070  0.05025  0.28474  0.91858  0.66783  0.25761  0.93553  1.55481  1987  0.01262  0.03305  0.06412  0.27517  0.97846  0.77381  0.31395  1.03052  1.95141  Continued on Next Page. . .  179  Chapter 2. Business Sector Data on Outputs and Inputs for Canada 1961-2007 Table 2.15 – Continued Year t  t τP  t τB  rt  t UICT  t UM E  t UN R  t UBI  t UAL  t UBL  1988  0.01253  0.03318  0.05905  0.25936  0.96079  0.79750  0.30321  0.96627  2.01340  1989  0.01267  0.03323  0.04822  0.22971  0.94929  0.78175  0.27209  0.92490  1.98116  1990  0.01303  0.03249  0.03650  0.21200  0.92299  0.74964  0.23376  0.87139  1.86982  1991  0.01360  0.03066  0.01785  0.17872  0.81795  0.64092  0.16443  0.67758  1.51093  1992  0.01418  0.03054  0.02931  0.17425  0.89545  0.70016  0.21221  0.79188  1.84691  1993  0.01418  0.03271  0.02850  0.17325  0.94110  0.71929  0.21742  0.81059  1.91629  1994  0.01392  0.03528  0.04441  0.17635  1.06954  0.84170  0.29711  1.04577  2.49121  1995  0.01369  0.03729  0.05185  0.16943  1.16489  0.90509  0.34405  1.23716  2.88751  1996  0.01371  0.04085  0.05500  0.15962  1.23083  0.97235  0.37844  1.42627  3.18027  1997  0.01366  0.04436  0.04773  0.15238  1.24910  0.97949  0.35092  1.48857  3.21548  1998  0.01386  0.04021  0.04498  0.14121  1.27485  0.97115  0.32674  1.47920  3.14252  1999  0.01395  0.04567  0.04749  0.13315  1.32792  1.04185  0.36292  1.66008  3.54736  2000  0.01320  0.05092  0.06134  0.13434  1.44719  1.19200  0.44748  2.00514  4.35800  2001  0.01296  0.04095  0.06042  0.12901  1.43815  1.13880  0.41265  1.88744  4.19919  2002  0.01255  0.03963  0.07025  0.12765  1.50961  1.21244  0.45488  2.07595  4.67601  2003  0.01239  0.04193  0.06610  0.11536  1.41639  1.22833  0.42137  2.08922  4.80516  2004  0.01212  0.04536  0.07973  0.10916  1.47468  1.42330  0.49287  2.43457  5.81222  2005  0.01167  0.04517  0.08577  0.10307  1.48894  1.55589  0.52207  2.58973  6.55621  2006  0.01110  0.04621  0.07823  0.09493  1.43474  1.62001  0.50423  2.52426  6.66814  2007  0.01065  0.04605  0.07965  0.09316  1.40462  1.71843  0.52524  2.62584  7.14985  Average  0.01351  0.03569  0.04827  0.26874  0.78978  0.64956  0.23351  0.95212  1.91810  180  Chapter 2. Business Sector Data on Outputs and Inputs for Canada 1961-2007 Note that the sample average of the balancing after tax real rates of return rt  was a rather large 4.827% per year.112 The average property tax rate τPt  was 1.351% while the average business tax rate on assets was 3.569%. Thus the before business tax real rate of return averaged 8.396%. Thus it appears that governments are taking about 42.5% of the before tax return to capital assets on average.113 However, it must be kept in mind that these balancing rates of return may not be very reliable; they contain the net effect of all the measurement errors that were made in constructing this data set. The volatility in the above real rates of return is a source of concern since it is likely that a considerable proportion of the volatility is caused by various measurement errors. The volatility in the real rates of return also causes volatility in the user costs and possible volatility in productivity growth rates. However, we repeated our productivity calculations using a constant after tax real rate of return (equal to the sample average real rate of 4.827%) and found no material difference in our productivity growth rates. Hence the volatility in the productivity growth rates is mainly due to volatility in our output measures. 112 The corresponding balancing real rate of return for Australia averaged around 3 percent; see Diewert and Lawrence (2006). Normally, after tax real rates of return are in the 1 to 3 percent rate whereas our average rate is close to 5 percent. This suggests that our estimates of the value of output are too high or that the value of labour input are too low or that our estimated asset values for business sector capital inputs are too small. We think that the last possibility is the most probable one. Using the data tabled for this chapter, we calculated a business sector nominal and real value of business sector output and we also calculated the corresponding business sector nominal and real capital stock inputs where the real measures were calculated using chained Fisher indexes. We found that the nominal business sector capital output ratio fell from 2.417 in 1961 to 1.861 in 2007 while the real capital output ratio fell from 2.417 in 1961 to 1.538 in 2007. These falls in the capital output ratio seem unlikely. See Diewert and Fox (2001) for a discussion of output mismeasurement problems. 113 This relatively high rate of business taxation has two negative effects: (i) it raises the user cost of capital and hence lessens the beneficial effects of capital deepening and (ii) the high rates lead to a relatively large loss of productive efficiency; i.e., the deadweight losses of such large tax rates are likely to be large. See Diewert and Lawrence (2002) for a methodology for estimating the deadweight losses due to capital taxation.  181  Chapter 2. Business Sector Data on Outputs and Inputs for Canada 1961-2007  2.6  Sources of Error  There are many problems with the data constructed in this paper. Some of the more important possible sources of error are listed as follows, • Our adjustments for converting final demand prices (those facing the final demanders of the goods and services produced by the business sector) into basic prices (prices facing the producers of the goods and services) were rather crude and some aggregation error will be associated with our procedures. In particular, only crude adjustments for the effects of indirect taxes on the components of consumption were made. Also our method for estimating the net supplies of the business sector to the non-business sector is rather indirect and subject to some error.114 • Our tax adjustments for the price of imports and exports were also not completely satisfactory due to various aggregation errors; i.e., we were not able to assign taxes accurately to the various components of imports and exports. • Our measure of labour input relies on the Statistics Canada KLEMS program estimates for quality adjusted labour and there may be some amount of error in these estimates. In particular, it is very difficult to account for the hours of work and labour compensation for the self-employed. • It proved to be difficult to reconcile balance sheet information with investment information. Our treatment of investment and capital services was highly aggregated and hence contains some aggregation errors. We also relied heavily on the Statistics Canada Balance Sheet estimates and these estimates are highly aggregated; in particular, there is not enough detail on the allocation of land. Moreover, the Balance Sheet stocks appear to give asset values that are too small.115 114  In particular, we did not have access to chained price indexes for the non-business sector for the years prior to 1997 and this will lead to some aggregation errors. 115 Evidence of this possible undercounting of asset values in the Balance Sheet accounts  182  Chapter 2. Business Sector Data on Outputs and Inputs for Canada 1961-2007 • Our treatment of property taxes is very approximate. • Our user costs of capital were constructed using a particular set of assumptions (no capital gains and endogenous real rates of return) and these assumptions are not universally accepted. • The roles of infrastructure capital and R&D investments were not taken into account. • The role of resource depletion was also not taken into account. The next international version of the System of National Accounts (SNA) will recognize capital services in the production accounts. This will be a big step forward since it will allow inputs in the SNA production accounts to be decomposed into price and quantity components and hence the revised SNA will facilitate the development of productivity accounts for each country that implements the revised SNA. However, just introducing capital services into the SNA will not be sufficient in order to develop accurate sectoral productivity accounts. The revised SNA also needs to consider the following problems: • More attention needs to be given to the development of basic prices by industry and by commodity; i.e., we need accurate information on the exact location of indirect taxes (and commodity subsidies) by commodity and industry on both outputs and intermediate inputs. • In order to deal adequately with the complications introduced by international trade, the existing Input Output production accounts need to be reworked so that the role of traded goods and services can be tracked by industry. are the declining capital output ratios that are implied by our data. Moreover, the assessed value of real property (land and structures) in British Columbia for 2007 was just over one trillion dollars. If we add up the value of land and structures in the National Balance Sheets for the beginning of 2007, we get a value of about 4 trillion dollars. If we multiply the British Columbia value by a factor of 8, it seems that the national value of real property should be equal to about 8 trillion instead of the 4 trillion in the accounts.  183  Chapter 2. Business Sector Data on Outputs and Inputs for Canada 1961-2007 • The treatment of inventory change in the present SNA seems inadequate for the needs of productivity accounts. Inventory change should be integrated with the balance sheet accounts and the user cost accounts. • The investment accounts need to be integrated with the corresponding balance sheet accounts, both in nominal and real terms. • The treatment of land in the balance sheets requires additional work; i.e., there are problems in obtaining information on the quantity of land used by each industry and sector and valuing the land appropriately.116 • Difficult decisions must be made on the exact form of the user cost formula to be used when measuring capital services; i.e., the revised SNA should make specific recommendations on how user costs should be constructed so that some measure of international comparability can be achieved in the accounts. • The problems involved in making imputations for the labour input of the self-employed (and unpaid family workers) should also be addressed. The introduction of capital services into the SNA will provide challenges for statistical agencies. However, as national statistical agencies make productivity accounts a part of their regular production of the national accounts, there will be benefits to the statistical system as a whole since a natural output of the new system of accounts will be balancing real rates of return by sector or industry. These balancing real rates of return will provide a check on the accuracy of the sectoral data: if the rates are erratic or very large or very small, this can indicate measurement error in the sectoral data and hence will give the statistical agency an early indication of problems with the data. 116  There are some difficult conceptual and practical problems involved in separating structure value from land value; see Diewert (2007a) for a discussion of some of these problems.  184  Chapter 2. Business Sector Data on Outputs and Inputs for Canada 1961-2007 Statistics Canada already has an extensive productivity program. It is to be hoped that as the program evolves in the future, the data will be presented to the public in some detail and hopefully, at some level of aggregation, revised series will be made available back to 1961.117  2.7  Recommendations for the Statistics Canada Productivity Program  There are substantial difficulties in accessing data on the prices and quantities of primary inputs used by the business and non-business sectors from CANSIM. Also it is evident that the coverage of primary input usage by industry by Statistics Canada is not nearly as extensive as the corresponding coverage of gross outputs and intermediate inputs. With the next revision of the System of National Accounts recommending a decomposition of gross operating profits into price and quantity components, it seems time for Statistics Canada to devote more effort into improving measurement with respect to primary inputs used by industries in the Canadian economy. Without accurate information on the flow of labour and capital services by industry, governments and businesses will not be able to plan ahead for Canada’s future. It is important to know past trends in TFP growth by industry so that future trends can be anticipated and so that budgetary planning can be carried out on a more rational basis. Hopefully, other national departments interested in Canadian productivity growth (the Bank of Canada, the Department of Finance and Industry Canada to name a few) will support an initiative that will put more resources into the hands of Statistics Canada so that they can provide better information on productivity growth. Important priorities for improving Statistics Canada’s productivity program include the following ones: 117 It is important to have data back to the early 1960’s since the 1950’s and 1960’s were decades of very high productivity growth. Hence if we want to explain the productivity slowdown that took place in the 1970’s, it is important to have comparable data for the 1960’s.  185  Chapter 2. Business Sector Data on Outputs and Inputs for Canada 1961-2007 • The National Balance Sheet accounts need to be fully integrated with the productivity program; i.e., Statistics Canada collects information on 30 classes of assets with some degree of industry breakdown but publishes only a crude four type of asset by households, corporations and governments breakdown. The household sector needs to be split into a self employed business component and a “consumer of goods and services” component and the corporate sector should be decomposed into industries with price and quantity information for the 30 classes of asset made available by quarter and by industry. • The National Balance Sheet information on the value of land, residential structures and non-residential structures needs to be greatly expanded so that more information on the price and quantity of real property by industry is made available.118 The problems associated with finding adequate constant quality price indexes for residential and non-residential structures are formidable119 but given the importance of real property in the Canadian economy, it is necessary to put additional resources into this area of economic measurement. • The KLEMS program has developed very useful price and quantity information on 56 types of labour used by the Canadian business sector but has only made a highly aggregated form of this information (the information on three types of labour service used in this study) available on CANSIM II. However, this information is extremely useful to the general community. If it is felt that the disaggregated information is not reliable enough to be released in this form, then it should be aggregated up and released at some level of detail that is more detailed than the present three price and quantity series that are available on CANSIM II. Furthermore, corresponding information on disaggregated labour input by type of worker should also be provided 118  We have some concerns that the National Balance Sheets are perhaps missing some growth in the value of real assets. Indirect evidence that points in this direction includes declining capital output ratios for the Canadian business sector. Part of the problem may be the very high depreciation rates that are being used by the KLEMS program. 119 For a review of these problems, see Diewert (2007a).  186  Chapter 2. Business Sector Data on Outputs and Inputs for Canada 1961-2007 for the non-business sector.120 • More information on the incidence of taxes needs to be provided in the Input-Output accounts; i.e., we need to know exactly in which cell of the Input-Output accounts various indirect and direct taxes are applied.121 Not only is this information required to reconcile final demand indexes with production accounts indexes, it is also required in order to evaluate the efficiency of our tax system.122 • The estimations in Chapter 1 has shown that over short periods of time, changes in the real price of exports and imports can have substantial effects on living standards. The methodology used in the paper is applied only to the aggregate business sector. In Section 1.8 of Chapter 1, we showed how the methodology can be extended to the industry level but in order to implement this methodology to show the effects of changes in the terms of trade by industry, it will be necessary to expand existing Input-Output tables to include information on exports produced and imports used by industry.123 Government departments who have an interest in productivity measurement by industry will have to consider whether it would be worthwhile extending the production accounts in this direction. These extended accounts would enable researchers to study issues related to outsourcing and globalization in a more scientific manner. • Baldwin and Gu (2007; 15-22) have a nice discussion about many of 120 Statistics Canada has been a pioneer in developing and publishing very detailed information on the prices and quantities of outputs produced and intermediate inputs used by industry back to 1961 in its input output tables. What we are asking here is that these tables be extended to also cover the 56 types of labour input and 30 types of capital input that are being used in the Statistics Canada KLEMS program. Note that extending the Input-Output tables to cover primary input allocations will also involve extensions to the corresponding final demand accounts, which in the case of inputs, will be corresponding household and government supplies of labour and capital. 121 Recall that we were forced to make guesses about the incidence of various consumption, import, property and capital taxes in order to reconcile final demand prices with producer prices. For additional material on how to accomplish this reconciliation, see Diewert (2006b)(2007b). 122 See Diewert (2001; 97-98) for an elaboration of this point. 123 Diewert (2007b)(2007c) explains these expanded production accounts in more detail.  187  Chapter 2. Business Sector Data on Outputs and Inputs for Canada 1961-2007 the unresolved issues in constructing an appropriate user cost formula in order to price capital services and note that an unambiguous “best practice” measure has not yet emerged. Given this state of affairs, we recommend that Statistics Canada provides not only the actual user costs by asset and year that they used in the KLEMS program but that they provide supplementary information on the various ingredients (interest rates, property taxes, business taxes, asset price appreciation terms and asset prices) that go into the making of the user costs so that researchers can construct their own preferred versions of user costs. Eventually, a view will form on what the “best practice” user cost is but we are not at this point yet and hence it is essential that Statistics Canada provides analysts with information on the various components of user costs.  188  Chapter 3  Does Lobbying Affect Antidumping Case Determinations in Canada? 3.1  Introduction  A large number of petitions against dumping124 were filed globally in the last decade125 indicating that many countries are involved either as plaintiffs or defendants in antidumping litigation. Canada is no exception. In fact, Canada is a seasoned user of antidumping legislation. According to the World Trade Organisation (WTO), Canada ranks 6th among its 45 reporting members in the total number of antidumping initiations from 1995 to 2007.126 GATT Article 6 of the WTO stated that all countries are allowed to introduce their own antidumping legislation. Thus, a domestic firm that believes it is victimised by dumping can file an antidumping petition with its government. In Canada, antidumping cases are determined according to the antidumping legislation mandate which states that factors such as prices and market share cannot be used in the injury determination. Most importantly, the determinations are supposed to be immune from political pressure. This suggests that there should be no need for firms to apply 124  Dumping means an export good is being sold below the cost of its production or below the price it is normally charged in the country of production. 125 According to the World Trade Organization Antidumping Gateway, there were over 2,800 antidumping initiation filed by member countries between 1998 to 2008. 126 Statistics taken from the World Trade Organisation Antidumping Gateway.  189  Chapter 3. Does Lobbying Affect Antidumping Case Determinations in Canada? political influence after they file antidumping petition. In fact, however, Canadian firms have often lobbied antidumping agencies after they filed petitions. This indicates that Canadian antidumping legislation mandate may not be as apolitical as it was designed to be. This paper examines antidumping cases in Canada from 1996 to 2003, focussing on the relationship between the outcomes of the cases and various economic and political factors, in order to determine whether the mandate is independent of political influences. Using the Canadian Lobbyists Registration data, we construct a new political variable to test for the effect of the political influence on Canadian antidumping cases determinations. We find that antidumping case determinations in Canada are not independent of political influence. This paper contributes to the literature in two ways. First, while several previous studies have found evidence that political influence affects antidumping cases in the United States, this is the first paper to study the role of lobbying in affecting antidumping cases in Canada. Finger et al. (1982) is the pioneer of empirical studies in antidumping and found some evidence that political pressure was able to influence U.S. antidumping cases.127 Leipziger and Shin (1991), Hansen and Park (1994), Hansen and Prusa (1997), Bown, Hoekman and Ozden (2003), Drope and Hansen (2004), Francois and Niels (2004), and Evans and Sherlund (2006) also found evidence of non-economic influences on U.S. antidumping case determinations. In particular, Hansen and Park (1994), Hansen and Prusa (1997), and Drope and Hansen (2004) found that political factors are significant in affecting the decisions made by the antidumping agencies in the U.S. These studies used a large variety of political variables ranging from indus127 Most subsequent empirical studies shifted the focus towards economic determinants and their influences on antidumping cases, such as Herander and Schwartz (1984), Feinberg (1989), Coughlin et al. (1989), Leidy (1997), Prusa and Skeath (2001), Knetter and Prusa (2003) and Feinberg (2005). These studies found that microeconomic variables such as industry capacity utilisation rate, wage to value-added ratio and unionisation coverage; and macroeconomic variables such as inflation, real exchange rate and real GDP growth are significant in influencing the decision of filings.  190  Chapter 3. Does Lobbying Affect Antidumping Case Determinations in Canada? try concentration to number of political party representatives to political contributions by industry to represent political pressures.128 However, a problem with these variables is that it is hard to tell if the political pressure is truly targeted to influence antidumping cases.129 Second, unlike previous empirical studies that have had to rely on proxies to construct political pressures, this paper is able to utilise a new set of data on Canadian lobbyists. The details of the Canadian Lobbyists Registration data make it possible to identify the lobbyists who lobbied antidumping legislation agency, thus making the political variable constructed in this paper more appropriate than those previously used. The rest of the paper is organised as follows: Section 3.2 introduces the background of antidumping legislation, petition filing process and lobbyists registration in Canada. Section 3.3 includes the empirical analysis with details on data choices and methodology, and the discussion of results. Finally, in Section 3.4, there is a brief conclusion of the empirical analysis.  3.2  Antidumping and Lobbying  Dumping is said to occur if the exporter charges a price that is lower than either the price at which it is sold in the exporter country or the cost of production. Dumping can lead to consequences such as reduced domestic prices, reduced sales, reduced market share or reduced profit, all of which can injure the domestic producers. Domestic producers who believed they are victimised can file antidumping petitions with the government agency 128 There is also a small theoretical literature that investigates the role of political influence in antidumping cases. For example, Anderson (1994) developed a two-stage game which allows the firm to make a decision on filing in the first stage and then consider whether to lobby the antidumping agency for a specific outcome in the second stage. Gasmi et al. (2004) also examined a model in which a domestic firm can choose to lobby an agency that administers international antidumping legislation in order to influence a determination. 129 Nelson (2006) pointed out that empirical works in antidumping have very little connection with the theoretical development due to reasons like this.  191  Chapter 3. Does Lobbying Affect Antidumping Case Determinations in Canada? responsible for antidumping cases. Upon the receipt of a petition, the agency will start an investigation and can levy antidumping duties if injury from dumping is found. In Canada, the Special Import Measures Act (SIMA) that went into effect on December 1, 1984, states that industries producing the same goods as imports and suspect the imports are dumped, may file complaints with the Canada Border Services Agency (CBSA). Suspected cases are usually filed by the industry or by a representative firm that believes it has been victimised. At least 25% of the Canadian production of that industry must support the complaint, and the proportion of support must also be larger than the opposition. A complaint must include information on the domestic goods, the imports, the domestic industry and the conditions in the Canadian market. The complaint must also include evidence of injury. If satisfied that the complaint is properly documented, CSBA will launch an investigation to determine whether the named imports are dumped by the named countries. CSBA is also responsible for determining the amount of antidumping duties to be imposed.130 Another antidumping agency that works independently from the CBSA is the Canadian International Trade Tribunal (CITT). The tribunal consists of nine full time members appointed by the government. Their responsibilities include conducting inquiries into whether dumped imports have caused injury to domestic producers and hearing appeals of cases determined by CSBA. It also conducts antidumping injury inquiry through public hearings and is responsible for the final decision on injury after the inquiries. The SIMA restricts the factors that can be included in the determination of injury to include only domestic production values, market prices, market share of the complaining firms, profits, employment level, capital utilisation and investment expenditure. 130  See CBSA website for details of antidumping cases investigation procedure.  192  Chapter 3. Does Lobbying Affect Antidumping Case Determinations in Canada? Upon preliminary affirmative ruling, a temporary duty will be imposed by Canada Border Service Agency. In the case where a final injury decision is issued in favour of the Canadian industry, an antidumping duty will be implemented on imports from the named countries that were dumped.131 Exporters named in the investigation can alternatively, before full investigation is instigated, choose to work on an undertaking that commits them to change their pricing practices that are causing harm to the Canadian industry. Any proceedings or rulings of the cases are published in the Canada Gazette. Over 150 complaints had been filed with the Tribunal and were given rulings since the introduction of the Special Import Measures Act in 1984. For the past two decades, about two-third of the antidumping cases filed in Canada eventually led to injury findings and only a few of the cases resulted in a price undertaking. Table 3.1 shows a summary of the antidumping cases filed from 1990 to 2006 and Table 3.2 shows the country classifications of the named countries from 1990 to 2004.132 There were a total of 87 complaints that named 74 countries during this period.133 Most of the antidumping complaints in Canada were initiated by four industries: steel products, textiles, food products and chemical products. The most named trade partners were the U.S. followed by China and Taiwan. European countries like the United Kingdom and Germany were named regularly but less frequently. However, the U.S. and France were the only countries that ever accepted undertaking during this period.  131  This is different from the U.S. where duties begin at the stage before ruling. Classification is from International Monetary Fund Economic Outlook 2007. Bown, Hoekman and Ozden (2003) reported that in the U.S., developing countries are often named and they believed this is partly because those countries have limited resources to retaliate, are unable to raise the already high protection further and are unfamiliar with litigation procedure for appeal. However, in Canada, advance economies have been named slightly more often than developing countries have been named. 133 Author’s own recoding, details described in the Section 3.3. 132  193  Chapter 3. Does Lobbying Affect Antidumping Case Determinations in Canada?  Table 3.1: Antidumping Cases in Canada 1990-2006 Times Namedb  Injury Found  Undertaking Accepted  United States  56  29  8  China  24  15  0  India  11  8  0  Taiwan  14  7  0  Country  Germany  13  8  0  United Kingdom  11  8  0  France  12  8  1  Other Countries  175  122  0  Total  316  205  9  b Each case may consist of multiple named countries and multiple NAICS6 industries, the combination of one country and one industry is considered as one count.  194  Chapter 3. Does Lobbying Affect Antidumping Case Determinations in Canada?  Table 3.2: Antidumping Cases in Canada by the Named Country’s Economic Status 1990-2004 Initiation Year  c  Advanced Economiesc  Developing Countries  1990  6  9  1991  12  0  1992  39  17  1993  18  6  1994  3  0  1995  36  1  1996  3  7  1997  8  5  1998  5  6  1999  11  18  2000  8  16  2001  9  16  2002  1  4  2003  7  15  2004  6  5  Total  172  125  Classification comes from International Monetary Fund’s World Economic Outlook 2007.  Canada’s antidumping legislation process is quite different from that in the United States. First, no ex-ante duties are collected, and the named countries are allowed to raise export prices and accept undertaking before any final decisions are made. Second, named countries with dumping margins less than a threshold set by CBSA are excluded from injury determination.134 Thus, not all named countries will have duties imposed if the determination is affirmative. However, once an affirmative decision is made, Canada tends to impose a heavier duty to eliminate the dumping margin. 134 U.S. calculates its dumping margin using a practice called “zeroing” which eliminates any negative dumping margins and includes all named countries, thus possibly overestimating the margin.  195  Chapter 3. Does Lobbying Affect Antidumping Case Determinations in Canada?  Figure 3.1: Antidumping Cases in Canada by the Named Countries 19902006  200  175  180  160  140 122 120 Named Injury found Undertaking accepted  100  80  60  56  40 29  24 15  20  11  8 0  14 8  13 7  0  11  8  0  12  8  0  8 1  0  0  0 US  China  India  Taiwan  Germany  UK  France  Other  196  Chapter 3. Does Lobbying Affect Antidumping Case Determinations in Canada? In Canada, the Lobbyists Registration Act became effective in 1995 and required all lobbyists to register with Industry Canada. They must also list which company they represent, what purpose they have and what actions they plan to take. Actions usually include meetings with legislators and attending committee meetings where the politicians who have the power to influence policies attend. Table 3.3 shows the number of active lobbyist registrations in Canada from 1996 to 2005, and it can be seen that the total number of active lobbyists that registered with Industry Canada has doubled since 1996. Unlike direct political contributions, there is no restriction on lobbying expenditure. Thus, industries can spend as much on lobbying as they prefer to reach their goals and may hire as many lobbyists as they wish to represent them. Table 3.3: Active Lobbyist Registrations in Canada 1996-2005  d  Year  Consultant  In-house (Corporate)  In-house (Organisation)d  1996-1997  1774  380  295  1997-1998  2012  367  327  1998-1999  2060  352  362  1999-2000  2401  336  382  2000-2001  2682  300  363  2001-2002  3003  234  357  2002-2003  3095  296  316  2003-2004  3287  298  330  2004-2005  3417  192  271  Data from Office of Registrar of Lobbyists, recorded as of March 31st in the year  In Canada, firms and industries that petition for antidumping are observed hiring lobbyists to lobby CITT for their cases. Table 3.4 shows the percentage of Canadian lobbyists between 1996-2003 who had indicated they planned to lobby the CITT. 18.7% of all lobbyists on average showed intention to lobby the CITT. The percentage in manufacturing is higher, with  197  Chapter 3. Does Lobbying Affect Antidumping Case Determinations in Canada? 23.8% of all lobbyists on average planned to lobby the CITT.135 Table 3.4: Percentage of Lobbyists that Lobbied Canadian International Trade Tribunal 1996-2003  All  Manufacturing Only  Year  Lobbies CITT  Lobbies CITT & Foreigne  1996  15.5%  0.3%  1997  17.5%  0.2%  1998  18.1%  0.3%  1999  20.2%  0.7%  2000  20.5%  0.9%  2001  19.0%  0.9%  2002  18.8%  1.0%  2003  19.8%  1.1%  1996  18.9%  0.6%  1997  21.4%  0.5%  1998  22.4%  0.8%  1999  26.3%  2.2%  2000  26.4%  2.7%  2001  24.5%  2.3%  2002  24.6%  2.7%  2003  25.5%  3.3%  e Author’s own recoding, each registration is counted as one lobbyist, raw data from Public Registry of Lobbyists, Industry Canada  135  Blonigen and Prusa (2003) suggested that directly unproductive profit-seeking activity like lobbying is more explicit in administered protection like antidumping duties than in traditional protection like tariffs.  198  Chapter 3. Does Lobbying Affect Antidumping Case Determinations in Canada?  3.3  Empirical Analysis  In this section, we analyse the factors influencing antidumping case determinations in Canada during 1996-2003. The economic and political determinants that can affect an affirmative determination are drawn from previous empirical analyses of antidumping case determinations. A key assumption underlying our work is that firms only file based on their likelihood of success in getting an affirmative decision.136 We assume that the relation between affirmative determinations and the independent variables can be described by ni = F (pi , ci , ei , di )  (3.1)  where ni is the dependent variable representing number of antidumping cases by industry i that are determined affirmative, pi represents the political variables, ci represents the determinants CITT uses for injury determination, ei represents the additional economic determinants and di represents dumping indicators. The dependent variable is constructed as the number of antidumping petitions that were determined affirmative. Data on petitions are taken from the Historical Listing of Antidumping Cases in the Canada Border Services Agency website. Cases that had been filed, investigated and ruled on in Canada are listed in the data base with the Harmonised System (HS) 10 digits product code of dumped products, the initiation dates of the investigation, the specific accusation (dumping or subsidised), the countries named, the disposition, the dumping margins and the final rulings. To construct this variable at industry level, each HS 10 digits product is matched with its corresponding North America Industry Classification System 6 digits code (NAICS6) using a concordance provided by Statistics Canada. Each petition may have more than one exporting country named and more than one 136  This assumption means industries self-select into filing antidumping petitions given the expected likelihood of success in affirmative decision.  199  Chapter 3. Does Lobbying Affect Antidumping Case Determinations in Canada? product identified as dumped. In the construction of the dependent variable, each named country in each complaint is to be treated as an individual petition. For example, the complaint initiated on March 1, 1996 named two products, HS codes 4823.51.00.00 and 4820.10.00.00; and named two countries, Indonesia and Brazil. The equivalent NAICS6 industry codes for these products are respectively, 323119 and 323116. Therefore, the number of petitions would be two for each of the two industries. By the same logic, and excluding complaints that were terminated before decision or complaints filed by non-manufacturing industries, there were 163 petitions filed by 50 industries, among which 127 cases were determined affirmative or had undergone price undertakings. Hansen and Prusa (1997) suggested that cases that were settled (went through price undertakings) should also be included because political pressure plays heavily in explaining these cases.  3.3.1  Economic and Political Determinants  According to the CITT mandate, all determinations of injury must be made purely based on evidence that the imports are dumped. Reasons that are unrelated to the discrepancy of import prices relative to prices in the exporter countries should not be considered.137 Profits, capacity utilisation138 and capital investment139 are included in the analysis as industry-specific CITT economic injury determinants. Instead of the total number of employees, the annual growth rate in employment is used for analysis, as it is a better indication of industry injury than the total employment. Average labour earnings growth and output growth 137 Industry specifics such as profits, employment, capacity utilisation and investment can be considered for determinations of injury as they are indirectly affected by imports dumped. However, changes in domestic demand, technology, export performance of Canadian firms, volume and prices of non-dumped imports, trade restrictiveness and performance of firms should not be included. 138 A capital utilisation rate that is less than 100 percent would indicate the industry has excess capacity available to increase production for replacing any displaced imports. 139 Capital investment data are only readily available in subcategories of NAICS, thus requires re-weighing by the industry’s proportion in total output of the subgroup to determine an approximate amount of capital investments.  200  Chapter 3. Does Lobbying Affect Antidumping Case Determinations in Canada? are also included as additional economic determinants. Industry export intensity is used to proxy the economic interdependence described by Finger et al. (1982). Export intensity is calculated as the ratio of domestic exports to manufacturing shipments. Therefore, a high export intensity indicates an industry is highly dependent on exports and will probably experience less injury in case of dumping. Knetter and Prusa (2003) and Feinberg (2005) both found that macroeconomic indicators such as currency depreciation can be influential in antidumping cases. Thus a year dummy each for 1997 and 2001 is included to indicate a depreciation in Canadian dollar relative to the year before. The former is the year of Asian financial crisis and the latter is the year terrorists attacked the U.S. Both years saw overall deterioration in world economy and weakened Canadian dollars.  140  The “interest group” theory assumes that interest groups influence politicians for desirable policy outcome. The channel of influence is not specific. It can be contributions, lobbying or some other type of political influence. In earlier empirical studies, political influence was often represented by variables that are not exactly political in nature. However, as suggested by Baldwin (1989), better measures are needed to illustrate the extent of political pressures exerted by economies than indirect measures such as industry concentration ratio or sales growth. To address this, in this analysis, a political variable is constructed using the Canadian Lobbyists Registration database from 1996 to 2003. The Lobbyists Registration database discloses details of all active lobbyists including what particular concerns the lobbyists have and which government agencies they plan to lobby. Thus for this analysis, lobbyists who have listed their lobbying concerns as “Special Import Measures Act (SIMA)”, “antidumping regime” or those who planned 140  Knetter and Prusa (2003) also showed that real GDP growth increase antidumping cases in Canada but real GDP growth of the Canadian economy has been very steady over the sample period. It would be obsolete to include as macroeconomic determinant.  201  Chapter 3. Does Lobbying Affect Antidumping Case Determinations in Canada? to lobby CITT are selected. Out of the 163 antidumping cases initiated in the sample period, about one-quarter of them had lobbyists that lobbied CITT or had SIMA concern. Each of these lobbying firms is manually matched with its NAICS6 industry code by searching through the Canadian Companies Capability Database.141 Drope and Hansen (2004) suggest that lobbyists data should be complemented by political contributions data to identify how much industries spent in seeking protection. Therefore, a second political variable formed by the amount of lagged political contributions is included in the analysis. The total amount of political contributions by each industry is matched with the corresponding NAICS6 code as in the case of the lobbyists data.142 Political contributions may not directly influence the antidumping case determinations as they differed from direct lobbying in nature. However, recent literature, like Austen-Smith (1995) has suggested that political contributions and lobbying are complementary, i.e. the hard-money contributions are used as access fee for future lobbying. All values of the political contributions are transformed into logarithms for easier interpretation. As CITT mandate is assumed to be independent of political pressure, therefore, any political determinants should not be significant in determination. In other words, if the antidumping legislation is correctly implemented, the effects of both lobbying and contributions variables should be zero. Many previous empirical studies (such as Finger et al. (1982) and Hansen and Prusa (1997)) included variables that reflected the collective action hypothesis (Olson (1965)). The hypothesis says a more concentrated industry is able to organise more easily, and that organised industries fare better in 141 If a firm is not found in the Canadian Companies Capability database, other online sources such as Lexis-Nexis Academic and Manta will be searched. If the industry code is still unfound after these sources are exhausted, information listed on the company websites will be checked and an appropriate code will then be assigned. 142 Political contributions data are taken from Election Canada website. However, unlike the lobbyists data, political contributions by businesses and trade unions cannot be identified by what the donations were for, since each contribution will only be associated with the donor’s name and the intended party.  202  Chapter 3. Does Lobbying Affect Antidumping Case Determinations in Canada? obtaining affirmative determinations than the non-organised ones. While these studies did not find strong support on the collective action theory, an industry concentration ratio is included in this analysis to test the collective action hypothesis in Canadian antidumping case determinations. A Herfindahl-Hirschman index would be the best indicator of concentration of an industry as it is weighted by the market share. However, due to limited availability of data, we use a four-firm industry concentration instead. In addition to the CITT economic injury determinants and other economic determinants and political determinants, import price growths and import penetration ratios are included as dumping indicators. Import unit value, calculated by dividing the total import value with total import volume, is used as a proxy for import prices.143 Import penetration ratio is calculated as the imports of goods as a ratio to the domestic consumption, i.e. how much of domestic demand is satisfied by imports.  144  Industries that  have high import penetration ratio are hypothesized to be able to obtain protections more easily than those that have low import penetration. Bown, Hoekman and Ozden (2003) showed that developing countries were named more frequently than developed countries in antidumping petitions in the U.S. However, as shown in Table 3.2, developing countries are not named more often than the advanced economies in Canadian antidumping cases. The economic natures of the named countries are therefore irrelevant in Canada. A possible reason is that Canada trades mostly with the U.S. and European countries, and less so with developing countries. There is a possibility of endogeneity issue that needs to be addressed before moving on to our main analysis. If increased lobbying can generate more affirmative determinations in antidumping cases, then it is possible that increased affirmative determinations encourage more lobbying in the future. 143  Import values and volumes are reported in HS code, these are recoded into NAICS6 using the Statistics Canada concordance. 144 Domestic consumption equals the sum of domestic output and imports less exports.  203  Chapter 3. Does Lobbying Affect Antidumping Case Determinations in Canada? We examine this possibility as follows. Assume that the number of lobbyists hired to represent an industry can be loosely taken as the amount of lobbying effort exerted by an industry and that the lobbying expenditure by an industry is positively correlated to the number of lobbyists representing the industry. Thus, an increased number of lobbyists hired would indicate a larger amount of effort. We show in the following analysis that lobbying efforts are influenced by industry-specific characteristics by regressing the number of lobbyists hired by each industry on three types of characteristics. We also include two variables to account for the influence of past affirmative determination in antidumping cases, a variable that represents the cumulative number of affirmative determinations and a dummy that indicates there was ever an affirmative determination. There are three types of industry-specific characteristics: political factors, market-related factors and production-related factors. The political factors include variables that represent whether there is counter-lobbying in the industry, whether there is previous experience in lobbying and whether the industry had contributed politically before. They are included to examine how these political involvements affect the industry lobbying efforts. The market-related factors are the import elasticities the industry faces, the import penetration ratio the industry experiences, the amount the industry sells to the government, the amount of subsidy the industry receives from government and the number of complaints the industry files with government agencies. They are included to examine how the market environment affects the amount of lobbying efforts. Finally, the production-related factors are the industry concentration, the amount of industry output and value-added, the number of employees and establishments in the industry, the amount of capital investment of the industry, the capital-labour ratio, the cost of energy and R&D expenses and the industry’s export propensity. These are included to examine the effects of industry production characteristics on lobbying efforts.  204  Chapter 3. Does Lobbying Affect Antidumping Case Determinations in Canada? Note that there are some industries that have never lobbied, which yields some zeros in the data for the number of lobbyists. To deal with overdispersion caused by the excessive zeros, unobserved heterogeneity and serial dependence, a random effects negative binomial regression is used to reduce the possibility of biased estimates. The results of the regression of lobbying efforts are presented in Table 3.5.  205  Chapter 3. Does Lobbying Affect Antidumping Case Determinations in Canada?  Table 3.5: Regression of Lobbying Efforts 1  2  3  4  Cumulate affirmative  0.025 (0.005)  1.680 (0.525)  0.025 (0.005)  1.636 (0.506)  Ever affirmative  −0.059 (0.080)  −1.525 (0.543)  0.099 (0.072)  −1.830 (0.526)  Lagged counter-lobbying  0.073 (0.052)  0.070 (0.069)  Experience in lobbying  0.036 (0.006)  0.035 (0.013)  Lagged contribution  0.056 (0.016)  0.055 (0.022)  Independent Variablesf  Models  Import elasticities  0.502 (1.224)  1.603 (1.019)  Import penetration  0.356† (0.190)  0.511 (0.331)  Sales to government  0.215 (0.049)  0.116 (0.054)  Government subsidy  0.003 (0.026)  0.020 (0.027)  Complaints  0.012 (0.010)  0.007 (0.009)  Industry concentration  −0.006 (0.002)  0.001 (0.003)  Output  0.466† (0.250)  0.404 (0.284)  Employees  0.002 (0.005)  −0.009 (0.008)  Establishments  0.051 (0.041)  −0.073 (0.070)  Capital investment  0.034 (0.160)  0.005 (0.199)  Capital labour ratio  −0.001 (0.004)  −0.004 (0.006)  Cost of energy  0.001 (0.000)  −0.000 (0.001)  Value-added  −0.090 (0.102)  0.006 (0.117)  Export propensity  −0.075 (0.120)  −0.004 (0.342)  R&D expenses  0.051 (0.042)  0.061 (0.060)  Concentration*export propensity  0.004 (0.002)  −0.004 (0.005)  Log-likelihood ratio Number of observations f  Standard errors are in parentheses, ,  −2452.628  −1461.565  −3242.040  −1091.282  n=1310  n=761  n=2055  n=500  and † represent the significance levels of 1%, 5% and 10%.  206  Chapter 3. Does Lobbying Affect Antidumping Case Determinations in Canada? The first three specifications are regressions of the number of lobbyists on each of the three sets of characteristics plus the two variables accounting for past affirmative determinations. The results show that the cumulative number of affirmative determinations is always significant and positive, while the dummy that indicates if there was ever an affirmative determination is not. Note that in each specification, a few additional variables are significant, such as lobbying experience and lagged contributions in the political factors; import penetration ratio and sales to government in the marketrelated factors; and industry concentration ratio, industry output and cost of energy in the production-related factors. The final specification includes all three sets of factors and an interaction term of industry concentration and export propensity. The interaction term is used to capture the resultant effect of the opposing effects of industry concentration and export propensity. The two variables that represent influences of past affirmative determinations are significant but with opposite effects. The cumulative number of affirmative determinations increases the lobbying effort. However, if there was ever an affirmative determination, the industry will lower the number of lobbyists they hire. It will be difficult to say whether the opposing effects will net out. But as both factors are significant, endogeneity between affirmative determination and lobbying seems to exist. We will take this into account in the empirical work below. Almost all other factors, except for the political factors, become insignificant when regressed together. Both the lobbying experience and political contributions increase lobbying effort. Thus, industries that are experienced in lobbying and have contributed will exert more effort. The existence of counter-lobbyists does not seem to have significant effect on lobbying effort. The only other significant factor is the sales to government. Overall, the above results show that in Canada, the amount of lobbying efforts exerted by the industries is highly motivated by the previous success in obtaining determinations and what previous political involvements they had. Lastly, the overdispersion test results show that the choice of negative binomial 207  Chapter 3. Does Lobbying Affect Antidumping Case Determinations in Canada? regression is appropriate in all of the specifications. Next we describe the econometric model used for analysing the historical cases.  3.3.2  Econometric Analysis  A negative binomial model that can utilise the count data nature is used to study the effect of the determinants on the number of occurrences of an event (affirmative determinations). The model assumes the mean and the variance are unequal (as opposed to equal mean and variance in Poisson regression) which can be used to overcome the problem of overdispersion that is common in count data. Moreover, data in trade and political economics typically have large variations and excess zeroes.145 The structure of a negative binomial is in the form of  yi ∼ Poisson(μi ∗)  (3.2)  μi ∗ = exp(xi β + ui ) 1 1 , eui ∼ gamma α α  (3.3) (3.4)  where ui represents an omitted variable such that eui follows a gamma distribution, α is the overdispersion parameter. Another advantage of using a negative binomial model is that it allows for unobserved heterogeneity between events through ui . A likelihood ratio test will be performed to justify the use of negative binomial regression. If the null hypothesis of non-unity dispersion factor cannot be rejected, then the model reduces to a simple Poisson. 145  Examples of excess zeroes in this analysis would be there are some industries that have never filed for antidumping, which creates a skew in the distribution. The mean and variance would undoubtedly be unequal.  208  Chapter 3. Does Lobbying Affect Antidumping Case Determinations in Canada? The data for all cases initiated from 1996 to 2003 are pooled for analysis and all variable values are deflated to 1997 constant dollars using industrial product price index. There are a total of 127 affirmative determinations out of 163 cases among 258 manufacturing industries. The countervailing cases contributed to less than 1% of all cases, and thus will be excluded from the analysis. Although lobbying is often done at firm level, most antidumping cases are raised at industry level. Analyses are therefore performed at industry level. Furthermore, the lack of detailed data for characteristics at firm level made it difficult to carry out analyses at firm level. Table 3.6 shows the averages of the determinants and their expected signs as regression coefficients in the negative binomial analysis. Table 3.6: Expected Signs of Determinants Determinant Lobbyist  g  Averageg  Expected Sign  1.3  +  Political contributions (dollars)  9317  +  Industry concentration ratio  52.6  +  Profit (bil. dollars)  -1.4  −  Employment growth  -3.1%  −  Capacity utilisation rate  82.9  −  Capital investments (bil. dollars)  12.1  +  Growth in average earnings  1.9%  −  Output growth  4.7%  −  Import price change  6.6%  −  Import penetration ratio  0.5  +  Export intensity  47%  −  Several industries do not have information on import prices.  We tried several specifications of the model and the estimation results are shown in Tables 3.7 and 3.8. The first three specifications each regresses on a different set of variables that represents the political variables, the CITT economic injury determinants and the additional economic determinants. 209  Chapter 3. Does Lobbying Affect Antidumping Case Determinations in Canada? The fourth specification regresses on all determinants with an addition variable, the interaction between industry concentration and export intensity. In order to account for possible endogeneity between lobbying and affirmative determinations, we include an independent variable that is equal to the lagged percentage of affirmative cases in terms of all successful antidumping petitions in the fifth to eighth specifications which are repeats of the first four specifications.  210  Chapter 3. Does Lobbying Affect Antidumping Case Determinations in Canada?  Table 3.7: Negative Binomial Regression on Antidumping Case Determinations Independent Variables h  Models  1  2  3  4  Lobbyist  0.050 (0.018)  0.053 (0.019)  Industry concentration ratio  0.001 (0.007 )  −0.003 (0.012)  Lagged contributions  0.249 (0.116)  0.247 (0.118)  Profit (bil. dollars)  0.093 (0.155)  0.003 (0.023)  Employment growth  0.018 (0.012)  0.028† (0.015)  Capacity utilisation rate  0.032 (0.033)  0.031 (0.038)  Capital investments  0.001 (0.010)  −0.006 (0.010)  Growth in average earnings  −0.014 (0.020)  −0.026 (0.019)  Output growth  −0.002 (0.011)  0.008 (0.015)  Import price change  −0.002 (0.005)  −0.002 (0.005)  Lagged import penetration  0.003 (0.031)  0.004 (0.033)  −0.009† (0.005)  −0.010 (0.015)  Export intensity Concentration*Intensity Year dummy Log-likelihood Number of observations h  Standard errors are in parentheses, ,  0.000 (0.001) No  No  No  Yes  −147.534  −186.800  −179.736  −136.456  n=1465  n=2048  n=1857  n=1307  and † represent the significance levels of 1%, 5% and 10%.  211  Chapter 3. Does Lobbying Affect Antidumping Case Determinations in Canada?  Table 3.8: Negative Binomial Regression on Antidumping Case Determinations (Cont’d) 5  6  7  8  % of affirmative cases  0.855 (0.904 )  1.908 (0.667)  1.747 (0.667)  0.675 (0.941)  Lobbyist  0.047 (0.019)  0.048 (0.019)  Industry concentration ratio  0.008 (0.010 )  0.002 (0.013)  Lagged contributions  0.255 (0.126)  0.274 (0.124)  Independent Variables i  Models  Profit (bil. dollars)  0.084 (0.155)  0.004 (0.024)  Employment growth  0.020 (0.013)  0.028 (0.016)  Capacity utilisation rate  0.060† (0.034)  0.057 (0.040)  Capital investments  0.001 (0.010)  −0.001 (0.010)  Growth in average earnings  −0.017 (0.019)  -0.026 (0.020)  Output growth  −0.005 (0.012)  0.003 (0.015)  Lagged import penetration  0.005 (0.030)  0.005 (0.031)  Export intensity  −0.010 (0.006)  −0.008 (0.016)  Import price change  −0.000 (0.000)  Concentration*Intensity Year dummy Log-likelihood Number of observations i  Standard errors are in parentheses, ,  No  No  No  Yes  −127.527  −162.330  −163.517  −121.992  n=1274  n=1794  n=1750  n=1233  and † represent the significance levels of 1%, 5% and 10%.  212  Chapter 3. Does Lobbying Affect Antidumping Case Determinations in Canada? The result from the first specification shows that the number of affirmative determinations increases when the industries that filed the petition also hired lobbyists to lobby CITT and when they had contributed monetarily before. Thus, political involvement is positively associated with administered protection in Canada. Both lobbyists and lagged contributions are significant though the industry concentration does not seem to influence administered protection. The next two specifications show some surprising results. Of all the economic determinants, only export intensity is found to be significant and of expected sign. This is surprising because the economical determinants are the only determinants that are supposed to be used in determinations. The results indicate that being economically independent (as suggested in Finger et al. (1982)) can lower the number of affirmative determinations. The fourth specification includes all determinants as regressors. Both lobbyists and lagged contributions continue to be significant. Employment growth is the only economic determinant that turns out to be significant except that the sign is not as expected. Thus, if we ignore the endogeneity between affirmative determinations and lobbying, increased amount of political involvements of industries increases the number of affirmative determinations. The insignificance of economic determinants is quite unanticipated, especially the ones used by CITT. However, this is similar to what Hansen and Prusa (1997) found: that is, increasing economic injury may not improve the chance of protection. The growth in employment is expected to have a negative effect on determination because if dumping injures an industry, one would expect a decline in employment to follow. The estimation result shows otherwise. A possible explanation is that growth in employment represents increase in the size of an industry that needs protection. If that is the case, the results of the positive relation can be justified and can be interpreted as the size effect overpowering the economic effect. Widely accepted dumping indicators like import price change and lagged import penetration are not significant even though they are of expected 213  Chapter 3. Does Lobbying Affect Antidumping Case Determinations in Canada? signs. There is a possibility that on average, import prices did not fluctuate a lot in the sample period even though import prices for some commodities such as machinery had decreased quite significantly. Similarly, the average import penetration ratio is only 0.5, showing that the Canadian economy overall does not face a high import penetration. The lack of significance in economic determinants also shows that antidumping cases may not be determined based on the proof of unfair trade practices by the foreign exporters (such as lower import prices), but rather on other determinants. The year dummies do not seem to have any effect, which can be easily explained by the fact that Canadian economy remained rather steady in the sample period. Therefore, the economic setbacks in 1997 and 2001 were not sufficient to affect antidumping case determinations in Canada. Overall, when ignoring the endogeneity issue, the empirical results indicate that antidumping case determinations may have been influenced by factors other than those stated in the CITT mandate. The next four specifications take into account the endogeneity between affirmative determinations and lobbying (note that in the seventh and eighth specifications, import price growth is removed to maintain the concavity assumption). We find that, when the number of affirmative determinations is regressed on the lagged percentage of affirmative determinations and the political variables, only the number of lobbyists and lagged contributions are significant. However, when the regression is carried out on economic determinants, the lagged percentage of affirmative determinations became significant. This shows that if cases are determined only on economic determinants, the affirmative rate of past antidumping cases can influence the number of future affirmative determinations. In the eighth specification, when all regressors are included, the lagged percentage of affirmative determinations returns to being insignificant. Both the number of lobbyists and lagged contributions are significant while other economic determinants are not. Thus, the result here does not support there 214  Chapter 3. Does Lobbying Affect Antidumping Case Determinations in Canada? is an endogeneity between lobbying and determinations. Political involvement is much more important in terms of affecting determinations. The results from the last four specifications are intriguing because they indicate that if the mandate is truly apolitical, the determinations of antidumping cases is influenced by the rate of affirmative determination in the past. Yet when firms are lobbying to the CITT, the influence is switched to the political strategies. Summarising the results of all specifications, we have evidence that political factors have influences on the number of affirmative case determinations in Canada. One question that remains is whether lobbying leads to industries obtaining affirmative determinations or whether it also leads to more antidumping petitions. We tried the same analysis on the number of successful antidumping petitions. A successful petition is one that had been filed, investigated and determined but not necessarily affirmative. The results are shown in Table 3.9.  215  Chapter 3. Does Lobbying Affect Antidumping Case Determinations in Canada?  Table 3.9: Negative Binomial Regression on Antidumping Petitions Independent Variables j  Models  1  2  3  4  Lobbyist  0.064† (0.037)  0.072 (0.039)  Industry concentration ratio  0.025 (0.009 )  0.009 (0.021)  Lagged contributions  0.419 (0.136)  0.361 (0.152)  Profit (bil. dollars)  0.009 (0.072)  0.005 (0.029)  Employment growth  0.001 (0.000)  0.077 (0.030)  Capacity utilisation rate  0.133 (0.043)  0.042 (0.049)  Capital investments  0.001 (0.010)  0.001 (0.010)  Growth in average earnings  −0.026 (0.033)  −0.075† (0.042)  Output growth  −0.043† (0.023)  0.029 (0.030)  Import price change  −0.035 (0.025)  −0.018 (0.015)  Lagged import penetration  0.354 (0.618)  0.006 (0.075)  Export intensity  −0.022 (0.014)  0.001 (0.022) −0.000 (0.000)  Concentration*Intensity Year dummy Log-likelihood Number of observations j  Standard errors are in parentheses, ,  No  No  No  Yes  −203.244  −256.832  −247.607  −185.678  n=1465  n=2048  n=1857  n=1307  and † represent the significance levels of 1%, 5% and 10%.  216  Chapter 3. Does Lobbying Affect Antidumping Case Determinations in Canada? Comparing the results in Table 3.7 and Table 3.9, it can be seen that the estimates are quite close. The determinants that are significant in Table 3.7 are also significant in Table 3.9. There are a few exceptions such as industry concentration ratio is significant in the first specification, the growth in output is significant in the third specification, and the growth in average earnings is significant and of expected sign in the fourth specification. There are two possible explanations to why the results are similar. First, it is possibly an indication that the number of successful petitions and the number of affirmative determinations are driven by similar determinants. Thus if political involvements like lobbying promotes affirmative determinations, it also promotes the successful petitions. Second, the affirmative determination rate is moderately high relative to the number of successful petitions. In fact, the percentage of affirmative determinations relative to successful petitions is almost 70% over the sample period. Nonetheless, both analyses provide evidence that in Canada, antidumping cases are not independent of political influences.  3.4  Concluding Remarks  Using the Canadian Lobbyists Registration data and the historical antidumping cases in Canada, this paper attempts to determine to what extent the economic and political factors are affecting administered protection. In the above empirical analyses, the number of CITT lobbyists is always significant and positive. Economic determinants on the other hand do not affect determinations as much. Similar studies on U.S. antidumping case determinations found rather different results. Economic determinants are often found equally important as political determinants, if not more so. The difference is most probably driven by the fact that Canada and the U.S. are very different in terms of their economic structures and antidumping legislations. First, Canada is very export dependent compared to the U.S. This means Canadian industries may be more economically independent in the Finger et al. sense and thus may endure lower amount of injuries in case of dumping. 217  Chapter 3. Does Lobbying Affect Antidumping Case Determinations in Canada? Second, Canada’s antidumping legislation is very straightforward, using a prospective system that publishes “undumped” prices that are sufficient to eliminate the need to pay dumping duties. Therefore, importers can simply follow these guidelines in setting prices. By contrast, the U.S. antidumping legislation is very complicated. It applies a retrospective system that collects an ex-ante dumping duties. Adjustments in duties such as collecting more from the importers or refunding them only occur in regular reviews after determination. Some of the U.S. practices may lead to importers being fined more (the zeroing practice in calculation of margin) and industries filing more antidumping petitions than needed (motivated by incentive such as the Bryd Amendment). The U.S. Department of Commerce also uses many different rules in its reviews than the original investigation, many of which are rules that could result easier in affirmative determinations. Furthermore, importers are allowed to lower their home prices instead of raising the U.S. import prices to avoid dumping duties which does not relieve any dumping injuries that domestic industries experience. The complexity in the U.S. antidumping legislation made it possible for industries to manipulate their levels of injury for affirmative determinations. In other words, Canada’s simple antidumping legislation made manipulation of injuries almost impossible. These two important differences combined increases the importance of using political strategies to influence determination. Another finding from the empirical analyses is that the determinants that are significant in influencing affirmative determinations are also significant in influencing successful petitions. This suggests political strategies such as lobbying the CITT affect both stages of antidumping cases. Lack of appropriate data has made it difficult to empirically analyse the relation of political strategies with policy outcomes. This paper, like many of its predecessors, reflects the same difficulty. Although the Lobbyists Registration data has provided a suitable set of detailed data to construct a political variable, an ideal political variable should be constructed using lobbying expenditure. Fortunately, one can still isolate the lobbyists who are interested in influencing particular outcomes in order to conduct more 218  Chapter 3. Does Lobbying Affect Antidumping Case Determinations in Canada? focussed studies. In this case, only lobbyists who lobbied CITT were selected to study whether the antidumping case determinations are independent of influences other than those allowed as stated in the mandate. Nonetheless, use of the Lobbyists Registration data allows the study of political strategies to go beyond the traditional use of industry concentration or political contributions. For the economic determinants, there are also many situations where less detailed measures must be used. The measure of capital investments is one of the determinants CITT is allowed to consider regarding injury and has been shown to be an influential determinant in previous studies. However, capital investment data by industry are often not released due to information protection. In this case, this variable has to be constructed by weighing the aggregate level of capital investments using industry output, which greatly reduces its accuracy. Other variables such as the industry concentration ratio and the capacity utilisation rate are also only available at NAICS 3 digit level instead of the NAICS 6 digit level needed for the analysis. Furthermore, much of the data transformation requires applications of concordance or manual matching, either of which could translate into sources of error. Another weakness of our analysis is that only a short time series of data is available. The antidumping cases that were initiated during this period amount to only a small portion of the total number of antidumping cases in Canada. If data were available for a longer time series, the estimates would be more reliable. Further research may need to consider problems such as identifying individual firms’ intents to lobby CITT (whether it is for or against protection) and whether the firms that lobby are also the petitioners of the cases. Other considerations may include incorporating the institutional structure of the agencies responsible for antidumping investigation and injury determination, and developing ways to quantify the benefits and costs associated with antidumping cases.  219  Bibliography [1] Aggarwal, Aradhna (2004) ‘Macro Economic Determinants of Antidumping: A Comparative Analysis of Developed and Developing Countries.’ World Development 32(6), 1043–1057 [2] Alterman, W., W.E. Diewert, and R.C. Feenstra (1999) ‘International Trade Price Indexes and Seasonal Commodities.’ Technical Report, Washington, US Department of Labor, Bureau of Labor Statistics [3] Andere, Eduardo (1992) ‘The Political Economy of Protectionism: Antidumping in the Mexican-U.S. Trade Relationship.’ PhD dissertation, Boston College [4] Anderson, James E. (1994) ‘Strategic Lobbying and Antidumping.’ Working Papers in Economics 269, Boston College [5] Anderson, Simon P., Nicholas Schmitt, and Jacques-Francois Thisse (1995) ‘Who Benefits from Antidumping Legislation?’ Journal of International Economics 3-4, 321–337 [6] Archibald, R.B. (1977) ‘On the Theory of Industrial Price Measurement: Output Price Indexes.’ Annals of Economic and Social Measurement 6, 57–72 [7] Baldwin, J.R., and W. Gu (2007) ‘Multifactor Productivity in Canada: An Evaluation of Alternative Methods of Estimating Capital Services.’ Research Paper, The Canadian Productivity Review Catalogue No. 15206-XIE-No. 009, Ottawa: Statistics Canada  220  Bibliography [8]  (2009) ‘Productivity Performance in Canada 1961-2008 An Update on Long Term Trend.’ Research Paper, The Canadian Productivity Review Catalogue No. 15-206-X-No. 205, Ottawa: Statistics Canada  [9] Baldwin, J.R., W. Gu, and B. Yan (2007) ‘User Guide for Statistics Canada’s Annual Multifactor Productivity Program.’ Research Paper, The Canadian Productivity Review Catalogue No. 15-206-XIE-No. 014, Ottawa: Statistics Canada [10] Baldwin, R.E. (1989) ‘The Political Economy of Trade Policy.’ Journal of Economic Perspectives 3(4), 119–135 [11] Baldwin, R.E., and J.W. Steagall (1994) ‘An Analysis of ITC Decisions in Antidumping, Countervailing Duty and Safeguard Cases.’ Review of World Economics (Weltwirtschaftliches Archiv) 130(2), 290308 [12] Balk, B.M. (1998) Industrial Price, Quantity and Productivity Indices (Boston: Kluwer Academic Publishers) [13] Blonigen, Bruce A. (2006) ‘Working the System: Firm Learning and the Antidumping Process.’ European Journal of Political Economy 22(3), 715–731 [14] Blonigen, Bruce, and Thomas Prusa (2003) ‘The Cost of Antidumping: the Devil is in the Details.’ The Journal of Policy Reform 6(4), 233–245 [15] Bown, Chad P. (2007) ‘Canada’s Antidumping and Safeguard Policies: Overt and Subtle Forms of Discrimination.’ The World Economy 30(9), 1457–1476 [16] Bown, Chad P., Bernard Hoekman, and Caglar Ozden (2003) ‘The Pattern of U.S. Antidumping: The Path From Initial Filing to WTO Dispute Settlement.’ World Trade Review 2(3), 349–371 [17] Christensen, L.R., and D.W. Jorgenson (1969) ‘The Measurement of U.S. Real Capital Input, 1929-1967.’ Review of Income and Wealth 15, 293–320 221  Bibliography [18] Christensen, L.R., D.W. Jorgenson, and L.J. Lau (1971) ‘Conjugate Duality and the Transcendental Logarithmic Production Function.’ Econometrica 39, 255–256 [19] Coughlin, Cletus C., Joseph V. Terza, and Noor Aini Khalifah (1989) ‘The Determinants of Escape Clause Petitions.’ The Review of Economics and Statistics 71(2), 341–347 [20] Denison, Edward F. (1974) ‘Accounting for United States Economic Growth 1929-1969.’ Technical Report, Washington, D.C.: The Brookings Institution [21] Diewert, W.E. (1973) ‘Functional Forms for Profit and Transformation Functions.’ Journal of Economic Theory 6, 284–316 [22]  (1974) ‘Application of Duality Theory.’ vol. II (Amsterdam: North Holland) pp. 106–171  [23]  (1977) ‘Walras’ Theory of Capital Formation and the Existence of a Temporary Equilibrium.’ (Reidel Publishing Company) pp. 73–126  [24]  (1978) ‘Superlative Index Numbers and Consistency in Aggregation.’ Econometrica 46, 883–900  [25]  (1980) ‘Aggregate Problems in the Measurement of Capital.’ (Chicago: The University of Chicago Press) pp. 433–528  [26]  (1993) ‘Symmetric Means and Choice Under Uncertainty.’ vol. Contributions to Economic Analysis 217 (Amsterdam: North-Holland) pp. 355–433  [27]  (1997) ‘Commentary on Mathew D Shapiro and David W. Wilcox, “Alternative Strategies for Aggregating Price in the CPI”.’ The Federal Reserve Bank of St. Louis Review 79:3, 127–137  [28]  (2001) ‘Which (Old) Ideas on Productivity Measurement are Ready to Use?’  vol. 63 of NBER Studies in Income and Wealth  (Chicago: University of Chicago Press) pp. 85–101 222  Bibliography [29]  (2002) ‘Productivity Trends and Determinants in Canada.’ (Calgary: University of Calgary Press) pp. 31–57 (2005a) ‘Issues in the Measurement of Capital Services, Deprecia-  [30]  tion, Asset Price Changes and Interest Rates.’ vol. 65 of NBER Studies in Income and Wealth (Chicago: University of Chicago Press) pp. 479– 542 [31]  (2005b) ‘On Measuring Inventory Change in Current and Constant Dollar.’ Discussion Paper 05-12, Department of Economics, The University of British Columbia, Vancouver, Canada  [32]  (2005c) ‘Welfare, Productivity and Changes in the Terms of Trade.’ Presented at the Bureau of Economic Analysis, Washington D.C., November 17  [33]  (2006a) ‘Comment on Aggregation Issues in Integrating and Accelerating the BEA’s Accounts: Improved Methods for Calculating GDP by Industry.’ vol. 66 of NBER Studies in Income and Wealth (Chicago: University of Chicago Press) pp. 287–307  [34]  (2006b) ‘The Measurement of Income.’ Chapter 7 as Tutorial Presented at the University Autonoma of Barcelona, Spain, September 21-22,2005  [35]  (2007a) ‘Price Indices Using an Artificial Data Set.’ Draft of Chapter 19 of the Export Import Price Manual, IMF, Washington D.C.  [36]  (2007b) ‘The Economic Approach.’ Draft of Chapter 17 of the Export Import Price Manual, IMF, Washington D.C.  [37]  (2007c) ‘The Paris OECD-IMF Workshop on Real Estate Price Indexes: Conclusions and Future Directions.’ Discussion Paper 07-01, Department of Economics, The University of British Columbia, Vancouver, Canada  223  Bibliography [38]  (2008) ‘Changes in the Terms of Trade and Canada’s Productivity Performance.’ Discussion Paper 08-05, Department of Economics, The University of British Columbia, Vancouver, Canada  [39] Diewert, W.E., and A.M. Smith (1994) ‘Productivity Measurement for a Distribution Firm.’ The Journal of Productivity Analysis 5, 335–347 [40] Diewert, W.E., and D. Lawrence (2000) ‘Progress in Measuring the Price and Quantity of Capital.’ Essays in Honor of Dale W. Jorgenson (Cambridge: The MIT Press) pp. 273–326 [41]  (2002) ‘The Deadweight Costs of Capital Taxation in Australia.’ (Boston: Kluwer Academic Publishers) pp. 103–167  [42]  (2005) ‘Estimating Aggregate Productivity Growth for Australia: The Role of Information and Communications Technology.’ Occasional Economic Paper, Canberra: Department of Communications, Information Technology and the Arts, Australian Government  [43]  (2006) ‘Measuring the Contributions of Productivity and Terms of Trade to Australia’s Economic Welfare.’ Report by Meyrick and Associates to the Productivity Commission, Canberra, Australia  [44] Diewert, W.E., and K.J. Fox (2001) ‘The Productivity Paradox and Mismeasurement of Economic Activity.’ (London: MacMillan Press) pp. 175–197 [45] Diewert, W.E., H. Mizobuchi, and K. Nomura (2005) ‘On Measuring Japan’s Productivity, 1995-2003.’ Discussion Paper 05-22, Department of Economics, The University of British Columbia, Vancouver, Canada [46] Drope, Jeffrey M., and Wendy L. Hansen (2004) ‘Purchasing Protection? The Effect of Political Spending on U.S. Trade Policy.’ Political Research Quarterly 57(1), 27–37 [47] Edwards, E.O., and P.W. Bell (1961) The Theory and Measurement of Business Income (Berkeley, CA: University of California Press) 224  Bibliography [48] Eurostat, International Monetary Fund, Organisation for Economic Cooperation, United Nations Development, and World Bank (1993) ‘System of National Accounts 1993 .’ Technical Report, Luxembourg, New York, Paris, Washington D.C. [49] Evans, Carolyn L., and Shane M. Sherlund (2006) ‘Are Antidumping For Sale?’ International Finance Discussion Paper 888, Board of Governors of the Federal Reserve System [50] Feenstra, R.C. (2004) Advanced International Trade: Theory and Evidence (Princeton, NJ: Princeton University Press) [51] Feinberg, Robert M. (1989) ‘Exchange Rates and Unfair Trade.’ The Review of Economics and Statistics 71(4), 704–707 [52]  (2005) ‘U.S. Antidumping Enforcement and Macroeconomic Indicators Revisited: Do Petitioners Learn?’ Review of World Economics (Weltwirtschaftliches Archiv) 141(4), 612–622  [53] Feinberg, Robert M., and Barry T. Hirsch (1989) ‘Industry Rent Seeking and the Filing of Unfair Trade Complaints.’ International Journal of Industrial Organization 7(4), 325–340 [54] Fisher, F.M., and K. Shell (1972) The Pure Theory of the National Output Deflator [55] Fisher, I. (1922) The Making of Index Numbers (Boston: HoughtonMifflin) [56] Francois, Joseph, and Gunnar Niels (2004) ‘Political Influence in a New Antidumping Regime: Evidence from Mexico.’ Discussion Papers 4297, CEPR [57] Gasmi, Farid, Eric Malin, and Fran¸cois Tand´e (2004) ‘Lobbying in An´ tidumping.’ IDEI Working Papers 320, Institut d’Economie Industrielle (IDEI)  225  Bibliography [58] Gorman, W.M. (1968) ‘Measuring the Quantities of Fixed Factors.’ (Chicago: Aldine Press) pp. 141–172 [59] Hansen, Wendy L., and Kee Ok Park (1995) ‘Nation-state and Pluralistic Decision Making in Trade Policy: the Case of the International Trade Administration.’ International Studies Quarterly 39, 181211 [60] Hansen, Wendy L., and Thomas J. Prusa (1997) ‘The Economics and Politics of Trade Policy: An Empirical Analysis of ITC Decision Making.’ Review of International Economics 5(2), 230–245 [61] Herander, M.G., and J.B. Schwartz (1984) ‘An Empirical Test of the Impact of the Threat of US Trade Policy: The Case of Antidumping Duties.’ Southern Economic Journal 51(1), 59–79 [62] Hicks, J.R. (1946) Value and Capital, second edition ed. (Oxford: The Clarendon Press) (1961) ‘The measurement of Capital in Relation to the Measure-  [63]  ment of Other Economic Aggregates.’ (London: Macmillan) pp. 18–31 [64] Hill, R.J. (2006) ‘Superlative Indexes: Not All of Them are Super.’ Journal of Econometrics 130, 25–43 [65] Hotelling, H. (1931) ‘Edgeworth’s Taxation Paradox and the Nature of Demand and Supply Functions.’ Journal of Political Economy 40, 577– 616 [66] Irwin, Douglas A. ‘The Rise of U.S. Antidumping Activity in Historical Perspective.’ Technical Report, International Monetary Fund [67] Jorgenson, D.W. (1989) ‘Capital as a Factor of Production.’ (Cambridge, MA: The MIT Press) pp. 1–35 [68]  (1996a) ‘Empirical Studies of Depreciation.’ Economic Inquiry 34, 24–42  [69]  (1996b) Investment: Volume 2; Tax Policy and the Cost of Capital (Cambridge, MA: The MIT Press) 226  Bibliography [70] Jorgenson, D.W., and Z. Griliches (1967) ‘The Explanation of Productivity Change.’ The Review of Economic Studies 34, 249–283 (1972) ‘Issues in Growth Accounting: A Reply to Edward F. Deni-  [71]  son.’ Survey of Current Business 52:4, Part II, 65–94 [72] Knetter, Michael M., and Thomas J. Prusa (2003) ‘Macroeconomic Factors and Antidumping Filings: Evidence from Four Countries.’ Journal of International Economics 61, 1–18 [73] Kohli, U. (1978) ‘A Gross National Product Function and the Derived Demand for Imports and Supply of Exports.’ Canadian Journal of Economics 11, 167–182 (1990) ‘Growth Accounting in the Open Economy: Parametric  [74]  and Nonparametric Estimates.’ Journal of Economic and Social Measurement 16, 125–136 (1991) Technology, Duality and Foreign Trade: The GNP Function  [75]  Approach to Modeling Imports and Exports (Ann Arbor: MI: University of Michigan Press) (2003) ‘Growth Accounting in the Open Economy: International  [76]  Comparisons.’ International Review of Economics and Finance 12, 417– 435 [77]  (2004a) ‘An Implicit T¨ornqvist Index of Real GDP.’ Journal of Productivity Analysis 21, 337–353  [78]  (2004b) ‘Real GDP, Real Domestic Income and Terms of Trade Changes.’ Journal of International Economics 62, 83–106  [79] Kon¨ us, A.A. (1939) ‘The Problem of the True Index of the Cost of Living.’ Econometrica 7, 10–29. Translated from 1924 original [80] Lau, L. (1976) ‘A Characterization of the Normalized Restricted Profit Function.’ Journal of Economic Theory 12:1, 131–163  227  Bibliography [81] Leacy, F.H., ed. (1983) Historical Statistics of Canada, second edition ed. (Ottawa: Canada: Statistics Canada) [82] Leidy, M.P. (1997) ‘Macroeconomic Conditions and Pressures for Protection under Antidumping and Countervailing Duty Laws: Empirical Evidence from the United States.’ Staff Papers 44, International Monetary Fund. p. 132144 [83] Leipziger, Danny M., and Hyun Ja Shin (1991) ‘The Demand for Protection: A Look at Antidumping Cases.’ Open Economies Review 2, 27– 38 [84] Mah, Jai S. (2000) ‘Antidumping Decisions and Macroeconomic Variables in the USA.’ Applied Economics 32, 1701–1709 [85] McFadden, D. (1978) ‘Cost, Revenue and Profit Functions.’ vol. 1 (Amsterdam: North-Holland) pp. 3–109 [86] Moore, Michael O. (1992) ‘Rules or Politics?: An Empirical Analysis of ITC Anti-dumping Decisions.’ Economic Enquiry 30(3), 449–466 [87] Morrison, C.J., and W.E. Diewert (1990) ‘Productivity Growth and Changes in the Terms of Trade in Japan and the United States.’ (Chicago: University of Chicago Press) pp. 201–227 [88] Moyer, B.C., M.B. Reinsdorf, and R.E. Yuskavage ‘Aggregation Issues in Integrating and Accelerating the BEA’s Accounts: Improved Methods for Calculating GDP by Industry.’ vol. 66 of NBER Studies in Income and Wealth (Chicago: University of Chicago Press) [89] Nelson, Douglas ‘The Political Economy of Antidumping: A Survey.’ European Journal of Political Economy [90] Niels, Gunnar (2000) ‘What is Antidumping Policy Really About?’ Journal of Economic Surveys 14(4), 467–492 [91] OECD (1972) ‘Labour Force Statistics.’ Technical Report, Paris: OECD 228  Bibliography [92]  (1993) ‘Labour Force Statistics.’ Technical Report, Paris: OECD  [93] Prusa, Thomas J. (2001) ‘On the Spread and Impact of Anti-dumping.’ Canadian Journal of Economics 34(3), 591–611 [94] Prusa, Thomas J., and Susan Skeath (2001) ‘The Economic and Strategic Motives for Antidumping Filings.’ Working Paper 8424, NBER [95] Rymes, T.K. (1968) ‘Professor Read and the Measurement of Total Factor Productivity.’ Canadian Journal of Economics 1, 359–367 [96]  (1983) ‘More on the Measurement of Total Factor Productivity.’ The Review of Income and Wealth 29, 297–316  [97] Samuelson, P.A. (1953) ‘Prices of Factors and Goods in General Equilibrium.’ Review of Economic Studies 21, 1–20 [98] Samuelson, P.A., and S Swamy (1974) ‘Invariant Economic Index Numbers and Canonical Duality: Survey and Synthesis.’ American Economic Review 64, 566–593 [99] Sato, K. (1976) ‘The Meaning and Measurement of the Real Value Added Index.’ Review of Economics and Statistics 58, 434–442 [100] Schreyer, P. (2001) OECD Productivity Manual: A Guide to the Measurement of Industry-Level and Aggregate Productivity Growth OECD (Paris) [101]  (2007) Measuring Capital: Revised Manual OECD Statistics Directorate (Paris:OECD). Paper presented at the Meeting of the Working Party on National Accounts  [102] Secretariat, WTO (2007) ‘Semi-Annual Report of Antidumping Actions.’ Technical Report, World Trade Organization [103] Solow, R.M. (1957) ‘Technical Change and the Aggregate Production Function.’ Review of Economics and Statistics 39, 312–320  229  Bibliography [104] Statistics Canada (1987) The Input-Output Structure of the Canadian Economy 1961-1981 Ottawa:Statistics Canada [105]  (1997) National Balance Sheet Accounts; Annual Estimates 1996 Ottawa:Statistics Canada  [106]  (2007) Canadian Economic Observer, Historical Statistical Supplement 2006/07 Ottawa:Statistics Canada  [107] T¨ornqvist, L. (1936) ‘The Bank of Finland’s Consumption Price Index.’ Bank of Finland Monthly Bulletin 10, 1–8 [108] T¨ornqvist, L., and E. T¨ ornqvist (1937) ‘Vilket a¨r f¨orh¨ allandet mellan finska markens och svenska kronans k¨ opkraft?’ Ekonomiska Samfundets Tidskrift 39, 1–39. Reprinted as p. 121-160 in Collected Scientific Papers of Leo T¨ornqvist, Helsinki: The Research Institute of the Finnish Economy, 1981 [109] Woodland, A.D. (1982) International Trade and Resource Allocation (Amsterdam: North-Holland) [110] Wooldrige, Jeffrey M. (2002) Econometric Analysis of Cross Section and Panel Data (Cambridge, MA: The MIT Press) [111] Zanardi, Maurizio (2005) ‘Antidumping: A Problem in International Trade.’ European Journal of Political Economy 22(3), 591–617  230  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.24.1-0069933/manifest

Comment

Related Items