The Canadian Productivity Review

Volume 2007
Number 7

Abstract

Introduction

Provincial productivity growth,
1997 to 2005

Provincial labour productivity
1997 and 2005

Conclusion

References

Figures

More information

PDF version

Other issues
in this series

About this
series

Introduction

A recent Statistics Canada release of provincial labour productivity and hours worked data¹ noted that short- and longer-term estimates of productivity growth can yield substantially different impressions of the economic progress being made in certain areas of the country. This earlier release compared estimates of provincial productivity growth for 2005 to annualized growth averages covering the 2000-to-2005 period. It noted that the comparatively large productivity gains experienced in 2005 in Western and Central Canada lose some of their luster when the focus shifts to the longer term, as the post-2000 growth averages in these regions are, on balance, far more modest than the annual rates for 2005. The reverse is true of the Atlantic provinces: while many experienced relatively lackluster productivity growth in 2005, average productivity gains for the post-2000 period have been much stronger. In several Atlantic provinces, these longer term growth averages are similar to those posted in Western and Central Canada.

Observations of the sort noted above invariably raise questions about how, in the final analysis, productivity performance compares in different regions of the country, and whether much has changed over time. The central focus of Statistics Canada's productivity program rests with the estimation of annual growth rates — statistics that are designed to measure short-run changes in the efficiency with which an economy transforms its production inputs into market outputs. Historically, less emphasis has been placed on examining the absolute differences in productivity that exist at different stages of the growth process — i.e., shifts in relative position over the short, medium and longer term.

This paper aims at bringing medium-term changes in the productivity performance of Canadian provinces into clearer view. In what follows, we report on the average growth in provincial labour productivity from 1997 to 2005, and examine how differences in productivity growth across provinces have affected relative levels of labour productivity over this nine-year period. Two methodological issues warrant emphasis.

First, our choice of a nine-year reference period² is not arbitrary, as this corresponds to the period for which the Canadian Productivity Accounts publishes data that are consistent with the new North American Industry Classification System. While a longer time series would also be advantageous for comparative purposes, here we limit our focus to the better part of the last decade.

Second, both productivity growth rates and productivity levels are calculated using constant (1997) dollars. In order to compare output across two points in time in a world where there are multiple goods, these goods have to be aggregated. An aggregated index weights each output by a price. In order to abstract from price changes, these indices usually hold prices constant — at beginning-period prices, end-period prices or some combination of beginning and end periods. In this paper, we compare growth between 1997 and 2005 by holding prices constant at their values in the opening period (1997), thereby providing measures of 'real' or constant dollar relative productivity that abstract from (or remove) the effect of price changes.

While it is standard to report growth rates in real terms, differences in the output levels across provinces are usually examined in current dollars because these reflect the actual price conditions relevant to the time period being used for the comparison. The relative position of a province may change over time because it increases its output with prices held constant, and/or because the prices it receives for this output change relative to those received by other provinces. The former is often referred to as the result of growth in 'real' output; the latter as the effect of relative price changes.

These price movements can affect perceptions of relative productivity. A province may experience comparatively strong growth in real productivity along with a decline, relative to other provinces, in its nominal productivity, if it specializes more than other provinces in the production of commodities whose prices are in relative decline. Or the reverse can happen. A province may not experience as much growth in real terms as other provinces but its relative position may improve when examined using the prices at which transactions take place, if the prices it receives increase relative to those in other provinces. In this paper, we report on level differences between provinces in 2005 based on 1997 prices — in order to avoid the distorting effects of relative price movements. But we also ask what the relative position of each province is in 2005 using the prices that existed in that year.

1. See "Hours worked and labour productivity in the provinces and territories," The Daily, May 9, 2006.



2. Our analysis period encompasses nine separate annual estimates of gross domestic product (GDP) and hours worked, from 1997 to 2005 inclusive.