A Sectoral Analysis of Ontario's Weak Productivity Growth

Since 2005, labour productivity growth in Ontario’s business sector has been zero, greatly under-performing the rest of Canada and being single-handedly responsible for most of what has been described as “Canada’s dismal productivity growth.” This article examines the issue through detailed sectoral data, and finds a wide range of variation underlying the average productivity growth rate. Some important sectors have maintained decent productivity growth. Other sectors, especially in manufacturing, saw the level of productivity decline significantly. Empirical evidence suggests that weak aggregate demand—due to the high Canadian dollar, the U.S. recession, and global restructuring—was the main cause of weak productivity. Weak demand led to lost economies of scale, particularly due to compositional shifts in the economy. ONE OF THE MOST WIDELY used economic performance indicators is labour productivity, de f ined as rea l GDP per hour wor ked . 2 Ontario's performance in terms of labour productivity has been very poor over the past several years. This has caused a considerable amount of concern. It is widely noted that, in the long run, increases in real wages and living standards come mainly from productivity growth.3 When Canadian economists talk about competitiveness, they almost always bring in the subject of productivity. Many argue that Canada's poor productivity growth has caused international competitiveness to decline. Certainly, our economy appears less competitive than it used to be, based on the evidence of Canada's large trade deficit. If Canadians could be induced to work harder or smarter to boost productivity, it is claimed, competitiveness would improve. But what if it sometimes works the other way around? What if it is competitiveness that affects productivity growth? That is to say, what if an overvalued exchange rate, by reducing competitiveness and the scale of output, causes lower productivity growth? This article argues that productivity, when constrained by demand factors, is not an inde1 Peter S. Spiro is Executive Fellow, Mowat Centre for Policy Innovation, School of Public Policy and Governance, the University of Toronto. An earlier version of this article was presented at the annual meeting of the Canadian Economics Association in Montreal, May 31-June 2, 2013. The author has benefited from the comments of Andrew Sharpe, Jianmin Tang, Wulong Gu, Christopher Ragan, Alvaro del Castillo, Matthias Oschinski, and two anonymous referees. Email: [email protected]. 2 All references to productivity in this article refer to labour productivity. 3 It should be noted, however, that there is no fixed relationship between productivity and real wage growth. Such a relationship would only exist if the economy was characterized by a simple Cobb-Douglas type production function. Most empirical evidence does not support it. Changes in the relative prices of imports and exports can also play a significant role in raising the standard of living. This is the main reason why Alberta has been doing so well in terms of real income growth. 20 NU M B E R 26 , FA L L 2013 pendent causal factor that determines standards of living. While there are many reasons for anxiety about the Ontario economy, an excessive focus on low productivity growth, as if it was always a driving factor, is misplaced. There is both a demand side and a supply side to productivity. From the supply side, there is potential for productivity growth when technology improves, and when workers become more educated and have more and better capital equipment to work with. However, this potential will not be realized if there is insufficient demand. If highly educated individuals are relegated to driving cabs or selling shirts, their investment in education will be wasted. If more output cannot be sold, it will not make sense for companies to invest in more and better equipment to increase output. Business commentators generally focus on the average productivity growth for the whole economy. If most industries were clustered close to this average, it would be a meaningful indicator. In fact, this average can be misleading, inasmuch as it is the random outcome of a wide range of different sub-components and may therefore not be representative. In order to understand productivity, it is necessary to see what lies beneath, and to look at its performance at the detailed sectoral level. The possibility of doing this has been facilitated by a new experimental database from Statistics Canada that provides detailed sectoral productivity by province.4 As we peel away the layers, it will be possible to better understand why productivity growth has slowed so sharply in Ontario. The article consists of seven sections. The first sect ion provides a brie f overview of Ontario’s productivity performance. The second section illustrates how the composition of output can affect aggregate productivity. The third section examines the effect of export demand on the composition of output, along with implications for productivity. The fourth section looks at service sector productivity. The fifth section presents a three-digit NAICS decomposition of Ontario’s productivity growth. The sixth and most important section provides a regression analysis of the relationship between output and productivity growth in manufacturing. The seventh and final section concludes and provides directions for future research. A Brief Historical Overview Over the past ten years, Ontario has diverged very sharply from the performance of the rest of Canada (henceforth abbreviated as ROC, which is the total for Canada minus Ontario). Ontario has fallen behind both in labour productivity growth and in a host of other indicators such as employment growth and investment growth. This is due to the fact that Ontario's economy was the most open to international trade in general, and dependent in particular on exports of finished goods and services rather than raw commodities. It was therefore the most susceptible to both the sharp upward valuation of the Canadian dollar that began in 2003, and the deep recession among developed countries (and in particular the United States) that began in 2008. Ontario had much more to lose than the rest of Canada, starting with a greater dependence on exports. The early 2000s were already weak because of the dot-com recession in the United States. Between 2002 and 2008, Ontario's international exports as a share of GDP fell by nearly 10 percentage points (Chart 1). In contrast, there was hardly any decline in the rest of Canada, where the rising dollar's effect was largely offset by rising international commodity prices. The recession reduced the export/GDP share by an additional 5 percentage points or so, causing 4 The author is indebted to Hugh Finnigan for providing the data, and to Qaizar Hussain and Hugh Finnigan at the Ontario Ministry of Finance for their earlier analysis of the data. I N T E R N A T I O N A L PR O D U C T I V I T Y MO N I T O R 21