Main article

View the most recent version.

Archived Content

Information identified as archived is provided for reference, research or recordkeeping purposes. It is not subject to the Government of Canada Web Standards and has not been altered or updated since it was archived. Please "contact us" to request a format other than those available.

Archived

This page has been archived on the Web.

1. Introduction
2. Methodological issues
3. Importance of small, medium and large firms
4. Nominal gross domestic product per hour worked by firm size
5. Effect of changes in size distribution on labour productivity
6. Conclusions

1   Introduction

Discussions of Canada’s economic performance have asked whether changes in industrial structure towards larger firms would have a positive impact on productivity levels. Because larger firms are generally more productive than small firms, shifts in the firm size distribution toward larger firms might be expected to have a favourable impact on overall productivity.

Productivity gaps between small and large firms have been attributed to a number of factors: scale economies related to plant size; shorter production runs; higher prices of capital relative to labour that give rise to differences in capital per worker; and differences in managerial efficiency.

The factors behind and the effect of Canada’s relatively large concentration of smaller firms have been investigated by a number of researchers (Baldwin and Gorecki 1986; Inwood and Keay 2005; Leung, Meh and Terajima 2008). ¹  The issue is usually approached indirectly by developing estimates of improvements in labour productivity that derive from being able to exploit scale economies, and by then correcting for the impact of differences in firm or plant size on estimates of the relative levels of labour productivity.

Previous studies have been limited by a lack of comprehensive data on differences in productivity across firm size classes. This paper overcomes that limitation by developing the first set of Gross Domestic Product (GDP) estimates by firm size and combining them with estimates of employment in order to generate labour productivity estimates by firm size for an extensive set of industries. The effects of shifts in the distribution of firm size on overall productivity is assessed. The study covers the 2002-to-2008 period. GDP per hour worked is used as the measure of labour productivity. ² 

The study builds on Leung and Rispoli (2012) who estimate the contributions to GDP made by small, medium-sized and large businesses from 2001 to 2008. A firm, or business, in this study is defined as all units controlled by a parent firm. Firms are classified by the employment of the ultimate parent: small firms have 0 to 99 employees; medium-sized firms, 100 to 499 employees; and large firms, 500 or more employees. The contribution of large firms to business sector GDP increased steadily from 44.4% in 2002 to 47.9% in 2008. Small and medium-sized firms accounted for the other 52.1% of business sector GDP in 2008. The resource boom figures prominently in explaining how the contribution of large firms in the mining and oil and gas industry increased. The developments in mining and oil and gas outweighed the impact of the challenges in manufacturing, an industry where large firms have dominated historically.

The present study extends the analysis to examine labour inputs (measured by hours worked) and aggregate labour productivity for each firm size class. In Canada, large firms in the corporate sector are more capital intensive than are small and medium-sized firms. The paper, therefore, investigates how labour productivity differs across firm size classes and how it changed over the period. This information is also used to investigate the magnitude of changes in aggregate productivity that would be expected to accompany changes in the distribution of firm size. This exercise is intended to inform the debate on whether an increase in the percentage of large firms might be expected to have much influence on Canadian productivity levels.

2   Methodological issues

2.1  Coverage

In Canada, the business sector comprises all corporate businesses and unincorporated businesses that are organized for profit, and that produce goods and services for sale at a price intended to at least approximate the costs of production. Government business enterprises are included as part of this definition. The business sector accounted for 78.2% of total GDP in Canada in 2008. ³ 

2.2  Definition of productivity

The focus is labour productivity, an indicator of the efficiency with which the economy uses labour to produce goods and services. Labour productivity will be higher in sectors where workers have more capital, and where firms exploit economies of scale, employ more skilled workers, or use more advanced technologies.

Labour productivity is defined here as output per unit of labour input. Output is measured using GDP calculated at basic prices. Labour input is measured using hours worked. The labour productivity of industry i is the weighted average of the labour productivity in the small, medium-sized and large business categories for industry i, where the weights are the share of hours for each business-size category in industry i.

where LP represents labour productivity, H represents hours worked, and j indexes the business size categories. Labour productivity for the business sector as whole is the summation across all industries of the labour productivity of an industry multiplied by its share of business sector hours worked:

2.3  Measurement of gross domestic product

Business sector GDP is measured at basic prices and in nominal terms. Value added in the Input-Output Accounts is one of several GDP measures produced in the Canadian System of National Accounts. Value added in the Input-Output Accounts is the sum of gross value added of all resident producer units. ⁴ 

The Input-Output Accounts provide estimates of GDP at the industry or national level, but they are not available by firm size. To obtain measures of the income components of GDP by industry and by firm size, data from several sources are used to estimate the components included in the industry aggregates: wages and salaries; portions of supplementary labour income; other operating surplus; and indirect taxes less subsidies.

This study generally relies on administrative data from the Canada Revenue Agency (CRA). The General Index of Financial Information income statements included with the T2 corporate tax filings of firms are the main data source used to generate measures of revenue and operating surplus and indirect taxes on production less subsidies. The main data source for labour income is the T4 Statement of Remuneration Paid forms issued by every business to its employees for income tax purposes. The employment of each firm is obtained from PD7 payroll deduction accounts.

The estimates for the firm size classes are derived from Leung and Rispoli (2012). Each location in the firm is allocated to a size class based on the employment size of the commonly controlled group of enterprises (ultimate parent) to which it belongs, and all locations are summed to provide an estimate of the GDP of the ultimate parent. ⁵ 

The resulting shares across firm sizes for each component of GDP are benchmarked industry by industry to the Input-Output Accounts. ⁶  The aggregate estimates for the business sector are the sum of the benchmarked industry estimates.

2.4  Measurement of labour input

Labour input in this analysis is measured as hours worked for paid workers, as well as, for the unincorporated self-employed.

This is calculated in two steps for paid workers—first, with an estimate of jobs, and then, with an estimate of hours worked per job. The product of the two yields an estimate of total hours worked for paid workers.

In the first step, hours worked are estimated for paid workers. This requires an estimate of jobs of paid workers by firm size. The estimate is derived from the Payroll Deductions file (PD7) (business sector only) and benchmarked to the major industry level to number of jobs, as published by Statistics Canada’s Productivity Accounts.

The CRA requires that each employer open and maintain a payroll account and be responsible for deducting, remitting and reporting payroll deductions for each employee. This account has the employer’s Business Number (BN), a unique identifier assigned by the CRA. Some enterprises have simple business structures, in which the enterprise and establishment are the same, and generally have only one BN. By contrast, complex enterprises may have many BNs. In such cases, allocation of employment to locations is based on the structure of the enterprise, which is taken from the Business Register.

In the second step, hours worked per paid worker are estimated by firm size and industry using the Labour Force Survey.

Total hours of paid workers are calculated as the product of jobs and hours worked per job. These are benchmarked to the business sector hours worked by paid workers by industry, as published by Statistics Canada’s Productivity Accounts.

Finally, total hours worked by the unincorporated self-employed are taken from the Productivity Accounts. For this analysis, all the unincorporated self-employed are placed in the small-firm category.

3   Importance of small, medium and large firms

3.1  Gross domestic product

Over the 2002-to-2008 period, the share of small firms in GDP declined. Small and medium-sized firms generated 55.6 % of business sector GDP in 2002; this fell to 52.1% in 2008 (Tables 1 and 2). The GDP of small firms grew 5.3% per year on average over this period, and of medium-sized firms, 3.4% per year. The output growth in both groups was less than the 7.3% annual growth rate of large firms.

Large firms’ share of GDP differed considerably across industries, ranging from less than 4% in agriculture to 93% in utilities (Table 3). Industries whose large-firm share of GDP surpassed 40% throughout the 2002-to 2008-period were: utilities; mining and oil and gas; information and culture; manufacturing; transportation; arts and entertainment; and finance and real estate.

The most notable increase in the share of GDP produced in large firms was in mining and oil and gas, where the large-firm share rose by 13.6 percentage points from 69.3% in 2002 to 82.9% in 2008. The growth of large businesses in mining and oil and gas occurred during the resource boom. A number of industries experienced more moderate increases (6 to 8 percentage points) in their large-firm contributions to GDP: transportation, finance and real estate, information and culture, wholesale trade, and administrative services.

Some industries—utilities, manufacturing, other services, education, and agriculture—experienced a decline in the large-firm share of GDP over the 2002-to-2008 period. The decline in the large-firm share in manufacturing is noteworthy, given manufacturing’s substantial contribution to business-sector GDP. The decline in manufacturing GDP over the 2002-to-2008 period that coincided with the significant appreciation of the Canadian dollar relative to the U.S. dollar occurred mainly in large businesses. The share of large firms in manufacturing fell from 61% to 56%. The changes in manufacturing coincided with absolute declines in nominal GDP in this sector (Table 4). The GDP of large manufacturing businesses decreased at an average annual rate of 2.2% between 2002 and 2008. By contrast, small manufacturing businesses increased their GDP by an average of 3.1% per year.

3.2  Hours worked

The growth rates of hours worked and GDP were similar across firm size classes. Like GDP, the increases from 2002 to 2008 in hours worked for large firms exceeded those for medium-sized and small firms. Total hours worked in large firms rose by 3.4% per year, compared with 1.0% per year for medium-sized firms and 0.6% per year for small firms (Table 5).

The similarities in the growth rates of hours worked and GDP across firm size classes prevails at the industry level. The industries with strong or moderate gains in the GDP share of large firms (mining and oil and gas, transportation, finance, wholesale trade, and administrative services) also showed stronger hours growth in the large-firm category. For example, in mining and oil and gas, the average growth rate in hours worked for large firms was 9.4%, compared with 3.7% for medium-sized firms, and 6% for small firms.

Moreover, in industries where the large-firm share of GDP declined, growth of hours worked in large firms tended to be weaker than that in small and medium-sized firms. For example, in manufacturing, the average growth rate of hours worked in large firms was -2.5%, compared with -1.9% in medium-sized firms and -1.6% in small firms.

The relative share of hours worked in large businesses rose from 28.3% in 2002 to 31.6% in 2008 (Table 6). Similar to GDP, increases in the hours share of large businesses were highest in mining and oil and gas, wholesale trade, retail trade, finance, transportation, information, and administrative services. And the hours share of large businesses in manufacturing declined as did its GDP.

The share of hours worked in medium-sized firms was generally constant from 2002 to 2008. The two exceptions were in construction, where the share rose from 9% to 12%, and in mining and oil and gas where the share fell from 19% to 16%.

The share of hours worked of small firms declined overall from 58.4% to 55.5%. The largest declines were in transportation, information, construction, wholesale trade, retail trade, administrative, and accommodation.

4   Nominal gross domestic product per hour worked by firm size

Large firms differ from small firms in terms of capital intensity. The growth that transforms small into large firms often involves the application of more capital per worker to mechanize processes or to develop intellectual capital (Caves and Pugel 1980). Larger firms are more likely to make intangible investments, such as in advertising, skill enhancement, and research and development. ⁷  As well, larger firms tend to hire workers with more education and skills and pay higher wages (Brown and Medoff 1989, and Morissette 1991). These practices lead to differences in labour productivity between large and small firms.

In 2008, as measured by nominal GDP per hour worked, labour productivity in large firms ($72) exceeded that in medium-sized ($42) and small firms ($35) (Table 7). The labour productivity of medium-sized firms was 59% that of large firms; the labour productivity of small firms was 48% that of large firms (Table 8).

In Canada, large firms have a greater presence in industries that require substantial capital than do small and medium-sized firms. Dividing industries into those with “higher large-firm concentration” (GDP shares for large firms of about 50% or more) and those with “lower large-firm concentration” roughly splits industries into those that are capital-intensive and those that are non-capital-intensive. The industries in the higher large-firm concentration group are infrastructure industries (utilities, information and transportation), resource-based industries (mining and oil and gas) and manufacturers that are heavy consumers of capital (Baldwin and Dixon 2008). The labour productivity of large firms in these industries was $100 per hour worked in 2008, more than twice that in industries where large firms are less dominant ($39).

Scale effects are sufficiently important that small firms in capital-intensive industries experience greater productivity disadvantages than firms elsewhere. In 2008, small firms were only 50% as productive as large firms in the industries where large firms were more important. However, in industries where large firms were less dominant, the labour productivity of small firms was 77% that of large firms.

Evidence suggests that nominal labour productivity increased in small and medium-sized firms relative to large firms in industries where smaller firms were more important. The labour productivity of small relative to large firms increased in 9 of the 10 industries in the “lower large-firm concentration” group. The growth rates of labour productivity for small and medium-sized firms surpassed those of large firms.

This catch-up phenomenon was less prevalent in industries where large firms were more important. The exception was manufacturing, where the catch-up was probably related to the declining fortunes of large manufacturers.

5   Effect of changes in size distribution on labour productivity

The tendency for small firms to be less productive than large firms has led analysts examining differences in Canadian and U.S. productivity levels to ask whether industrial structure might provide an explanation. A full answer to this question is beyond the scope of this study, as it would require a detailed comparison of Canada–United States firm-size differences by industry and a calculation of the proportion of the productivity difference attributable to differences in firm-size distributions.

Nonetheless, it is possible to shed some light on the issue by calculating the amount by which an arbitrary shift of total hours from small and medium-sized firms to large firms would increase average labour productivity, while holding constant the labour productivity of each size class. ⁸ 

A counterfactual is created by arbitrarily increasing the large-firm share of hours worked and recalculating aggregate labour productivity, and expressing the new productivity level as a percentage of the original level. This is done by calculating average productivity for large firms and for small and medium-sized firms combined, and then, reweighting the two. Three scenarios are considered: hours shares for large firms are increased by 10%, 25% and 50% from those existing in 2002. To put these scenarios in context, over the 2002-to-2008 period, the hours share of large business in the Canadian business sector increased by about 10%.

The counterfactual is calculated for each industry separately and the results are summed to estimate the impact for the business sector overall (Table 9). ⁹  For the entire business sector, increasing the hours share of large firms by 10%, 25% and 50% would increase productivity in 2002 by 2%, 4%, and 7%, respectively. ¹⁰ 

In the “higher large-firm concentration” industries, changes in industrial structure have a substantial impact on labour productivity because of the wide gaps in productivity between small and large firms. In mining, labour productivity would increase by 4%, 9% and 18%, respectively, under the three scenarios. In manufacturing, it would increase by 3%, 8%, and 17%.

In industries where large firms have a smaller presence and less of a productivity advantage, changes in the size distribution generally have much less impact. In construction, the first two scenarios lead to less than a 1% gain in productivity. The largest increases occur in wholesaling—2%, 4%, and 8%.

The results of these scenarios need to be set against estimates of the size of the productivity gap between Canada and the United States. Baldwin, Leung and Rispoli (2011) estimate that the gap in real business-sector GDP per hour worked in 1998 was about 12 percentage points—Canadian real GDP per hour worked at the end of the 1990s would have to have increased by about 10% to have closed the gap. About a third of this was due to sole proprietorships in the unincorporated sector—a group classified here in the smallest firm size category. ¹¹ 

The changes derived from the counterfactual experiment would not eliminate this gap entirely, but they would reduce it. A 10%, 25% and 50% increase in the hours share of large firms would close the gap in 1998 by about 12%, 27% and 56%, respectively. But the Canada/U.S. labour productivity gap has widened substantially since 1998 and the increase in nominal labour productivity from an increase in large firms will therefore have less of an impact on the closing of the gap. Further detail on how this impact played out over the post 2000 period will be presented in a forthcoming research paper that compares GDP per hour worked in Canada and the United States over the 2002 to 2008 period (Baldwin, Leung, Rispoli, 2013).

6   Conclusions

Canada has often been described as having a lack of large firms, a consequence of which, it has been argued, is lower aggregate productivity. Analysis of this issue requires comprehensive and consistent data on firm-size distributions and differences in productivity levels across firm size classes. Until now, these data have not been available.

This study and related papers in this series (Rispoli 2009; Leung and Rispoli 2011) examine the importance of firm size by measuring the GDP generated, the employment provided, and consequently, the labour productivity by firm size class.

In 2008, productivity, as measured by nominal GDP per hour worked was $72 for large firms, well above that for medium-sized ($42) and small firms ($35).

Nominal GDP per hours worked of small and medium-sized firms is much lower than that of large firms for capital-intensive industries (mining and oil and gas, utilities, manufacturing, transportation and information).

Important changes occurred post-2002. First, large firms increased their share of both GDP and labour. Second, some of the differences between the labour productivity of large and small firms declined over the period—especially in industries where large firms were less important.

An examination of the effect of changes in industrial structure (distribution of hours worked across firm size categories within each industry) on labour productivity finds that a 10% increase in the hours share of large firms (similar to the change post-2002) would have a relatively small impact on Canadian labour productivity (2%), assuming that changes in relative hours worked in the different firm size classes are not accompanied by changes in the relative productivity of the different size classes.