Main article
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.
1 Introduction
The relative economic activity generated by small and large firms has received considerable attention in recent decades. Researchers have related differences in the size distribution across countries to differences in dynamism, maturity, and efficiency. 1 The rise in the unincorporated self-employed (without paid workers) from the late 1980s to the late 1990s has also contributed to the attention given to the small-firm population. 2
The importance (measured by employment) of small firms in the mid-1980s and the 1990s has been well documented. In the 2011 vintage of the Longitudinal Employment Analysis Program (LEAP), the share of large-firm employment in the total economy remained slightly below half throughout the 2000s (See subsection 5.1 of the Appendix, Table 16), reaching 46% in 2011.
Industry Canada’s Key Small Business Statistics provide measures of the employment of small and large businesses using the Survey of Employment, Payrolls and Hours (SEPH). These statistics indicate that the employment shares of large firms in the business sector went from 34% in 2002 to 35% in 2010 (See subsection 5.1 of the Appendix, Table 17).
To measure the importance of firm size, previous studies have relied predominately on employment size (an input to the production process) rather than on a measure of output (GDP or value added). In contrast, Leung and Rispoli (2012) examined the relative importance of small and large firms using GDP. 3
Since firm size is an often-cited structural feature that distinguishes the Canadian economy from the U.S. economy, precise measurement of the differences therein and analysis of the impact of different metrics—employment versus GDP—is important. 4 If different metrics yield inconsistent results, the debate about causes of structural differences needs to be nuanced. More importantly, differences in the metrics may provide additional information on the nature of the disadvantages that Canada’s industrial structure produces.
Changes in the economic environment and how it impacts the relative contribution of small and large firms over time also need to be examined.
This paper presents estimates of the contribution of small and large firms to Canadian gross domestic product (GDP) for the period from 2002 to 2008 that are comparable to those produced for the United States. In this paper, small firms are those with fewer than 500 employees, and large firms are those with 500 or more employees.
The objective is to determine whether the Canadian economy contains a greater proportion of smaller firms than the U.S. economy. The study finds that the relative importance of small firms, whether measured by GDP or by hours worked, is larger in Canada. In 2008, Canadian small firms generated 53.4% and 70.8% of business-sector GDP and hours worked, respectively; U.S. small firms generated 46.1% and 55.6% of business-sector GDP and hours worked, respectively.
Users of the GDP-by-size estimates should recognize that these estimates are point estimates and are sensitive to the methods, the assumptions, and the particularities of the source data used to derive them.
The next section, Section 2, outlines the methodology employed for calculating GDP and business hours worked by firm size. Section 3 presents the results for each country’s GDP estimates and hours worked by firm size, and Section 4 concludes.
2 Methodology
2.1 Coverage
This paper focuses on the business sector. In Canada, the business sector consists of all corporate businesses and unincorporated businesses organized for profit that produce goods and services for sale at a price intended to at least approximate the costs of production. Government business enterprises (GBEs) are included in this definition. Under this definition, the business sector in Canada accounted for 78% of total GDP in 2007. In the United States, it accounted for 80% of total GDP. 5
2.2 Measurement of gross domestic product
Value added in the Canadian and U.S. systems of national accounts is the sum of gross value added of all resident producer units. Canadian GDP is valued at basic prices while U.S. GDP by industry produced by the Bureau of Economic Analysis (BEA) is valued at market prices. All of these estimates are produced by industry but not by firm size. Furthermore, the official Canadian business-sector GDP does not include nonprofit organizations while the official U.S. private-sector GDP includes nonprofit organizations. 6
The next two subsections discuss how business-sector GDP by firm and industry was derived for 2002 to 2008.
2.2.1 Canadian gross domestic product
Value added in the Input-Output Accounts is the sum of gross value added of all resident producer units and is one of several measures of GDP in the Canadian System of National Accounts. 7 These estimates are produced at the industry level but not by firm size. 8 This paper produces GDP by industry and by firm size using the income approach for measuring GDP. Data from a number of sources are used to estimate the components included in the industry aggregates: wages and salaries; portions of supplementary labour income; mixed income; other operating surplus; and indirect taxes less subsidies.
To measure these components, the paper uses administrative data from Canada Revenue Agency files edited for purposes of the National Accounts. The main data source for operating surplus and indirect taxes on production less subsidies is the General Index of Financial Information (GIFI) income statements included with the T2 (Corporation Income Tax Return) corporate tax filings of firms and the T1 (General - Income Tax and Benefit Return) unincorporated tax filings. The main data source for labour income is the T4 (Statement of Remuneration Paid) slips. Employment of each firm is obtained from PD7 (statement of account for current source deductions) payroll deduction accounts.
The methodology is based on that developed by Kobe (2007, 2012). 9 The estimation of labour income was relatively straightforward. Salary and wages (the largest component of labour income) from each firm’s establishments (locations) were assigned to the appropriate size category of the firm. 10 The size class of the firm was based on the employment size of the commonly-controlled group of enterprises (ultimate parent) to which it belongs. Salary and wages from all locations within a firm were summed together to provide an estimate of the ultimate parent. The NAICS of the ultimate parent was assigned by determining the dominant industry that exhibited the largest salary and wages of an industry (s-level aggregation) belonging to the ultimate parent.
A different approach is employed in this paper to derive the estimate for the other GDP components for Canada. Following Kobe (2007), the non-compensation (primarily operating surplus) was produced mainly from tabulations of revenue and the components of GDP by revenue size class and industry published by the Internal Revenue Service (IRS) 11 as well as from tabulations of receipts and payroll by employment size class and industry from the U.S. Census Bureau. 12
In this study, the Canadian estimates of non-labour compensation were produced by tabulating the revenue and GDP components of the General Index of Financial Information (GIFI), T1, and PD7 by revenue size class (linked using the Business Register). The resulting shares across firm sizes for each component of GDP are benchmarked industry-by-industry to the Input-Output Accounts. 13 The aggregate estimates for the business sector are the sum of the benchmarked industry estimates.
2.2.2 U.S. gross domestic product
Value added in the Bureau of Economic Analysis (BEA) Industry Accounts is derived as the sum of gross value added of all resident producer units and is valued at market prices. These estimates are produced by industry but not by firm size. This paper uses the Kobe (2007) methodology (based on the income approach) to produce GDP by industry and by firm size. The BEA includes all industries in the private sector and nonprofit institutions.
Kobe’s (2007, 2012) estimates for the United States were produced mainly from tabulations of revenue and the components of GDP by revenue size class and industry published by the IRS as well as from tabulations of receipts and payroll by employment size class and industry from the U.S. Census Bureau. These tabulations were used to break down the BEA data on GDP by major component and industry.
The Statistics of U.S. Businesses (SUSB) program of the U.S. Census Bureau provided a special tabulation of payroll and receipts by employment size of firm, by industry, and by legal form of organization, which is used in this paper. All firms in the private sector are covered, including corporations and non-corporate enterprises as well as nonprofit institutions. 14 The public sector was excluded from the tabulations. For the wage and salaries component of GDP, the U.S. Census Bureau’s estimates of payroll by employment size of firm is used directly to split wage and salaries across firm size classes. For the non-labour compensation portions of GDP for corporations, an indirect approach is adopted since these data are not available from the U.S. Census Bureau. The government sector in the SUSB data was included because, in Canada, this sector operates government business enterprises, which are part of the business sector.
The IRS produces non-compensation components of GDP by size class, but only by revenue size, not by employment size. To combine the two sources of data, Kobe (2007) first took the share of receipts for large firms from the U.S. Census Bureau and multiplied it by total revenue reported by the IRS. This calculation gave the fraction of revenue accounted for by large firms. The amount was then used to find the number of revenue size classes that large firms account for. 15 The estimates of components of GDP by size obtained from the U.S. Census Bureau and IRS data are then benchmarked industry-by-industry to the BEA Industry Accounts.
2.3 Measurement of labour input
Labour input in this paper is measured as hours worked. The hours worked in official productivity programs of Canada and the United States are produced at the industry level, but not by firm size.
The next two subsections discuss the methodology used to estimate the labour input of the Canadian and U.S. business sectors from 2002 to 2008, and describe how the business sector is divided into small and large firms. To be consistent with the GDP estimates, nonprofit organizations are included with the business sector.
2.3.1 Canadian labour input
Labour input in this paper is measured as hours worked. This is done in two steps—first, by estimating jobs according to firm size for both paid workers and self-employed owners; and, second, by estimating hours worked per job for these two groups. The product of the two generates an estimate of total hours worked by firm size and industry for the period from 2002 to 2008.
At the first stage, hours worked are estimated for paid workers. This first requires an estimate of jobs of paid workers by firm size. The estimate of jobs by firm size by industry is derived from the PD7 file (business sector only) and benchmarked to the major industry level and number of jobs as published by Statistics Canada’s Productivity Accounts (SCPA). These include paid workers for both corporate and unincorporated enterprises. This was added to the estimate of employment of nonprofit organizations which came from the SCPA. 16
At the second stage, hours worked per paid worker were estimated. The latter is estimated by firm size and industry using the Labour Force Survey (LFS). 17 Total hours are calculated as the product of jobs and hours worked per job. These, in turn, were benchmarked to the business-sector hours worked of paid employees by industry as published by the SCPA. Again, an estimate of hours worked for nonprofit organizations was multiplied by usual hours worked of paid workers obtained from the SCPA.
The next step was to estimate the total hours worked of the unincorporated self-employed from the SCPA. In this paper, all of the unincorporated self-employed with zero employees were allocated into the small-size category while those with paid employees were derived as mentioned in the previous paragraph. Paid hours worked and unincorporated hours worked were added together in order to arrive at business-sector hours worked for small and large businesses.
2.3.2 U.S. labour input
As was the case for Canadian labour input, U.S. hours worked by firm size were estimated in two steps for the period from 2002 to 2008. Jobs of paid workers by firm (employment) size by industry were obtained from a special tabulation by the SUSB program of the U.S. Census Bureau that covered all industries in the private sector, including corporations and non-corporate enterprises as well as nonprofit institutions. The government sector was also included, since government business enterprises are included in the business sector in Canada. Non-employers obtained from the SUSB program were placed in the small-size category. These were then benchmarked to the official jobs of paid workers and self-employed derived from the BEA.
The next stage involved estimating hours worked by firm size and by industry. Usual hours worked for paid workers and non-corporate self-employed by firm size were obtained from the U.S. Current Population Survey (CPS) for the period from 2002 to 2008. These were then multiplied by the jobs data obtained in step one, in order to arrive at estimates by firm size. They were then benchmarked to the official hours worked of paid workers from the BEA and also to the Bureau of Labor Statistics (BLS) for the owners of firms in the non-corporate sector (i.e., self-employed owners). The two were then combined in order to arrive at the number of hours worked in the business sector by firm size and industry. To facilitate an accompanying comparison of Canada/U.S. productivity, (Baldwin, Leung and Rispoli, 2014), the hours worked by firm size were benchmarked to WORLD KLEMS. 18
3 Overview of the Canadian and U.S. economies
3.1 Nominal gross domestic product by industry
The following sections compare the structure of Canadian and U.S. industries in terms of nominal GDP from 2002 to 2008 for the entire business sector and by firm size.
3.1.1 Overview of nominal gross domestic product by industry
This section compares the structure of industries between Canada and United States from 2002 to 2008. It first looks at business-sector growth by industry in each country. The differences in industry composition are then analyzed.
From 2002 to 2008, nominal GDP in both the Canadian and the U.S. economies grew by an average of 6.0% per year for Canada and 4.9% per year for the United States (Table 1). During this period, the goods-producing and services industries grew by an average of about 6.4% per year in Canada and by an average of about 4.9% per year in the United States.
Most industries contributed to economic growth in both countries, with the exception of manufacturing, which declined in Canada. Canadian manufacturing growth was hampered by the appreciation of the Canadian dollar relative to the U.S. dollar during the early 2000s. GDP growth in most industries was similar between Canada and the United States. The exceptions were agriculture, which grew faster in the United States, and construction, which grew faster in Canada. The fastest-growing industry in both countries was Mines, Oil and Gas, which benefited from the resource boom starting in 2003: it grew by an average of 19.4% per year in Canada and by an average of 19.5% per year in the United States from 2002 to 2008.
The industry distribution (measured as a percentage of business-sector GDP) was very different in the goods-producing industries and the services industries (Table 2). The proportion of business-sector GDP for goods-producing industries was larger in Canada (42.3% in 2008) than in the United States (27.5% in 2008), while the proportion for services (57.7% in 2008) was much smaller in Canada than in the United States (72.5% in 2008).
In the goods-producing industries, the proportion of GDP in mining (including oil and gas) and construction increased in Canada. By 2008, the proportion of GDP in mining, oil and gas (13.1%) was four times larger in Canada than in the United States (2.9%). By 2008, the proportion of GDP for construction (9.1%) was almost twice as large in Canada as in the United States (5.6%).
For services, the proportion of GDP in financial intermediation in Canada (5.5% in 2008) was about three-fifths as large as that in the United States (9.5% in 2008). 19 Other differences occurred in education and in health and social work, where private education and private health services (i.e., private hospitals) are more prevalent in the United States than in Canada. 20 Real estate, renting, and business services accounted for a significant proportion of GDP in both countries, reaching 17.8% in Canada and 22.1% in the United States in 2008.
3.1.2 Overview of nominal gross domestic product by industry and by firm size
Over the period from 2002 to 2008, the small-firm share of GDP declined in both Canada and the United States. In Canada, small firms generated 57.1% of business-sector GDP in 2002; this fell to 53.4% in 2008 (Table 3). In the United States, small firms represented 47.8% of GDP in 2002. This fell moderately throughout the 2000s, to 46.1% in 2008 (Table 3). Relative to large firms, small firms play a larger role in the Canadian economy than their counterparts in the United States. The rest of the analysis examines industry detail to determine where these differences occurred. 21
The large-firm share of GDP at a more detailed industry level was generally greater in the United States than in comparable Canadian industries (Table 5). The United States had more industries (eight industries) whose large-firm share was more than 50% in 2008 than Canada did (five industries). The exceptions were electricity, gas, and water supply (Canada’s share is slightly larger), and mines, oil and gas. The highest large-firm share occurred in very capital-intensive industries (electricity, gas, and water supply; mines, oil and gas; manufacturing; transport and information) and in financial intermediation.
There were considerable differences across industries between the two countries in terms of the importance of large firms. The Canadian share of large-firm GDP was less than that in the United States more frequently in the services industries than in the goods-producing sector. In 2008, the large-firm share of the services industry in Canada was 40.2%, compared to 52.8% in the United States, while it was 55.4% in the goods-producing industries in Canada, compared to 56.8% in the United States. 22
The most notable increase in the share of large-firm GDP in Canada between 2002 and 2008 occurred in mines, oil and gas: the large-firm share of GDP in this industry increased by 12.2 percentage points, from 72.1% in 2002 to 84.3% in 2008. The increase in the growth of large firms in mines, oil and gas occured during the resource boom starting in 2003. A number of industries experienced a more moderate increase of roughly 6 percentage points in their contribution from large firms (transportation, real estate, and business services; wholesale and retail trade). This increase was partly offset by a decline in manufacturing as its large-firm share fell from 62.4% in 2002 to 56.6% in 2008. Apart from these two significant occurrences, increases in the share of large-firm GDP was widespread and occurred across many industries (ten industries). Similarly, increases in the United States share of large-firm GDP were spread across different industries (eight industries).
In Canada, large-firm nominal GDP growth for both goods-producing industries and services industries substantially outpaced that in the United States from 2002 to 2008 (Table 6). Canadian large-firm nominal GDP growth was much stronger in construction and in most services industries, but weaker in manufacturing.
Differences in the importance of large firms at the aggregate total-economy level depend not only on differences at the industry level but also on differences in the industrial structure (the importance of different industries). Comparisons of the source of large-firm GDP (or employment) reveal the joint effect of industry intensity and concentration of overall activity in that industry.
The percentage of total business-sector GDP in large firms is concentrated in specific industries (Table 7). There is more GDP in large firms in the goods sector in Canada than the United States and less GDP in large firms in the services industries. A notable difference was observed in the mines, oil and gas industry, where large-firm shares were about seven times larger in Canada than in the United States. Another notable advantage for Canada occurred in transportation, storage, and communication. Conversely, the United States had a higher large-firm share of GDP in health, in financial intermediation, and in real estate, renting, and business activities. 23
For small firms, nominal GDP in most Canadian industries from 2002 to 2008 grew at rates similar to those of their U.S. counterparts in services and for the entire business sector (Table 8). Larger differences were observed in the goods-producing sectors. GDP growth of small firms was substantially stronger in the U.S. than in Canada for agriculture, electricity, gas, and water supply, and mines, oil and gas, while small-firm GDP growth was stronger in Canada than in the United States in the construction sector.
The difference in the distribution of small-firm GDP between Canada and the United States was concentrated in a few industries (Table 9). The proportion of GDP in small firms was larger for construction in Canada while it was larger in health, financial intermediation, and real estate, renting, and business activities in the United States. These differences reflected differences in the overall industry structure of the two countries.
3.2 Hours worked
The following sections analyze the structure of Canadian and U.S. industries in terms of hours worked from 2002 to 2008 for the entire business sector and by firm size.
3.2.1 Overview
This section first reports business-sector growth by industry. This is followed by analysis of the similarities and differences in industry composition over this time period.
Hours worked in the business sector in Canada grew by an average of 1.5% per year compared to an average of 0.7% per year in the United States (Table 10). Most of the discrepancy was due to the faster pace of growth in the Canadian goods-producing industries (average growth of 0.7% per year) than in U.S. goods-producing industries (which fell by an average of 0.7% per year). Generally, hours worked grew in most industries in both countries. However, they fell in both countries for agriculture and manufacturing.
Growth in hours worked was substantially stronger across more industries in Canada than in the United States. This was the case for the following: electricity, gas, and water supply; construction; wholesale and retail; transportation, storage, and communication; financial intermediation; real estate, renting, and business activities; and other community, social, and personal services.
Industry structure (measured as a percentage of total business-sector hours worked) was similar in Canada and the United States for most industries and changed very little over the period (Table 11). The largest inter-country difference occurred in health, followed by wholesale and retail, transportation, construction, and financial intermediation.
The GDP and hours-worked proportions by industry were similar for most industries, with some notable exceptions. For example, the percentage of hours in mining was similar (about 1% to 2% in 2008) in the two countries while the percentage of GDP was 13.1% in Canada but only 2.9% in the United States.
3.2.2 Hours worked by firm size
Canadian small firms accounted for 70.8% of hours worked in the business sector in 2008. This was substantially larger than the result for the United States (55.6% in 2008) (See subsection 5.3 of the Appendix). 24 Hours worked by small firms declined in importance in both countries, falling from 73.0% in 2002 to 70.8% in 2008 in Canada and from 56.8% in 2002 to 55.6% in 2008 in the United States. Growth in business-sector hours worked for both small and large firms was stronger in Canada than in the United States from 2002 to 2008. Canadian small firms grew at 1.0% per year and large firms at 2.8% per year. U.S. small firms grew at 0.4% per year and large firms at 1.2% per year. 25
This difference in the share of hours worked of small firms between Canada and the United States occurs for two reasons. First, about two-thirds of hours worked by paid workers took place in small firms in Canada while in the United States the distribution between small and large firms is more even (Table 12). Second, there is a higher proportion of unincorporated self-employed owners in Canada (part of small firms) than in the United States.
Across goods-producing industries and services sectors, there are considerable differences between the two countries in the share of hours worked in large firms, with shares in the United States very much higher (Table 13). This occurred in every industry except for electricity, gas, and water supply, and mines, oil and gas. The United States had more industries (seven industries) whose large-firm share was more than 50% than Canada (three industries) in 2008. The tendency to employ proportionally more workers in large firms extends across much of the industrial spectrum.
Increases in the share of hours worked for large firms occurred across many industries in Canada (eight industries) and the United States (eight industries). In both countries, declines in the large-firm share of hours worked occurred in agriculture, manufacturing, and financial intermediation.
Overall, the difference in the industrial structure of hours worked (in terms of the percentage of total business-sector hours worked) for large firms is concentrated in certain industries (Table 14): Canada had larger proportions in wholesale and retail, transportation, mines, oil and gas, and manufacturing; and the United States had larger proportions in health, real estate, renting, and business activities, and hotels.
Contrary to the situation with respect to large firms, the differences in the industrial structure of hours worked (in terms of the percentage of total business-sector hours worked) for small firms across industries between Canada and the United States were small, with one notable difference—health (Table 15). 26 Cross-industry differences in the relative importance of an industry (as measured by hours worked) therefore are attributable to a greater extent to differences in the distribution of hours worked among large firms across industries.
4 Conclusion
This paper presents estimates of the contribution of small and large firms to Canadian business-sector GDP and hours worked that are comparable to those produced for the United States.
It investigates whether the Canadian economy exhibits a greater proportion of smaller firms (relative to large firms) than the U.S. economy. The relative importance of small firms in terms of GDP and hours worked is larger in Canada than in the United States. In 2008, Canadian small firms generated 53.4% of business-sector GDP and 70.8% of hours worked. U.S. small firms generated 46.1% of business-sector GDP and 55.6% of hours worked. The small-firm share of both GDP and hours worked declined in Canada and the United States from 2002 to 2008.
The United States had eight industries, whose large-firm GDP share was more than 50% where as Canada had five in 2008. The three industries where the large firm GDP share was more than 50% in the United States but not in Canada were health, education, and wholesale and retail.
The most notable increase in the share of GDP in large firms in Canada, between 2002 and 2008, occurred in mines, oil and gas. The relative importance of this industry (measured as a percentage of total business-sector GDP) was about seven times larger in Canada than in the United States. These trends were not observed for hours worked.
The relative importance (measured as a percentage of total business-sector GDP) of small firms in the construction industry was almost twice as large in Canada as in the United States. For both large and small firms, the relative importance of financial intermediation was substantially larger in the United States than in Canada.
The study also found that, from 2002 to 2008, business-sector nominal GDP and hours worked grew faster in Canada relative to the United States for both small and large firms. From 2002 to 2008, growth of business-sector nominal GDP was 4.9% among Canadian small firms and 7.5% among Canadian large firms and 4.3% among U.S. small firms and 5.5% among U.S. large firms. Furthermore, from 2002 to 2008, business-sector hours worked stood at 1.0% for Canadian small firms and 2.8% for Canadian large firms and at 0.4% for U.S. small firms and 1.2% for U.S. large firms.
Two separate metrics are used in this paper to measure the importance of small firms in the two North American economies. These are output (as GDP) and employment (as hours worked). Both metrics indicate that small firms are more important in Canada. Furthermore, the impression of a Canadian economy dominated by small firms is much more striking when employment, rather than output, is used. This difference indicates that the relative productivity of small firms vis-à-vis large firms is lower in Canada than in the United States. The related paper The Canada–United States Labour Productivity Gap Across Firm SizeClasses (Baldwin, Leung and Rispoli, 2014) will examine these differences in more detail and look at the extent to which the Canada–U.S. labour productivity gap is due to the differences between Canada and the United States in industry structure, the greater importance of small firms in Canada, and the gap in the productivity of small and large firms between the two countries.
5 Appendix
5.1 Various employment indicators
5.2 Hours worked by industry and by firm size in Canada and the United States
5.3 Proportion of total business-sector hours worked by firm size in Canada and the United States
5.4 Small and large firms’ business-sector gross domestic product and hours worked after removing health and education
5.5 Small and large firms’ share of business-sector gross domestic product and hours worked after removing health and education
5.6 Proportion of total business-sector (nominal) gross domestic product for large firms by industry in the Canadian and U.S. business sectors after removing health and education
- Date modified: