Notes

^1.

Baldwin and Gorecki (1986) and Leung et al. (2008) attribute about half of the difference between Canadian and U.S. manufacturing productivity to differences in size of plants. Inwood and Keay (2005) examine a longer period and find that plant size also contributes about half of the difference.

^2.

Explanations include, but are not limited to, the smaller size of the Canadian market.

^3.

Other sectors of the domestic economy, including government, households, and institutions, are part of the non-business sector. In the non-business sector, measurement of output is difficult since output is not sold at an explicit market price. As a result, it is generally valued at the cost of its inputs. Therefore, measures of growth in the volume of output differ little from measures of the growth of inputs. This implies that estimates of productivity growth are close to zero by construction.

^4.

The official Canadian estimates of gross domestic product for the business sector include an imputation for the service provided by owner-occupied dwellings. This estimate is moved from the business sector to the non-business sector for this analysis since it does not fit the definition of the business sector (i.e., entities selling goods at market price) that is used here.

^5.

See Concepts and Methods of the U.S. National Income and Product Accounts (published by the Bureau of Economic Analysis, U.S. Department of Commerce).

^6.

Quoted from the website of the Canada Revenue Agency: www.cra-arc.gc.ca/tx/bsnss/tpcs/slprtnr/prtnrshp/menu-eng.html.

^7.

See Luttrell et al. (2006) for a discussion of the tax treatment of various legal forms of organization.

^8.

Many of the definitions in this section are taken from U.S. Bureau of Economic Analysis (2009).

^9.

NIPA defines these entities as part of the non-corporate sector. In Canada, GBEs are a component of corporations. For comparability, GBEs in NIPA were moved to the corporate sector from the non-corporate sector for purposes of the comparisons produced in this paper.

^10.

The source for the U.S. data is the Bureau of Economic Analysis, U.S. Department of Commerce.

^11.

The estimate of sole proprietorships and partnerships for Canada was based on the aggregation of GDP items (labour income, interest payments, taxes paid, depreciation, and net income) from the redesigned T1 (T1 GeneralIncome Tax and Benefit Return) 2008 tax data for unincorporated enterprises. The redesign involved separating enterprises, as provided in the 2008 T1 tax returns, into sole proprietorships and partnerships. A database on partnerships was developed, which linked the individual T1 enterprises involved in partnerships to other partnerships by means of Form T5013 (Statement of Partnership Income) information or other partnership information reported on T1 tax returns. At present, these data are not available for years prior to 2008. As a consequence, in this paper, sole proprietorships and partnerships are assigned shares for years prior to 2008 on the basis of 2008 shares. For the period from 2002 to 2005, labour income with respect to sole proprietorships was drawn from T4 (Statement of Remuneration Paid) data. Sole-proprietorship GDP for the years 2002 to 2005 was generated by first estimating labour income from T4 data for the entire period and then applying the 2008 shares (excluding labour income) from the T1 tax returns to the unincorporated GDP (excluding labour income) for the period.

^12.

The division of U.S. GDP for the unincorporated sector into the parts generated by sole proprietorships and partnerships was based on the aggregation of GDP items (labour income, interest payments, depreciation, taxes paid, and net income) from Internal Revenue Service (IRS) Table 10 (Nonfarm Sole Proprietorship Returns: Selected Income Statement Items) and Table 11 (Partnership Returns: Selected Balance Sheet and Income Statement Items).

^13.

The IRS defines financial investments as securities, commodity contracts, and other financial investments and related activities.

^14.

Comparisons at basic prices were also carried out. The official U.S. estimates (valued at market prices) were adjusted to a basic price. Indirect taxes and subsidies on products and production were adjusted by keeping only indirect taxes on production. The required data for this exercise were obtained from BEA Table 900,Detailed Tax and Social Contribution Receipts by Type of Tax or Social Contribution and Receiving Subsector. The BEA does not produce estimates at basic prices. The conclusions derived using basic prices are not substantially different from the ones presented in this paper using market prices.

^15.

IRS Table 10 includes animal production but excludes crop production. In order to make IRS Table 11 consistent with IRS Table 10, non-farm industry totals for partnerships that exclude crop production were used.

^16.

The net income (loss) obtained from IRS Tables 10 and 11 represented about 95% of the net income (less loss) of non-farm proprietorships and partnerships (plus payments to partners) obtained from the IRS as reported in BEA National Economic Accounts Table 7.14. Relation of Nonfarm Proprietors' Income in the National Income and Product Accounts to Corresponding Measures as Published by the Internal Revenue Service.

^17.

Since unincorporated enterprises are smaller, they likely pay less per hour than do corporations. Therefore, the share of labour income may underestimate the hours worked in the unincorporated sector and overstate labour productivity in this sector relative to the corporate sector.

^18.

Using mixed income to separate self-employed hours worked by sole proprietorships and self-employed hours worked by partnerships would give slightly lower hours worked for sole proprietorships (about 1.0% to 3.5% lower from 1998 to 2005). Using mixed income to split hours worked assumes that owners of sole proprietorships earn about the same net income per hour as partners in partnerships. However, sole proprietorships are generally much smaller than partnerships and generate most of the sole-proprietorship GDP. Larger partnerships, on the other hand, generated most of the partnership GDP. Consequently, sole proprietorships likely earn less per hour than the partners in partnerships. Therefore, using mixed income to split hours worked may underestimate hours worked in sole proprietorships and overstate the latter's labour productivity.

^19.

An alternative method to using the wage bill for estimating Canadian hours worked is presented in Section 7. Instead of employee hours worked in sole proprietorships being estimated by means of labour shares, employment was arrived at by means of an estimate based on employment figures obtained from the Survey of Employment, Payrolls and Hours (SEPH) (Economic Analysis Division of Statistics Canada). However, data are not available on employment of sole proprietorships for prior years; the growth in the number of T4 slips issued to employees of sole proprietorships was used instead. Hours worked for sole proprietorships were derived by applying these ratios to the business-sector hours worked, as shown in Table 3.

^20.

IRS Tables 10 and 11 include both corporate and non-corporate partnerships. IRS (Estimated Data Line Counts Individual Income Tax Returns [Part II: Income or Loss From Partnerships and S Corporations] from IRS Schedule E [Form 1040] [Supplemental Income and Loss]) provide the number of non-corporate partners for the period from 2003 to 2005. This information was used along with data on the number of sole proprietorships from IRS Table 10 in order to obtain the share of sole proprietorships in the non-corporate sector.

^21.

When mixed income is used to separate self-employed hours worked between sole proprietorships and partnerships, a much lower estimate for the period from 1998 to 2005 is produced with respect to hours worked for sole proprietorships—ranging from 36% to 68% depending on the year. This assumes that net income of sole proprietorships is the same as that of owners of partnerships. However, sole proprietorships are generally much smaller than partnerships: they generated most of their GDP from smaller-sized enterprises while, in the case of partnerships, most of the GDP came from larger partnerships. Consequently, sole proprietorships likely earn less per hour than do partnerships. As a result, using the share of mixed income may underestimate the hours worked in sole proprietorships and overstate the latter's labour productivity.

^22.

An alternative to using the wage bill for estimating U.S. hours worked is presented in Section 7. Instead of using labour shares of sole proprietorships to estimate employee hours worked in this group of businesses, employment figures from the 2002 and 2007 Statistics of U.S. Businesses (U.S. Census Bureau) were used. Figures for the 2003-to-2006 period were estimated by means of straight-line interpolation. Hours worked for sole proprietorships were derived by applying these ratios to the business-sector hours worked, as shown in Table 4.

^23.

These do not correspond to estimates in Baldwin and Rispoli (2010), which exclude lessors of real estate. Both the Canadian and the U.S. GDP estimates used here include lessors of real estate because these could not be removed from the U.S. data.

^24.

When mixed income is used to separate self-employed hours worked for sole proprietorships and partnerships in the unincorporated sector, the results are similar (see footnote 22). GDP per hour worked for sole proprietorships is slightly higher—about 0.8 of a percentage point from 1998 ($19.1) to 2005 ($27.4). GDP per hour worked in the rest of the business sector falls by about 0.2 of a percentage point per year from 1998 to 2005.

^25.

The wage bill of sole proprietorships as a percentage of the business sector was very small—about 2.1% in 2005, based on IRS Table 10 and BEA Table 1.13.

^26.

When mixed income is used to separate self-employed hours worked for sole proprietorships and partnerships in the non-corporate sector, GDP per worker for sole proprietorships is overestimated (see footnote 22). GDP per hour worked for sole proprietorships is substantially higher—about 36% to 68% from 1998 ($32.2) to 2005 ($53.0).

^27.

The PPPs are derived by means of a basic-price concept. Baldwin, Gu and Yan (2008) provided a 1999 benchmark for the PPP. The 1999 estimates were projected forward using the relative movements of the GDP deflators for Canada and the United States.

^28.

The U.S. employees of sole proprietorships, calculated as a share of the total number of employees, are obtained from the Statistics of U.S. Businesses (U.S. Census Bureau). For Canada, employment for 2005 was arrived at by means of an estimate based on employment figures obtained from the Survey of Employment, Payrolls and Hours (SEPH) (Economic Analysis Division of Statistics Canada). However, data are not available on employment of sole proprietorships for prior years; the growth of the number of T4 slips issued to employees of sole proprietorships was used instead.

^29.

According to the 1997 Statistics of U.S. Businesses, the average wage per worker for corporations and partnerships in the United States was US$30,286 that year, while the average wage per worker for sole proprietorships stood at US$14,463.

^30.

Using the U.S. Current Population Survey, Berger et al. (1999) showed that 69% of private non-agriculture workers aged 16 and over employed by small firms (firms with fewer than 10 employees) worked 35 hours per week or more. In larger firms employing 500 or more, 81% of employees worked 35 hours per week or more.

^31.

Paid employees account for more of the hours worked in U.S. sole proprietorships than in Canadian sole proprietorships, regardless of whether the estimate of hours worked is split according to the wage bill or employment.

^32.

Evaluating differences in tax policies between the two countries would require comparing the personal-tax rates and corporate-tax rates of the two countries and examining the different incentives to incorporate. For example, the United States offers a form of organization that is not corporate but nevertheless provides the limited liability protection of corporations—the Limited Liability Company (LLC)—which does not exist in Canada.

^33.

For a discussion of outsourcing in the United States, see Dey, Houseman, and Polivka (2007).