The following describes the methodology used in the upcoming paper "Indicators of profit shifting by multinational enterprises operating in Canada" to be published on June 18, 2019. This methodology follows the one proposed by the Organization for Economic Cooperation and Development (OECD) in its Action 11 Report to create a dashboard of profit shifting indicators.
Section 1: Disconnect between financial and real economic activities
Discrepancies between financial and real economic activity within a country can be a sign that income is not reported, and therefore not taxed, where it was earned. The two indicators in this section rely on data aggregated to the country level and a list of countries with favourable corporate tax rates to explore these discrepancies.
Data sources for Indicators 1A & 1B: Data about real and financial economic activity is from Statistics Canada's Balance of Payments program, specifically the outward foreign direct investment statistics (NDM table 36-10-0008-01) and the activities of Canadian majority-owned affiliates abroad (NDM table 36-10-0470-01). We also used Gross Domestic Product (GDP) by country from the World Bank (World Development Indicators).
To identify jurisdictions with favourable corporate tax systems, we used data about foreign subsidiaries of Canadian corporations collected by the Canada Revenue Agency using the T1134 Information Return Relating to Controlled and Not- controlled Foreign Affiliates.
BEPS Indicator 1A: Mismatches between stocks of Canadian outward FDI and GDP of recipient countries for countries with favourable corporate income tax rates
This indicator is designed to gauge whether a significant proportion of Canada's outward FDI is driven by tax minimization. It compares Canada's outward FDI stock in 2016 to GDP of countries receiving the investments, for countries that have a favourable corporate income tax rates, and for those that do not.
Methodology: BEPS research often uses effective tax rates (ETRs) to identify jurisdictions with favourable corporate tax rates, because it is a measure of how much tax is actually paid for each dollar of income.Footnote 1 To determine whether countries had a favorable corporate income tax rate, we used data from the T1134 return from 2011 to 2016 to calculate the ETR (taxes divided by income)Footnote 2 for each foreign subsidiary. We weighted these ETRs by the subsidiary's assets to calculate the average ETR for each country over the period. We sorted countries by ETR, and labelled countries with the lowest- ETRs and up to 20% of the total assets held by foreign subsidiaries as having favourable corporate tax rates.
We also labelled countries with slightly higher ETRs and between 20% and 25% of total assets as having favourable corporate tax systems if they appeared on a list of countries compiled by the United States Government Accountability Office (2008) to study corporate tax evasion that is commonly cited.Footnote 3 Footnote 4
We collected data on the 10 countries with the largest stocks of Canadian outward FDI from Statistics Canada's FDI statistics program, and obtained GDP for each of these countries from the World Bank's World Development Indicatorstable.
We grouped countries according to whether they had favourable corporate tax rates, and added up the FDI and GDP for each group. We presented the sums for each group as a proportion of the total.
BEPS Indicator 1B: Mismatches between assets, employment and sales for countries with favourable corporate tax rates
This indicator looks for discrepancies between asset ownership, employment and sales in foreign subsidiaries of MNEs operating in Canada which may be associated with tax minimization. It compares the ratios of employees and sales to assets in subsidiaries operating in countries with favourable corporate income tax rates to those that are not.
Methodology: We collected data on total assets, employment and sales for the 10 countries where the subsidiaries of Canadian companies had the most assets from Statistics Canada's Activities of Canadian majority-owned affiliate's abroad program.
We grouped these countries using the list of countries with favorable corporate tax rates from Indicator 1A. We calculated averages of employment and sales per billion dollars' worth of assets.
Section 2: Profit rate differentials within MNEs
MNE affiliates with low tax rates that are highly profitable relative to their group overall may suggest that some of the group's income was transferred to minimize taxes. The two indicators in this section are recreated from Action 11 of the OECD's BEPS Action Plan using data about MNEs operating in Canada and their subsidiaries abroad.
Data sources for Indicators 2A & 2B: We used financial data about foreign affiliates from the T1134 Information Return Relating to Controlled and Not- controlled Foreign Affiliates. The variables were assets, income and taxes, and data was available for 2011 to 2016.Footnote 5
Financial data about the Canadian resident companies filling the T1134 return came from the T2 Corporation Income Tax Returns filed with the Canada Revenue Agency. The variables we used were assets, income and taxes for the same years where we had data from the T1134 available.Footnote 6
Combining these data sources allowed us to get a picture of a MNEs financials in Canada and abroad. A limitation was that financial data on foreign parents of the Canadian reporting company and other affiliates of these foreign parents was not available.Footnote 7
BEPS Indicator 2A: High profit rates of low-taxed affiliates of MNEs
This indicator compares income earned by MNE affiliates that are grouped based on how their profit rates and ETRs compare to their group. A high proportion of total income earned by affiliates with higher profit rates and lower effective tax rates than their groups is a sign of BEPS. This distribution of income suggests that MNEs could have strategically 'shifted' income to minimize taxes.
Methodology: The methodology for this indicator is adapted from Indicator 2 in Action 11 of the OECD's BEPS Action Plan. The OECD study used data from consolidated and unconsolidated financial statements of MNEs, and focussed on the largest 250 MNEs globally (OECD 2015).
Following this methodology, we calculated profit rate (income divided by assets) and ETR for each affiliate, and dropped observations with negative effective tax rates and income.
Next, we calculated profitability and ETR for each MNE group as a whole and compared the ratio of each affiliate to that of its group.Footnote 8 Based on these results, we sorted affiliates into the following groups (or quadrants):
Quadrant 1: Higher ETR, higher profitability,
Quadrant 2: Lower ETR, higher profitability
Quadrant 3: Lower ETR, lower profitability
Quadrant 4: Higher ETR, lower profitability
We then added up the income earned by the affiliates in each quadrant, and repeated the exercise for each year from 2011 to 2016. Since the results were variable year- over- year, we chose to present our result as an aggregate for the whole period.Footnote 9 Footnote 10
BEPS Indicator 2B: High profit rates of MNE affiliates in lower- tax locations
Indicator 2B compares the profit rates of affiliates within an MNE in low- tax areas to the profit rate of the MNE as a whole by calculating relative profit rates for each MNE, and combining these to create an average. If the indicator is above 1, profits are higher in low- tax areas than in high tax areas, meaning that MNEs may be shifting profits to minimize taxation.
Methodology: The methodology for this indicator is adapted from Indicator 3 in Action 11 of the OECD's BEPS Action Plan.
Following this methodology, we added up assets, income and taxes paid by affiliates to arrive at totals by country for each MNE group. We calculated the ETR by country for each MNE group by dividing the total tax paid by total income. Next, we ranked countries within each MNE by their ETR, and labelled the countries with the lowest ETRs and up to 20% of each MNEs assets as 'low-tax'.Footnote 11
Similarly, we calculated the profit rate by country for each MNE group by dividing the total income by total assets. We weighted profit rates for each affiliate in a 'low-tax' country by the total assets in that country and combined them into an average profit rate for each MNEs 'low-tax' affiliates.
Next, we added up assets and income by affiliates to arrive at totals for each MNE group, and calculated the overall profit rate for each MNE by dividing the total income by total assets. To obtain relative profit rates for each MNE, we divided the profit rate for the low- tax countries where the group had affiliates by the profit rate for the MNE as a whole.
Finally, we weighted each relative profit rate by the MNEs assets and combined these to obtain an overall average relative profit rate.Footnote 12 We repeated the exercise for each year from 2011 to 2016. As with Indicator 2A, results were variable year- over- year and we chose to present them for the period as a whole.Footnote13
We also calculated a weighted overall average relative profit rate for the 25% of MNEs with the highest relative profitability.
Section 3: MNE vs 'comparable' non-MNE effective tax rate differentials
BEPS Indicator 3: Effective tax rates of MNEs relative to non-MNE entities with similar characteristics
Unlike other BEPS indicators which focus on activities that reduce net income in countries where tax rates are high, Indicator 3 is designed to detect whether MNEs take advantage of their capacity to shift income to other jurisdictions, to minimize their tax rate. Lower ETRs for MNEs may also reflect non-BEPS behaviours such as the decision to carry out substiantial activity to benefit from certain preferential tax treatments (e.g. R&D tax subsidies, investment tax credits).
The indicator compares ETRs for two types of MNEs operating in Canada to ETRs for comparable enterprises without subsidiaries or parents abroad.
Data sources: We used consolidated financial statements of enterprises operating in Canada from the Annual Financial Taxation Statistics program, specifically the financial variables total assets, net income before taxes, and federal tax payable for 2011 to 2016. These data are consolidated at the enterprise level and cover activities that are 'booked in Canada'.
We also used a new 'flag' to identify MNEsFootnote 14 which was developed in collaboration with Statistics Canada's International Accounts and Trade Division. It relies on tax data to identify MNEs with affiliates abroad, and the data collected under the Corporate Returns Act to identify enterprises with foreign parents.
Methodology: The methodology for this indicator is adapted from Indicator 4 in Action 11 of the OECD's BEPS Action Plan. The OECD study used data from unconsolidated financial statements of MNEs and non-MNEs worldwide (OECD 2015).
To recreate this indicator for Canada, we set up the data for analysis by labelling enterprises as 'large' if they had over 25 million in total assetsFootnote 15. Following OECD methodology, we dropped all observations with negative total assets, net income, and taxes.
We calculated ETRFootnote 16 and profit rate for each enterprise, and sorted enterprises into six groups based on their characteristics: non-MNEs or enterprises without affiliates abroad (large and small), and two types of MNEs: Canadian- owned enterprises with one or more foreign affiliates (large and small) and foreign- owned enterprises (large and small). To discover whether differences between these 6 groups were statistically significant, we ran a regression that controlled for industry and profitability.Footnote 17
We present our findings as the differential between ETRs for comparable MNEs and non-MNEs for 2016, because results were stable year-over-year.