Break in the Agriculture–Population linkage net farm income data series

The Agriculture–Population linkage database provides a socio-economic profile of the farm population, farm operators, farm households and farm families. Income data for the Agriculture–Population linkage database is derived by linking the Census of Population long form (20% of the population) to farm operators on the Census of Agriculture. Weighted estimates are generated for the farm population, operators, households and families.

The proportion of net farm income that contributes to farm operator and farm family income has been steadily declining for a number of censuses. The shift between 2000 and 2005 (the income-reporting years for the 2001 and 2006 Censuses) has been particularly significant. Between 2000 and 2005, net farm income dropped to 8.1% of farm operator income on unincorporated farms, from 22.8% in 2000 (Table 1). Similarly, for farm families on unincorporated farms, the contribution of net farm income to total family income declined to 6.3% in 2005 from 17.5% in 2000 (Table 2).

The 2006 Agriculture–Population linkage database marks a break in comparability of the net farm income series for 2006 and previous Agriculture–Population linkages. Two factors have contributed to this break.

1. The major factor is the shift from respondent-supplied income data to a combination of data sourced from tax files and respondents. For the first time, the 2006 Census of Population offered respondents the choice of completing the income questions on the questionnaire or allowing their income tax data to be used instead. Income data was drawn from 2005 tax data for 79% of farm operators in the 2006 Census.

Data sourced from income tax files has many advantages over data provided directly by respondents, including increased precision and standardized inclusions and exclusions as defined by the Canada Revenue Agency (CRA). This shift, while improving data quality and consistency with the tax system, has introduced a conceptual change that limits comparability to preceding datasets. These limiting factors will be addressed in Section 1.

2. The second contributing factor to the reduced comparability of the income series is the gradual erosion of net farm income as a useful measure of a farm's contribution to total family income on unincorporated operations. The factors behind this trend will be discussed in Section 2.

Section 1
Shift to income tax data for the 2006 Census

For Censuses of Population prior to 2006, income data were supplied directly by the respondent. The 2006 Census gave respondents the option of allowing their income tax data to be used in lieu of filling in the questionnaire. Income data for 79% of farm operators were, therefore, drawn from 2005 tax data in the 2006 Census for the first time.

The move to tax data for income highlighted a number of conceptual reporting issues: it increased overall data precision (Section 1a), increased the frequency of reporting net farm income losses (Section 1b), standardized inclusion of capital cost allowance/depreciation in the calculation of net farm income (Section 1c), reflected the effect of tax laws (Section 1d), and generally improved the consistency of the response (Section 1e). The differences introduced by using tax data have resulted in a break in the data series.

Top of Page

Section 1a
Increase in overall precision

The data using tax files in 2006 are more precise than the data provided by respondents in 2001, which was often estimated or rounded, leading to peaks at $10,000 intervals (Figure 1).

Section 1b
Increased frequency in reporting net farm income losses

A higher proportion of farm operators reported negative net farm income in the 2006 Census compared to previous censuses.

Table 3 shows how the shift to using mostly income tax data in 2006 has reduced the percentage of farm operators reporting zero net farm income, while boosting the percentage reporting values less than zero.

Section 1c
Standardized inclusion of capital cost allowance/depreciation

More negative figures were reported after the shift to data derived from tax files for a number of reasons. One of the greatest factors is likely the automatic inclusion of capital cost allowance (CCA) in tax data so that net farm income figures reflect deductions for CCA/depreciation.

Filers who submit and claim the CCA/depreciation deduction can benefit from lower taxes. CCA/depreciation claimed by taxfilers is automatically incorporated in data drawn from tax files.

The Census Guide 2B (available on the Internet) specified that CCA/depreciation should be subtracted from gross farm receipts; however, CCA/depreciation is not mentioned explicitly on the questionnaire. Moreover, the Census of Agriculture questionnaire specifically instructs respondents to exclude depreciation from the operating expenses reported for the farm business. These different concepts may have affected the reporting of CCA/depreciation on the Census of Population: it appears that, prior to the 2006 Census, respondents were not often including CCA/depreciation in net farm income.

Top of Page

Section 1d
The effect of tax laws

Data from tax files are shaped by tax laws. One of the peaks in the 2005 net farm income distribution falls at -$8,750, the maximum amount for unincorporated farm losses that can be written off for tax purposes against other income when farming is not the major source of income. Peaks in Figure 1 at values between -8,750 and zero indicate that the deductible amount has been divided between operators in farm partnerships. Partners are eligible only for a proportion of the total loss for the whole farm.

Section 1d
Increased consistency

While there are a few different means of accounting (cash and accrual), most significant inclusions and exclusions will be reported consistently on income tax returns. Tax laws are well defined and the audit of returns provides an enhanced external data validation check.

Section 2
Erosion of net farm income as a useful indicator of unincorporated farm income

Another contributing factor to the reduced comparability of the income series is the gradual erosion of net farm income as a useful measure of a farm's contribution to total family income on unincorporated operations. Two principal reasons are behind the erosion: the increasing proportion of incorporated farms (Section 2a) and shifts in the way income is distributed from the farm to family members on unincorporated farms (Section 2b).

Section 2a
Increased incorporation

Incorporated farms pay wages and salaries, rent and/or dividends to disburse funds from the farm to operators and farm families—they do not disburse funds as net farm income. Between 2001 and 2006 the percentage of incorporated farms increased from 13.4% to 16.0%, according to the Census of Agriculture. While relatively few in number (though increasing), incorporated farms accounted for 54.5% of gross farm receipts in 2005, up from 46.2% in 2000 (Table 4).

Unincorporated farms may disburse funds to the farm operator and family as net farm income, wages and salaries or rent.

Top of Page

Section 2b
Wages and salaries paid to family

In order to minimize tax, small businesses can benefit by paying wages and salaries to family members for work performed for the business. These funds are usually subject to much lower taxation rates than if they had been held by the farm and taxed as net farm income in the hands of the farmer.

This tax planning strategy is well known and promoted. According to the Ontario Government Ministry of Agriculture, Food and Rural Affairs fact sheet on the subject, "Paying wages to family members involved in an agricultural business is an excellent way of recognizing their efforts, instilling a sense of participation and formalizing their role within the family business. Farm families may also be able to boost their after-tax family income by paying farm business wages or salaries to family members…".

Conclusions

A shift from respondent-supplied income data to a combination of data from respondents and income tax sources has, while improving data quality and consistency with the tax system, precipitated a break in comparability of the net farm income data series between 2006 and preceding censuses.

As the industry reorganizes over the longer-term and income disbursement trends change, net farm income data from the Agriculture–Population linkage database becomes increasingly limited in characterizing the farm's economic benefit to families. However, total family income remains a valid means of comparing the relative economic wellbeing of farm families with those in other population groups.

Table 1 Components of farm operator income on unincorporated farms

Table 2 Components of farm family income on unincorporated farms

Table 3 Percentage of farm operators reporting net farm income less than, equal to or greater than zero, 2000 and 2005

Table 4 Percentage of farm cash receipts by incorporation status, 1986 to 2006

Top of Page

Figure 1
Net farm income distribution, farm operators, 2006 and 2001 Censuses