Section 5
Differences between surveys
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.
According to the Survey of Household Spending (SHS), about 4 in 10 Canadian households sent voluntary inter-household transfers in 2008. However, only 12% of Canadian households responding to Survey of Labour and Income Dynamics (SLID) reported sending voluntary inter-household transfers in the same period, and barely 7% of households reported receiving voluntary inter-household transfers. Why is there such a difference?
Table 3 shows the counts of households (economic families for Survey of Financial Security (SFS)) and the total dollar amount of voluntary inter-household transfers sent and received by Canadian households in 2008 (2004 data were used for 2005 SFS). In terms of the donors that sent voluntary inter-household transfers, SHS has about three times more households than SLID and SFS, but the total dollars sent were in the same order of magnitude for all three surveys. In terms of the recipients, SLID and SFS have similar number of households receiving voluntary inter-household transfers; the total amounts received by households in SLID are twice as much as those in SFS. However, the counts for SHS cannot be compiled because the information on receiving voluntary inter-household transfers cannot be differentiated from the category of 'other income'. In short, there are some discrepancies between the three surveys in the treatment of voluntary inter-household transfers.
Each of the three ways of measuring voluntary inter-household transfers serves the purposes of their respective surveys. For example, SHS data is primarily used to compile current household expenditure for the System of National Accounts and less used for the measurement of economic well-being compared with SLID. This affects the reporting requirements; the SHS requires detailed reporting on every expenditure item, whereas SLID is looking for an overall estimate.
Also, each survey uses dissimilar wording when describing voluntary inter-household transfers. For example, SHS measures "money gifts" and it allows that the transfers have no usage restrictions. SFS measures the money sent to support the living expenses of the recipients. SLID measures money sent or received, plus the regular payments paid on the recipient's behalf.
The surveys also define the scope of voluntary inter-household transfers differently. Whereas SHS requires that respondents state the amount of "money gifts" sent to people who are not household members, SLID and SFS measure the amount of money sent to or received from family members not living with the respondents. Both SLID and SFS exclude a greater portion of voluntary inter-household transfers beyond a family relationship compared to the SHS. The relevant sections on voluntary inter-household transfers are listed in appendices 1 to 4 for reference.
While this paper is unable to determine to what degree these differences have affected responses, it can still compare the reporting behaviours among the three surveys. The difference in reporting behaviour is shown in Chart 3. The median amount of voluntary inter-household transfers sent is $2,200 in SLID, $2,000 in SFS, and $500 in SHS. While the SLID curve resembles the SFS curve, the SHS curve is distinct. Households started reporting sending voluntary inter-household transfers at the 60th percentile for SHS and above 85thpercentile for SLID and SFS. The SHS curve then converges with both the SLID and the SFS curves at the 90th percentile. The difference between the SHS curve and the other two curves from the 60th to 90th percentile indicates that SHS had many more households reporting transfers, with many of them reporting small transfer amounts sent.
Two visible plateaus can be observed on the SHS curve, with the first one at the $500 level (77th to 80th percentile), and the second one at the $1,000 level (86th to 88th percentile). This might reflect rounding behaviour in reporting or respondents' preference for sending $500 and $1,000.
All three curves show that a small percentage of households send large dollar amounts for voluntary inter-household transfers; less than 5% of all households sent over $5,000 in the reference year. This makes the total dollar estimates on voluntary inter-household transfers measured very dependent on a few households. These households may be misreporting inheritances and bequests (usually large in value) as current transfers rather than as capital transfers. A maximum limit of $200,000 is set for inter-household transfers on the SLID questionnaire. Meanwhile, no such limitation exists on the SHS or SFS income processing.
Chart 3 Average monetary value of voluntary inter-household transfers sent by household percentiles
Lastly, this section examines the balance of voluntary inter-household transfers sent and received within Canada. In a closed economy, the dollar amount of volutary inter-household transfers sent should be identical to the amount received. However, as Houle and Schellenberg (2008) show many recent immigrants remit money to support their families living outside Canada. The purpose of Table 4 is to examine the balance of voluntary inter-household transfers for 2008 SLID and 2005 SFS, by the destination (or origin) of payments, if such information is available. Alimony and support payments are listed for comparison, as they are involuntary inter-household transfers and are transferred within Canada. In 2008 SLID, most voluntary inter-household transfers sent and received are within Canada.
As a whole, the total estimated amount of voluntary inter-household transfers received and sent is not the same. From 2008 SLID data, it was estimated that Canadian households received 79% of all voluntary inter-household transfers sent. Of voluntary inter-household transfers within Canada, the estimated transfer amount received is 87% of that sent, which is comparable to the 94% for alimony and child support received.1 Sampling errors and non-standard errors such as misreporting are plausible reasons for the imbalance in transfer estimates within Canada.
Measuring international remittances, Canadian households sent more transfers to families and friends overseas than they received. Canadian households received only 44% of the amount of voluntary inter-household transfers sent abroad. This is not surprising, given Canada's sizable population of immigrants with family links aboard, and the fact that international remittance tends to flow from high-income countries to low-income countries. A similar trend, using World Bank data, is observed for other high-income Organisation for Economic Co-operation and Development countries.2
A similar imbalance for voluntary inter-household transfers is observed for the 2005 SFS, but it has a lower received-to-sent ratio. This might result from the combination of oversampling of high-income respondents (high-income families tend to remit more in dollars), a smaller sample survey than SLID (usually associated with larger sampling errors), and the fact that, often, reported transfers made exceed reported transfers received (Mason et al. 2006). Information on the destination (or origin) of the transfer was not collected in the SFS.
Table 4 The balance of voluntary inter-household transfers and, alimony and support payments
Notes
- In SLID 2007, the percentage was 100% for all inter-household transfers (105% for within Canada, 73% for international).
- Source: World Bank. Migrant Remittances Data, July 2009. http://www.worldbank.org/prospects/migrationandremittances (accessed: January 15, 2010)
- Date modified: