May 4, 2026

---

This is a companion piece to the article entitled Part 1 – Results from delivering CVITP service to clients (2025 update).

Individuals assisted

Data discontinuity:  Prior to 2021, I used several CRA data sources as the CRA did not offer comprehensive CVITP data in any one publication.  In the spring of 2023, the CRA started providing data from 2021 onwards on its CVITP webpages.  The new data uses a different annual time frame for collecting and reporting on the CVITP from the time frame the CRA previously used.  This creates a discontinuity between the data I cite before 2021 and the new data.  However, I do not believe the new data time frame creates a significant change in the trends observed.  Therefore, I have replaced the old data I cited by the new data starting with the year 2021. (For more information, see Great CRA Innovations: Annual Data on CVITP Results.) 

Absence of client information:  As I indicated in the main article, the CRA does not provide any details of client characteristics such as age, income levels or household structure.  As a result, it is impossible to say anything about the clients’ poverty profiles.  Nor does the CRA provide any data on new versus returning CVITP clients.  Although the CRA uses the COIN to identify returns filed within the CVITP, it presently has no capacity to take the information contained in the returns and analyze this separately.

Returns filed

Data discontinuity:  The issue is the same as for the data for “individuals assisted”.  Please see above for the explanation.

Absence of information on returns by tax year:  The CRA does not provide any information on the distribution of returns between various tax years or between CVITP clients in any given tax season.

Assumption about average returns filed:  I generally assume that all CVITP clients get their 2025 return filed and that the remaining prior years are distributed amongst all clients.  As a result, the average or mean number of returns per client represents a figure which, in reality, does not exist as it includes a fraction of one whereas returns are simply filed or not filed.  The average number of returns filed per client will be different from the median number which is very likely to be one.[i]  The prior year returns are probably clustered among a small number of clients.

Value generated

Data discontinuity:  The CRA only started listing the value generated on its CVITP statistics beginning with the 2021 tax season.  Data has been identified in various CRA publications for 2018 and 2019.  The CRA has not published data for 2020.

Assumption about calculation methodology:  I assume that the CRA used the same methodology for calculating its benefits throughout the entire period.

Assumption about average value:  The median value per individual is lower than the average value per individual.  This is because the value of all returns filed is used to calculate the average value.  However, as I have indicated above, the median number of returns filed per individual is probably one.  This means that most people file only one return.  So, the average value generated per individual will be above the median value.  However, it is impossible to say what the size of the difference is because the CRA provides no information on the distribution of returns on client family structure inclusive of children as the Canada Child Benefit probably has the largest impact on the value generated for households.

Absence of benefit details:  The CRA provides no breakdown on the specific federal and provincial benefits and credits administered by the CRA.  Therefore, it is impossible to highlight the year-to-year variations in the amounts generated by specific benefits and credits.

Assumption about refunds:  Refunds generally represent a small portion of the value generated as compared with the credits and benefits.  I assume that portion does not vary substantially from year to year.

Data only on CRA administered benefits:  The CRA provides information only on the benefits it administers, not on those administered by Service Canada, including the Guaranteed Income Supplement.  No information is provided on the indirect federal benefits, such as the Canada Disability Benefit, the Canada Learning Bond and the Canadian Dental Care Plan, or on the indirect provincial benefits, such as social housing rent subsidies, which depend are conditional on the client being current with the filing of their returns.

---

[i] This is a reminder for those who don’t remember their high school statistics lesson on the difference between average and median: the median represents the middle value in a data set, where 50% of the observations lie below and 50% lie above this value.  In this instance, at least 50% of the clients (and likely more) are getting only one return (the current year) prepared.