Skip to content

Literary funding · Canada · 2024–25 annual snapshot

Canada’s literary-grant average changes when you ask who the award row belongs to

The pooled mean is CAD 35,701.89; the 417 Individual award rows average CAD 17,662.11. Recipient classes and core/project mixtures change what the benchmark describes.

Published
Share
Share this page
EmailCite & reuse this study
Key findings as text
Reusable imagesCanada’s literary-grant average changes when you ask who the award row belongs to — 818: Published Literature/Grant award rows. 2024–25 national sheet only; not unique people, payments or books.Download sharing image (JPG) ↓

Key findings

818

Published Literature/Grant award rows

2024–25 national sheet only; not unique people, payments or books.

CAD 35,701.89

Mean across the pooled selection

CAD 29,204,150 recommended; median CAD 25,000.

CAD 17,662.11

Mean of 417 Individual award rows

Median CAD 24,000; all are Project Grant rows in this fixed selection.

32.15% / 63.84%

Core grants: share of rows /recommended amount

263 Organization/Core rows; composition, not a causal or fairness finding.

The result an applicant can actually use

The mean recommended amount across 818 published Literature/Grant rows in Canada Council for the Arts’ 2024–25 national workbook is CAD 35,701.89. Within the same fixed selection, the 417 rows classified Individual have a mean of CAD 17,662.11 and a median of CAD 24,000. These are different descriptions of the same published award table. The pooled figure combines individuals, groups and organizations, and combines project support with organizational core support. An individual deciding which historical comparison to examine should first match the recipient and funding categories, rather than treat the largest available average as a personal funding benchmark.

The recommended amounts sum to CAD 29,204,150 in this selection. Individual rows are 50.98% of its award rows but 25.22% of its recommended-amount sum. Organization rows are 48.41% of rows and 74.42% of that sum. The remaining five Group rows account for 0.61% and 0.36%, respectively. These contrasts describe the composition of an administrative table. They do not establish the number of distinct people supported, the fairness of allocation, a payment received or the amount any author earned.

This is an original grouped calculation from one exact official workbook, accompanied by an offline recipe. It is not a new survey, an application-success study or a ranking of named recipients. The central question is practical: how much does the meaning of a literary-funding average change when the award row’s recipient class and funding type are made explicit?

A literary field is not a list of writers

The selection uses three literal source fields: fiscal year 2024–25, support type Grant and field of practice Literature. The national worksheet is used once. Its provincial and other geographic worksheets are not added again, because that would duplicate the national records. The published national sheet has 6,546 data rows; 818 meet the fixed selection, 171 have a different support type, and 5,557 remaining rows have a different field. Those counts reconcile the source frame without publishing a recipient list.

“Literature” is the administrative field value used here. It is not a promise that every selected award funded the writing of one book, nor that the table contains every kind of literary activity financed in Canada. Individual is the source’s recipient class; Organization and Group are its other classes. The analysis does not infer occupation, gender, language, cultural background or artistic genre from a recipient’s name. It does not reclassify publishers, festivals or authors by reading their public profiles. That restraint keeps the result reproducible without adding undocumented judgments about recipients.

The workbook also distinguishes core and project grants. All 263 selected core-grant rows are in the Organization class. The 555 project-grant rows comprise 417 Individual, five Group and 133 Organization rows. Thus the contrast between recipient classes overlaps substantially with the contrast between support structures. It cannot isolate a recipient-class effect as though the programme mix were held constant.

The complete cross-tabulation also retains the Prizes and Fellowships category, although it contains no row in the fixed selection. That zero is not a statement about whether prizes or fellowships exist in Canada. It concerns only the combination of the filters used here with the selected categories. Readers should therefore neither fill an empty cell with a missing amount of money nor read it as proof that a whole funding area is absent. The visible classification is what makes the limited scope checkable.

Why the pooled mean reaches CAD 35,701.89

The 263 core-grant rows sum to CAD 18,643,076. Their mean recommended amount is CAD 70,886.22 and their median is CAD 58,000. The 555 project-grant rows sum to CAD 10,561,074, with a mean of CAD 19,028.96 and a median of CAD 23,500. Core rows therefore represent 32.15% of selected award rows and 63.84% of selected recommended amounts. Project rows supply the complementary 67.85% of rows and 36.16% of amounts.

An overall mean must weight these groups by their numbers of rows. It is not the midpoint between CAD 70,886.22 and CAD 19,028.96, and it is not the mean of the three recipient-class means. Dividing every cell’s recommended-amount sum by the common denominator of 818 gives an additive accounting of the pooled mean. The Organization/Core cell contributes CAD 22,791.05, Individual/Project CAD 9,003.79, Organization/Project CAD 3,778.39 and Group/Project CAD 128.67. These four contributions add to CAD 35,701.89, subject only to displayed rounding.

The word “contribution” here has a precise arithmetic meaning: cell sum divided by the pooled row count. It is not a causal estimate of what core funding does to an individual award. Removing a category would change both the numerator and the denominator, so the contribution is not the difference between the published mean and a hypothetical policy scenario. The exact fractional values are retained in the downloadable aggregate JSON for readers who want to verify the identity without rounding error.

This difference is decisive for reading the second figure. Its long bar for Organization/Core does not show an average grant within that cell. It shows the cell's sum, spread over all 818 selected rows. A small contribution can therefore arise from few rows without the individual recommended amounts being small. For example, the five Group/Project rows have their own mean of CAD 21,050 but contribute only CAD 128.67 to the pooled mean. Both follow from different denominators.

The useful comparison inside project support

Restricting attention to project grants removes the core-versus-project mixture, but it still leaves different recipient categories. Individual project rows have a mean of CAD 17,662.11 and median of CAD 24,000. Organization project rows have a mean of CAD 23,238.53 and median of CAD 20,000. The five Group project rows have a mean of CAD 21,050 and median of CAD 29,900. A project-only aggregate is consequently still not a stand-in for the typical individual award row.

The median ordering differs from the mean ordering for Individual and Organization project rows. That is useful evidence against reading the mean as an automatic description of the middle award. It does not by itself identify the shape of either distribution, the programme responsible for a difference or how many projects had any particular amount. The public bundle deliberately contains grouped statistics, not the underlying award-value vectors. Readers seeking distributional questions that require a histogram or individual record matching must undertake their own authorized source analysis.

The Group category is retained rather than absorbed into another class. Its five rows make the small size of the category visible. They are too narrow a basis for a general statement about groups applying for literary funding, especially because unsuccessful applications and eligibility are outside this frame. The table reports the finite published selection; it does not attach a confidence interval or imply that a sample was randomly drawn from all potential applicants.

Mean and median answer different questions

The pooled mean is the sum divided by the number of award rows. The median is the midpoint of the sorted award values; when the number of rows is even, this analysis averages the two middle values. For the 818-row selection, the median is CAD 25,000. Both measures are correct, but they answer different questions. The mean preserves the relationship between aggregate amounts and row count, whereas the median describes the middle position in the ordered published rows.

For the 417 Individual rows, the mean is below the CAD 24,000 median. A casual “large grants always pull the average above the middle” explanation would therefore be wrong for this subgroup. The calculation alone does not establish why the measures differ. A defensible explanation would require a separately defined examination of programme rules, amounts and perhaps project timing. The present result is a reason to show both measures, rather than reverse-engineer an institutional story from their ordering.

Likewise, comparing a mean from one recipient class with a median from another would confound the grouping choice and the summary measure. Our first aggregate table shows counts, sums, means and medians together. The full cross-classification preserves the three-by-three recipient/funding arrangement, including empty combinations. A zero row count has no defined mean or median. An empty field in the downloadable CSV represents that undefined quantity, not a zero-dollar award.

Recommended amounts are not cash in an author’s account

The official methodology describes a year-end administrative snapshot and warns that subsequent adjustments or cancellations are not reflected. Amounts are recommended amounts, and conditions can remain attached. Those definitions matter before any comparison to sales royalties, salary or an author’s annual income. This analysis supplies no evidence of when money was paid, how it was spent, whether it covered multiple years or what proportion became personal compensation.

The provider reports amounts above CAD 5 and excludes several forms of support, including Public Lending Right and Access Support. The source frame is therefore a published, bounded table, not every financial transaction or every support programme. One recipient can appear more than once; a co-recipient arrangement is not a complete list of everyone involved. Counting 417 Individual award rows as 417 different writers would silently change the unit. No unique-person total is calculated here.

A grant awarded to an organization also need not be income retained by that organization’s authors. Organizational operations can support a wider literary environment, but the row does not reveal that allocation. Conversely, the fact that an individual row is classified Literature does not make its recommended amount equivalent to a book advance. Neither the budget of a publication nor an author’s net earnings can be recovered from these columns alone.

What applicants, editors and researchers should do with it

For an applicant, the first useful action is to identify the current programme’s eligibility, funding type and allowable costs using the Council’s current official materials. This historical table can then provide a descriptive comparison for the relevant published category. It cannot predict an offer, establish an entitlement, indicate a current ceiling or replace the current application rules. There is no unsuccessful-application denominator here, so even a detailed award table cannot produce an application-success probability.

For a journalist, the strongest defensible angle is the mixture inside a frequently tempting headline number. A precise statement is that the pooled mean is CAD 35,701.89 across 818 selected published award rows, while the Individual subgroup mean is CAD 17,662.11 across 417 rows. The period, field, support filter, award-row unit and “recommended” qualification belong with the figures. Dropping those qualifiers changes the claim. Calling this the average Canadian writer’s grant or income would go beyond the evidence.

For a researcher, the useful contribution is the transparent decomposition and category audit. It is possible to verify how the pooled mean is assembled without republishing names, locations or all selected awards. Future questions about regional access, repeat recipients, application success or longitudinal change would need new protocols and suitable denominators. They are not automatically answered by the availability of more columns or earlier workbooks, and this single-year calculation should not be presented as a trend.

What this adds to existing official reporting

The Council already publishes annual funding summaries, programme reports and competition results. This analysis does not claim to discover its grant programmes or replace those reports. The contribution is a bounded, reproducible view of one Literature/Grant selection: recipient-class and funding-type summaries are placed alongside a complete cross-classification, and the pooled mean is reconstructed from their sums using the same denominator. That lets a reader see exactly why a national all-category average differs from a narrower applicant-relevant benchmark.

Different official reports can use different programme boundaries. A competition report for a particular part of Explore and Create is not the same frame as all Literature/Grant rows in the annual national workbook. Its application counts cannot be attached to this selection to manufacture a success rate. This publication uses only the fixed annual workbook for numerical results and treats other reporting as context. It does not claim that these familiar summary statistics are a new statistical invention or that nobody has previously studied grant allocations.

Reproduce the aggregates without redistributing the source

The public recipe requires a user-supplied copy of the exact official workbook, validated by SHA-256 and the expected selection fields. It makes no network requests and does not supply, cache or proxy the workbook. A changed fingerprint, category or amount representation stops the calculation for source review. The script emits only an aggregate JSON file and an aggregate CSV containing counts, sums, means and medians; its grouped shares and additive contributions are calculated from those counts and sums. It has no source-row export function.

The source workbook, recipient-level extracts, names, addresses, business identifiers, geographic classifications, source prose and source images are absent from the public bundle. Canada Council for the Arts is credited as the data source; the calculations, writing and chart geometry are original work by Alisson Bento. The Council’s source terms remain separate from the original analysis and code. No broad commercial source licence, official endorsement or permission to redistribute the workbook is asserted. Users obtaining an input must follow the applicable source conditions themselves.

This analysis follows an earlier private exploratory calculation. The original question, selection and metrics were fixed before that calculation, but the later aggregate-only public protocol was written after its results were known. The narrower output design must not be described as a blind preregistration. It changes what is distributed, not the selected award frame or the arithmetic. The downloadable method states those dates and preserves the difference between a verifiable factual summary and a recipient-level data service.

Recipient class changes the benchmark

Means and medians for the three source recipient classes, on one zero-based CAD scale.

Full chart and source
Individual: mean 17,662.11, median 24,000; Group: 21,050 and 29,900; Organization: 54,883.33 and 40,000 CAD. Three separate recipient classes in 818 published award rows.

Scroll horizontally to inspect the chart. ↔

Chart and values

Scroll horizontally to see all columns.

Recipient class: the same fixed Literature/Grant selection
Recipient classAward rowsRecommended sum CADMean CADMedian CAD
Individual4177,365,100.0017,662.11031224,000.000000
Group5105,250.0021,050.00000029,900.000000
Organization39621,733,800.0054,883.33333340,000.000000

CAD recommended amounts. Award rows, not unique recipients or paid cash. Means/medians rounded to two decimals in the article; six decimals here.

Data source: Canada Council for the Arts · 2024–25 Literature/Grant. Original aggregate analysis: Alisson Bento. Recommended amounts, not paid cash; award rows, not unique people.

Download chart (SVG) ↓ · Download chart (PNG) ↓

Where the pooled mean comes from

Four nonempty recipient/funding cells reconstruct the pooled mean using the common denominator of 818 award rows.

Full chart and source
Organization/Core contributes 22,791.05, Individual/Project 9,003.79, Organization/Project 3,778.39 and Group/Project 128.67 CAD to the 35,701.89 pooled mean. These are additive accounting contributions, not causal effects. Displayed contributions are rounded; exact fractions sum to the pooled mean.

Scroll horizontally to inspect the chart. ↔

Chart and values

Scroll horizontally to see all columns.

The full recipient/funding mixture
Recipient classFunding typeAward rowsRecommended sum CADMean CADMedian CADContribution to pooled mean CAD
IndividualCore Grant00.00——0.000000
IndividualProject Grant4177,365,100.0017,662.11031224,000.0000009,003.789731
IndividualPrizes and Fellowships00.00——0.000000
GroupCore Grant00.00——0.000000
GroupProject Grant5105,250.0021,050.00000029,900.000000128.667482
GroupPrizes and Fellowships00.00——0.000000
OrganizationCore Grant26318,643,076.0070,886.22053258,000.00000022,791.046455
OrganizationProject Grant1333,090,724.0023,238.52631620,000.0000003,778.391198
OrganizationPrizes and Fellowships00.00——0.000000

Contribution = cell sum /818, not a within-cell mean or causal estimate. Empty cells retained.

Data source: Canada Council for the Arts · 2024–25 Literature/Grant. Original aggregate analysis: Alisson Bento. Recommended amounts, not paid cash; award rows, not unique people. Displayed contributions are rounded; exact fractions sum to the pooled mean.

Download chart (SVG) ↓ · Download chart (PNG) ↓

Data & downloads

Scroll horizontally to see all columns.

Recipient class: the same fixed Literature/Grant selection
Recipient classAward rowsRecommended sum CADMean CADMedian CAD
Individual4177,365,100.0017,662.11031224,000.000000
Group5105,250.0021,050.00000029,900.000000
Organization39621,733,800.0054,883.33333340,000.000000

CAD recommended amounts. Award rows, not unique recipients or paid cash. Means/medians rounded to two decimals in the article; six decimals here.

Scroll horizontally to see all columns.

Funding type: do not average the subgroup averages
Funding typeAward rowsRecommended sum CADMean CADMedian CAD
Core Grant26318,643,076.0070,886.22053258,000.000000
Project Grant55510,561,074.0019,028.96216223,500.000000
Prizes and Fellowships00.00——

All categories retained. An empty cell has no defined mean/median.

Scroll horizontally to see all columns.

The full recipient/funding mixture
Recipient classFunding typeAward rowsRecommended sum CADMean CADMedian CADContribution to pooled mean CAD
IndividualCore Grant00.00——0.000000
IndividualProject Grant4177,365,100.0017,662.11031224,000.0000009,003.789731
IndividualPrizes and Fellowships00.00——0.000000
GroupCore Grant00.00——0.000000
GroupProject Grant5105,250.0021,050.00000029,900.000000128.667482
GroupPrizes and Fellowships00.00——0.000000
OrganizationCore Grant26318,643,076.0070,886.22053258,000.00000022,791.046455
OrganizationProject Grant1333,090,724.0023,238.52631620,000.0000003,778.391198
OrganizationPrizes and Fellowships00.00——0.000000

Contribution = cell sum /818, not a within-cell mean or causal estimate. Empty cells retained.

Evidence & method

How we calculated it

  1. Use the national sheet once, with literal year 2024-25, support Grant and field Literature. Do not append provincial copies. Use only the seven selection/grouping/amount columns for the analysis. Individual, Group and Organization are source recipient classes, not inferred occupations or demographics. Core Grant, Project Grant and Prizes and Fellowships are funding classes. Retain all nine cross-cells; a cell with zero rows has null mean and median.
  2. Amounts are parsed as exact CAD cents. Mean = cell sum / cell count. Median is the middle sorted value, averaging the two central values for even N. Row share = cell count /818; amount share = cell sum /29,204,150. Additive contribution to the pooled mean = cell sum /818. Exact fractions are included to separate arithmetic from six-decimal presentation. No sampling intervals or causal models are applied to this finite published frame.
  3. The source describes recommended amounts as at March 31. Later adjustments/cancellations, payment, recipient earnings, unique-person counts, book outputs and unsuccessful applications are not established. The provider publishes amounts above CAD 5 and excludes Access Support, Public Lending Right and some initiatives. No current eligibility or grant prediction follows from this historical selection.
  4. The original private question/metrics were fixed before selected metrics were calculated. This later public aggregate-only protocol was recorded after the private results were known. It narrows distribution, not the selection. No claim of blind preregistration or human peer review is made.
  5. The public recipe takes a user-supplied, authorized exact XLSX and requires no network or third-party packages. It emits only results.json and aggregate-summary.csv. No source workbook, award-row vectors, names, addresses, identifiers, source coordinates, geography, screenshots, source prose or agency images are distributed. Grouped source extrema and five-largest-value metrics in the earlier private pilot are omitted from this public version.
  6. Source compilation/expression rights remain with their holders. This bundle supplies original grouped facts, original expression and original code only; it does not confer a commercial licence to the workbook, warrant worldwide clearance or imply Council affiliation/endorsement. Users must obtain their own inputs through authorized routes and follow applicable source terms.

What the data cannot tell us

  • One 2024–25 snapshot, with recommended amounts as at March 31; not verified payments, author earnings, a time trend or current programme rules.
  • Award rows are not unique people or books. Repeated recipients and co-recipient arrangements prevent person-count inference.
  • Literal Literature/Grant selection excludes other support/fields, Public Lending Right, Access Support and certain initiatives; source amounts are above CAD 5.
  • Core versus project composition overlaps recipient classes; no recipient-class causal effect, fairness assessment, success probability or demographic inference.
  • Only five Group rows; finite-frame summaries do not generalize to applicants. Empty cells have undefined means and medians.
  • No source workbook, recipient-row extract, identifiers, source prose/images or geography is redistributed. Source rights remain separate; no blanket commercial source licence or Council endorsement.

Original sources

  1. Canada Council for the Arts: data tables

    Official annual source listing; not a blanket commercial source licence.

    Accessed October 31, 2026

  2. Open Data Table 2024–2025

    Exact private input for original aggregate calculations; not redistributed.

    Accessed October 31, 2026

  3. Open Data Tables: Methodology 2024–2025

    Recommended amount, year-end snapshot, repeat awards, exclusions and publication threshold; own concise paraphrase only.

    Accessed October 31, 2026

  4. Canada Council copyright notice

    Source expression and compilation rights remain separate; no broad commercial input-redistribution permission claimed.

    Accessed October 31, 2026

Cite & reuse this study

Alisson Bento. 2026. Canada’s literary-grant average changes when you ask who the award row belongs to. Original aggregate analysis of Canada Council for the Arts 2024–25 published grant data. https://alissonbento.com/research/canada-literary-grant-mixture/

Data source: Canada Council for the Arts. Original aggregate calculations, writing, code and chart design: Alisson Bento. Retain source, year, recommended-amount and award-row qualifications. Source compilation/expression rights remain separate; no official endorsement or input-redistribution permission is asserted.

Corrections or editorial questions

Pass it on.

Canada’s literary-grant average changes when you ask who the award row belongs to | Alisson Bentoalissonbento.com

WhatsAppFacebookXTelegramRedditThreadsEmailSMS