Frequently Asked Questions
The basics
What is this data actually showing?
This tool tracks how large charities in England and Wales (those with annual income over £500,000) fund themselves, and whether that funding position changes over time. It uses five years of official Charity Commission data (2020–2024) covering the same organisations across all five years.
The core idea is simple: where a charity gets its money from shapes what it can and can't do. An organisation that relies almost entirely on government contracts faces very different pressures from one that lives off investment income or public donations. This tool makes those differences visible and comparable across thousands of organisations at once.
Why does it only cover large charities?
Charities above £500,000 income are required to submit detailed financial returns to the Charity Commission under a standard called SORP. Below that threshold, the reporting is less detailed and less consistent, making meaningful comparison across organisations unreliable.
There is also a theoretical reason. Organisations above this size typically have the professional infrastructure, e.g. dedicated finance staff, strategic planning processes, to make deliberate choices about their funding mix. Smaller organisations often take whatever funding is available. The £500k threshold identifies the population where funding strategy is a meaningful concept.
What are the four funding streams?
Every charity's income is broken down into four streams:
- Donations
- Donations, legacies, and grants from trusts and foundations
- Government
- Government contracts and government grants
- Investment
- Returns on invested assets and endowments
- Market
- Fees for charitable services and trading income
Classifying organisations
What is an archetype and how is it assigned?
An archetype is a four-character label that summarises an organisation's funding position. Each character represents one of the four funding streams, and it is either H (high) or L (low) depending on whether that organisation relies on that stream more or less than the typical large charity.
An organisation labelled LHLH has low donation income, high government income, low investment income, and high market income. One labelled HLLL is primarily donation-funded with low income from everything else.
- Key takeaway
- Think of it like a postcode for funding strategy. Just as a postcode tells you roughly where something is without describing every detail of the building, an archetype tells you the broad structural position of an organisation without describing every line of its accounts.
What does high and low actually mean?
High and low are always relative to the average across all large charities in that year. The measure isn't relative to a fixed, absolute number. An organisation is H on government funding if it receives a larger share of its income from government than the typical organisation in this dataset. It is L if it receives less.
This means the labels are about structural position, i.e. where an organisation sits in the landscape, rather than about the absolute size of any income stream. A charity receiving 25% of its income from government grants might be H one year and L another if the sector average shifts, even if its own funding has not changed at all.
Why use a simple high/low split rather than more categories?
Because the underlying data supports it. When you look at how organisations are actually distributed on each funding axis, they do not spread evenly from 0% to 100%. They tend cluster at the extremes. Most organisations receive either very little from government or a lot. There are very few in the middle and the same is true for market, donations, and investments.
This means a high/low split is reflecting a genuine polarisation already present in the data rather than imposing an artificial division on a continuous spread. Organisations naturally separate into two groups on each axis, and the threshold sits in the sparse middle ground between them.
A more complex system with three or four categories per axis would generate hundreds of possible combinations, most of which would contain too few organisations to be meaningful. The binary system produces sixteen* real, populated archetypes that can be compared and tracked over time.
*While there are sixteen theoretical positions, HHHH has only been observed once across 27,515 archetypes (0.003%). Read more about what this means here.
Could the classification change if you drew the line slightly differently?
This was tested directly and the results are reassuring. The threshold on each axis was shifted up and down by the full amount it naturally varies from year to year (roughly 1–2% depending on the axis) and the classification was rerun each time.
Only 3.1–3.4% of organisations changed archetype when thresholds were shifted in either direction. The other 97% sat far enough from the boundary that a realistic threshold shift made no difference to their label.
More importantly, the top five archetypes appeared in exactly the same rank order across all three threshold variants, with proportions that moved by at most 2%. The dominant structural features of the sector are entirely insensitive to where exactly the threshold is drawn.
- Key takeaway
- Shifting the classification line by a full year's worth of natural variation only affects 3% of organisations and changes nothing about which funding configurations dominate the sector.
Classification stability and change
Do organisations stay in the same archetype over time?
Most do. Of the 5,503 organisations tracked across all five years, 57% held exactly the same archetype throughout. A further 31% moved between two archetypes, 9% between three, and a small number moved more.
The dominant finding is structural lock-in. Most large charities occupy the same funding position year after year, not because they have not tried to change, but because the funding environment they operate in makes certain positions self-reinforcing and others very difficult to reach.
When an organisation's archetype changes, does that mean something real happened?
Usually yes — but not always in the same way.
The 43% of organisations that change archetype at some point break down into three meaningfully different groups:
Genuine repositioning (22%生)
- The organisation moved in a consistent direction and stayed there i.e. its funding structure genuinely changed.
Cyclical oscillation (17%)
- The organisation crossed into a new archetype and held it for two or more years before returning. This may reflect specific funding dynamics (e.g. COVID example below) rather than a permanent structural shift.
Classification noise (4%)
- A single-year blip that reversed immediately. The most likely explanation is a one-off receipt or minor income fluctuation rather than any real strategic change.
Both archetype-level and the individual funding stream analyses produced identical classifications with no ambiguous cases which confirms these three groups are genuinely distinct rather than arbitrary categories.
What does "classification noise" mean and how much is there?
Classification noise refers to cases where the archetype label changes not because anything meaningful happened to the organisation's funding, but because a small income fluctuation pushed it just over or under the threshold for a single year.
The true noise level — changes that revert within a single year and cannot reflect any genuine shift in funding strategy — is approximately 2% of individual funding stream classifications, affecting around 4% of organisations. This is considerably lower than the 21% raw oscillation figure, because most organisations that oscillate do so over multi-year periods that reflect real funding dynamics rather than statistical outcomes.
- Key takeaway
- The classification system makes genuine errors for about 1 in 25 organisations. For the remaining 24 in 25, changes in archetype label reflect something real about the organisation's funding environment.
Which funding stream is the most stable classifier?
Investment income by a considerable margin. 94% of organisations never change their position on the investment axis across all five years. Only 2% show any oscillation at all.
This makes sense once you understand what investment income represents. An organisation either has an endowment or significant invested assets or it does not. There is very little middle ground and it doesn't change much year to year. Investment income is effectively a structural feature of an organisation rather than an operational variable.
Government income is the most active axis, showing both the highest genuine directional movement (13% of organisations trending consistently) and the highest longer-term oscillation (8%). As the sections below explain, this reflects specific and traceable policy dynamics rather than random instability.
Organisations caught between classifications
What is happening with the organisations who oscillate over multiple years?
These are organisations that cross into a new archetype, hold that position for two or more consecutive years, and then return to their original position. They are not random noise i.e. a genuine shift has happened, but they do not consolidate the change either.
The majority (78%) are moving on a single funding stream only, while remaining completely stable on the other three. Government income is the most common single driver, accounting for nearly half of these cases.
How did COVID affect the findings?
Looking at the 321 organisations that oscillate on the government axis alone, 48% (155 of them) show traces of COVID funding: they sat below the government threshold in 2020, crossed above it in 2021, and returned below it in 2022.
Government grant income for this group nearly doubled in 2021 — rising from 13% to 24% of total income — before falling straight back to 13% in 2022. The dataset as a whole also saw a grants increase in 2021 (from 9% to 13%), but the oscillating group's spike was almost twice as large, confirming these were organisations already close to the threshold that were pushed over it by emergency COVID grant funding and then returned when it ended.
These 155 organisations are not structurally unstable but they received a time-limited government intervention. As a result, their archetype label changed to reflect it and then it changed back when the intervention ended. The classification is registering a real funding event but the event was a once-in-a-generation pandemic rather than a strategic choice.
What about the other half of the government oscillators?
The remaining 52% (166 organisations) tell a different story about a structural shift in how government funds charities that was already underway before COVID and has continued since.
For the oscillating group as a whole, government contracts and grants tracked closely through 2020–2022. From 2023 onward, contracts overtook grants and have stayed ahead. The same crossover is visible across the full dataset. This is the long-documented shift from discretionary grant funding toward commissioned service delivery i.e. government funding progressively moving from giving charities money to deliver their own priorities, toward paying charities to deliver government priorities under contract.
The organisations in this group are caught mid-transition. They have enough government contract income to cross the threshold in some years but not enough to consolidate there permanently. Rather than failing to manage their funding, they are navigating a funding environment that is itself changing and their archetype movement is the visible trace of that change.