Divide and Rule

Primary bias

In-group/out-group bias: people naturally sort the world into "us" and "them," and judge "them" more harshly and less charitably by default. Divide and rule doesn't need to invent this tendency, it only has to activate it and point it at a specific rival group.

Secondary biases

Naive realism: each side assumes its own view of the conflict is simply how things are, so the other side's disagreement feels like bad faith rather than a different perspective.

Confirmation bias: once primed to see a rival group as hostile, people notice and remember what confirms that and skip past what doesn't.

Outgroup homogeneity effect: the rival group starts to feel more uniform and more unanimously hostile than it really is, "they all think that," which turns a mix of individuals into a single threat.

How it activates

1. The divider identifies an existing fault line in a population.
2. Each side is fed something that looks like support against the other side, not against the divider.
3. Each side reacts to the perceived threat from the other, and the hostility becomes real and self-sustaining.
4. Both groups direct their attention and organizing energy at each other.
5. The divider governs, profits, or avoids scrutiny largely unopposed, since no unified resistance forms.

Susceptibility factors

Strong: strength of group identity. The more strongly someone identifies with their group, the easier they are to mobilize against a rival one.

Moderate to strong: resource scarcity. Competition over jobs, land, or status makes "they're taking what's yours" narratives land harder.

Moderate: low cross-group contact. Groups that rarely interact directly are easier to caricature.

Moderate: an existing historical grievance. A divider rarely has to invent hostility from nothing, amplifying a grievance that's already there is cheaper.

What does not predict it

Education level alone. This technique doesn't rely on false facts so much as on identity and emotion, and well-informed people from every group involved have fallen for it. What matters more is the habit of asking who benefits if the two groups never team up.

Primary purpose

Mostly other: preventing unified opposition before it can form, distinct from discrediting or distracting an opposition that already exists. It often works alongside distraction too, since infighting is an effective distraction from the divider's own actions.

Immediate effect

Heightened suspicion and hostility toward a rival group, with attention pulled toward that rival instead of toward whoever is actually shaping events.

Long-term effect

Entrenched group identities and normalized mutual hostility that can outlast the original divider by generations, see the Rwanda example in Section 07.

Cui bono

Colonial administrations, rulers managing a large or diverse territory, employers managing a workforce, and modern state or non-state actors running influence operations against a rival country's electorate.

Offensive or defensive

Both. Offensively, it prevents an opposition from uniting against the divider. Defensively, a divided population has a much harder time organizing scrutiny of the divider itself.

How it works

Setup: the divider maps the social terrain and finds a real or latent fault line. Delivery: support, messaging, or favorable policy is fed to each side in a way that looks like it's about that side's own interests, not about the divider. Effect: both sides mobilize against each other, and unity against the divider becomes structurally difficult.

Preconditions

At least two identifiable subgroups with some basis for rivalry, real or latent, and a divider with enough reach, institutional power, budget, or media access, to feed messaging or favors to more than one side at once.

When it fails

If the target groups have strong cross-cutting ties, intermarriage, shared economic interests, mixed communities, the wedge often doesn't take. If the divider gets exposed, hidden coordination across "both sides" coming to light, that discovery can itself become a unifying grievance that turns both groups against the divider instead of each other.

Amplifiers

Crisis conditions (war, scarcity, election season) that already strain intergroup relationships; algorithms that reward emotionally charged, identity-based content; and anonymity, which lets a single actor present as a member of multiple opposing groups at once.

Natural habitat

Resources needed

Medium to high. State-level divide and rule needs real infrastructure, either institutional power to write laws and allocate favors, or a sustained information-operation budget. A small-scale version, a manager badmouthing coworkers to each other, needs almost nothing, but runs on the same underlying logic.

Scale dependency

Works at both ends. A single well-placed policy decision can have generational effects, but online versions typically need sustained, repeated, precisely targeted campaigns to keep the division active.

Digital mutation

Modern ad-targeting tools let one operator run opposing campaigns to opposite audiences from the same infrastructure, at a fraction of the cost and visibility of twentieth-century propaganda operations, and neither audience ever sees what the other was shown. See Manufactured consensus, Section 09.

First documented use

Genuinely contested rather than a single verified moment, this is one of the oldest strategies in recorded politics, older than any confirmed "first instance."

Named and theorized by

The maxim divide et impera is attributed to different rulers depending on the source, most commonly Philip II of Macedon, though some historians credit it to Julius Caesar or to Louis XI of France instead. Niccolò Machiavelli examined the same logic directly in The Art of War (1521), Book VI, advising commanders to divide an enemy's forces by making them suspicious of their own trusted men. Immanuel Kant later listed divide et impera as one of three cynical maxims of statecraft in Perpetual Peace: A Philosophical Sketch (1795), Appendix I, alongside acting first and apologizing later, and denying wrongdoing outright.

Tradition

Predates the IPA's twentieth-century technique taxonomy entirely, a classical political-strategy concept that modern propaganda studies later absorbed as a named structural technique.

Key evolution

Pre-mass-media (direct classification and favoritism through court politics and colonial administration), to mass-media era (state-sponsored press stoking regional, class, or ethnic tension at national scale), to social-media era (individually targeted, algorithmically delivered campaigns capable of showing opposite messages to opposing audiences from a single, often hidden source).

Cross-cultural notes

The Chinese military tradition attributed to Sunzi (roughly 5th century BCE) independently describes a comparable logic of dividing an enemy's forces, suggesting the strategy was arrived at separately in more than one ancient tradition. In German-language sociology, Norbert Elias coined the term Königsmechanismus ("king mechanism") in his book Die höfische Gesellschaft to describe how Louis XIV kept the nobility and the rising bourgeoisie competing against each other at his court, so neither group could unite against the crown. It's a related but more specific and later concept than Teile und herrsche, developed within sociology rather than propaganda studies.

Across all three, the same shape repeats: the outside actor doesn't need either group to be right. It just needs both groups busy fighting each other instead of asking who benefits from the fight continuing.

Pre-bunking note

Broader inoculation-theory research (from William McGuire's original work through more recent "prebunking" games by Sander van der Linden and Jon Roozenbeek) shows that warning people about a manipulation technique in advance measurably increases resistance to it. Harmony Square, a short online game built specifically around polarization tactics, was tested in field trials and found to reduce players' susceptibility to, and willingness to share, divisive manipulative content afterward.

Backfire risk

Real. Telling a group their grievance toward a rival group is "manufactured" can feel like dismissal, especially if the underlying grievance is at least partly real, and clumsy correction can deepen distrust rather than reduce it. The safer move is to name the technique and the divider, not to tell either side their grievance itself is fake.

Commonly paired with

Often precedes

Scapegoating and dehumanization. Once a group is primed to see a rival as a threat, it becomes easier to blame that group collectively or strip them of moral consideration.

Often follows

Fear-mongering or agenda setting, which first raise the salience of a threat or crisis that then gets pinned on the rival group.

Gets confused with

Primary sources

Accessible reading