How Do You Design a Democracy for 173 Million People?

In 1951, India attempted something many observers considered wildly optimistic.

It prepared to let every adult citizen vote.

The country was newly independent. Its first electoral rolls contained roughly 173 million people. Most could not read or write. Many had no formal proof of age. Millions of women were recorded in local registers not by their own names, but as “wife of” or “daughter of” somebody else.

There were remote villages, forests, islands, mountains, unreliable roads and hundreds of languages and dialects.

India did not merely need to announce an election.

It had to invent a way for an enormous and unequal population to participate in one.

This was democracy as a design problem.

The idea in 20 seconds

Inclusion is not achieved by declaring that everybody may participate.
A system is only universal if people can actually recognise, reach and use it.
Symbols can function as infrastructure, not decoration.
Sometimes the interface must change because the population should not have to.

Universal rights require deliberately non-universal solutions: different routes, formats and supports for different human realities.

A single ballot box branching into several differently shaped doorways, each doorway sized and shaped for a different kind of person to pass through equally, flat editorial conceptual illustration, warm cream and charcoal palette with a single amber-orange accent, generous negative space, 16:9

The scale was almost absurd

India’s first general election took place over several months in 1951 and 1952.

Election officials had to register voters, define constituencies, assign symbols, print ballots, manufacture ballot boxes, recruit and train workers, transport material and conduct polling across the country.

This was before electronic databases, mobile phones, GPS or instant communication.

The physical system involved hundreds of thousands of polling stations and roughly two million steel ballot boxes.

But scale was not the hardest part.

The harder problem was making one election legible to people with radically different levels of literacy, language, documentation and access.

The Constitution could grant equal voting rights in a sentence.

The interface had to make that equality real.

An elderly woman being identified by a polling officer at a striped-tent polling booth during India's first general election, 1952. — A voter is identified at a polling booth near Jama Masjid, Delhi, during India’s first general election, January 1952 — a system built for 173 million mostly first-time, largely non-literate voters. *1952, public domain.*

A name was not enough

A ballot made only of written candidate names would have excluded a huge portion of the electorate.

So candidates and political parties received visual symbols.

A pair of bullocks under a yoke. A hut. A tree. An elephant. A lamp.

The voter did not need to decode a paragraph. They could recognise a shape.

In the earliest system, each candidate had a separate ballot box displaying the candidate’s name and symbol. The voter received a ballot paper and placed it into the chosen box.

This created a direct relationship:

recognise symbol → find matching box → cast ballot

The symbol was not a logo added after the system had been designed.

It was part of the voting mechanism itself.

A voter in 1952 leaning over a row of separate ballot boxes, each printed with a different candidate's symbol, choosing where to post a ballot paper — A voter in front of several ballot boxes during India’s first general election, 1951-52, each box printed with a different candidate’s symbol. *1952, public domain.*

Symbols are interfaces when language cannot be assumed

Designers often discuss icons as space-saving devices.

India’s election symbols did something more important. They crossed literacy and language boundaries.

That does not mean symbols are naturally universal. A symbol still has to be culturally recognisable, visually distinct and easy to reproduce at many sizes and in poor printing conditions.

Two symbols that look similar can create the same problem as two similar words.

So the system needed a visual vocabulary with enough difference between options to survive:

hurried viewing,
low-quality printing,
unfamiliarity,
weak eyesight,
and translation across regions.

This is closer to designing road signs than designing a brand identity.

The goal is not to look distinctive in a presentation.

The goal is to remain unmistakable in the world.

Two nearly identical election symbols side by side, a tree and a slightly different tree, blurred and smudged as if poorly printed, with a single crisp unmistakable road sign style symbol beside them, flat editorial conceptual illustration, warm cream and charcoal palette with a single amber-orange accent, generous negative space, 16:9

The women missing from their own names

While creating the electoral rolls, officials encountered millions of women registered only through their relationship to a man.

“A’s mother.”

“B’s wife.”

The Election Commission insisted that eligible voters be enrolled under their own names. Records that did not identify the woman herself could not simply pass as complete.

This exposed a deeper truth about system design:

A database does not merely store the world. It decides what counts as a person inside the world.

If a citizen cannot be represented correctly in the data model, their formal right becomes difficult to exercise.

The problem was not a missing text field.

It was a social convention embedded in the information architecture.

We still encounter versions of it:

forms that assume everybody has a surname,
addresses that require a Western format,
identity systems that cannot represent chosen names,
family fields built around only one kind of household,
verification flows that assume everybody possesses the same documents.

Exclusion often begins long before the final screen.

It begins in the schema.

A queue of voters, including women observing purdah in white veils, waiting to cast their votes at a polling booth near Jama Masjid, Delhi, in January 1952 — Women observing purdah wait in a mixed queue to vote near Jama Masjid, Delhi, on polling day, 14 January 1952 — the same election in which officials had to fight to enter women in the rolls under their own names. *1952, public domain.*

Access included geography

The polling station was also part of the interface.

A perfectly clear ballot is useless if the voter cannot reach it.

Election material had to travel by road, rail, boat and on foot. Polling schedules had to account for weather and local conditions. Physical access was not a secondary operations problem; it was part of whether the democratic system worked.

Digital product teams often separate “experience” from “operations.”

The election makes that separation look artificial.

For the voter, the experience included:

learning that they were registered,
finding the polling place,
reaching it,
identifying the correct symbol,
casting the ballot,
and trusting that it would be counted.

Every handoff mattered.

The service was larger than the artefact.

A relay of election workers carrying a ballot box by road, by rail, by boat, and finally on foot across a mountain path toward a distant remote village, flat editorial conceptual illustration, warm cream and charcoal palette with a single amber-orange accent, generous negative space, 16:9

It was not elegant—and that is interesting

Separate ballot boxes for candidates required enormous amounts of material and coordination.

By contemporary standards, the system looks cumbersome.

But elegance depends on the constraint you are solving for.

If the primary goal is minimum hardware, the solution is inefficient.

If the goal is allowing a first-time voter with limited literacy to make a secret and understandable choice, the physical redundancy starts to make sense.

Good design is not the fewest objects or the shortest flow.

It is the clearest path through the actual conditions.

Sometimes simplicity for the user requires complexity behind the scenes.

A cramped storeroom stacked high with dozens of separate labeled ballot boxes for every single candidate, deliberately excessive machinery built so one confused first time voter can choose simply, flat editorial conceptual illustration, warm cream and charcoal palette with a single amber-orange accent, generous negative space, 16:9

The inclusion test

When designing a supposedly universal service, ask:

Who exists in the data model?

Can the system represent their real name, identity, household and address?

What ability are we quietly assuming?

Literacy, vision, language, mobility, documentation, connectivity or confidence?

Is the symbol doing real work?

Can it be recognised without accompanying text and distinguished under poor conditions?

Can people reach the service?

Access includes devices, transport, timing, assistance and physical location.

Where has complexity been placed?

The system should absorb complexity when doing so makes participation simpler.

A five item checklist rendered as a simple ledger page, each line marked with a small icon for a name, an eye, a symbol, a road, and a gear, flat editorial conceptual illustration, warm cream and charcoal palette with a single amber-orange accent, generous negative space, 16:9

One final thought

India’s first election is often described as the world’s largest experiment in democracy.

It was also one of the world’s largest acts of interaction design.

The state could have demanded that the voter adapt: learn to read, acquire the right paperwork, understand a complicated ballot and travel however far necessary.

Instead, imperfectly but deliberately, the system adapted itself.

It created symbols where words were insufficient. It created polling infrastructure where geography resisted. It confronted records that failed to name women as individuals.

That is the difference between access as a promise and access as a design.

A system becomes inclusive not when everybody is invited, but when the conditions of entry stop excluding them.

The ballot box was only the final object.

The real design was the path that allowed a citizen to reach it.

A single ballot box sitting at the end of a long winding path shaped from symbols, footprints, and paper, showing the whole journey a citizen took just to reach it, flat editorial conceptual illustration, warm cream and charcoal palette with a single amber-orange accent, generous negative space, 16:9