One More System We Don’t Need to Build

When Bangladesh talks about modernizing education, the conversation almost always turns to devices.

First, it was laptops. Now, it is tablets.

The instinct is understandable. If teachers have technology, classrooms will improve. If devices are distributed, learning will follow. It is a clean, visible solution to a complex problem.

But it is also the wrong abstraction.

The real problem is not device ownership. It is device access. And how you design that difference determines whether a policy quietly succeeds or loudly fails.

I saw this up close.

While serving on the Curriculum 2022 committee, I proposed something that, at the time, felt almost too simple. The government should not provide devices to teachers at all. Instead, it should provide an allowance.

Each teacher would receive a stipend aligned with market rates, enough to purchase a laptop that meets a government-defined minimum configuration. This allowance would be renewed every three years, allowing teachers to upgrade as technology evolves.

Alongside this, I proposed that teachers be given access to a Teletalk SIM with zero-tariff internet, ensuring that connectivity would not become the bottleneck.

At the time, there was interest. Because the logic is hard to ignore.

If the goal is to ensure that teachers have reliable access to digital tools, why should the government own those tools? Why should it procure them, distribute them, track them, repair them, and eventually replace them? Why should it build an entire administrative system just to manage hardware? See the attached image for example.

The traditional model — the one we are now doubling down on with “One Teacher, One Tab” -- assumes that central control leads to efficiency. That standardization leads to reliability. That ownership leads to accountability.

But experience suggests otherwise.

We have already seen what happens when the government becomes a device manager. Laptops were procured in bulk, vendors were selected centrally, and when those vendors could not support the devices, the system had no fallback.

Spare parts were unavailable. Repairs stalled. Devices broke and stayed broken. Many ended up sitting idle, quietly removed from the daily life of classrooms.

More recently, the system has gone even further. Central agencies have issued instructions to institutions to modify BIOS-level configurations on those devices, such as removing vendor logos.

This is not just unnecessary. It is revealing. A simple goal -- give teachers access to digital tools -- has evolved into a system that requires firmware-level intervention from the center. See attached image for reference.

That is not modernization. That is overengineering.

And overengineering has consequences.

Every additional layer of control creates additional work. Procurement teams must negotiate with vendors. Administrative units must track inventory across thousands of institutions.

Technical teams must issue directives. Local staff must execute them, often without the required expertise.

Each step introduces delay, inconsistency, and the possibility of error. More work means more mistakes. More mistakes increase costs. And higher costs make the system harder to sustain.

In the end, the system becomes busy, but not effective. The allowance model avoids this entirely.

It shifts the problem from the state to the user, where it belongs. Teachers, like any professionals, are capable of managing their own tools when given the resources to do so.

When a teacher purchases a device with their own allowance, the incentives change. The device is no longer an external asset to be managed.

It becomes a personal tool that must function. Maintenance becomes immediate, because the user depends on it. Repairs happen through local markets, which are far more responsive than centralized bureaucracies.

Replacement happens naturally as technology evolves, not through delayed procurement cycles.

The system becomes decentralized, flexible, and resilient.

There is no inventory to maintain. No vendor lock-in. No central directives about BIOS settings. No single point of failure.

The addition of zero-tariff internet addresses the other half of the problem.

Without connectivity, even the best device is underutilized. By ensuring that teachers can access the internet without cost barriers, the policy guarantees that access is meaningful, not just symbolic.

Taken together, these ideas solve the problem at its root, without creating a parallel system to manage hardware.

This is what good policy design looks like. It solves the problem without creating new ones.

The “One Teacher, One Tab” initiative does the opposite. It creates a system that must be continuously managed, updated, and corrected. It assumes a level of administrative capacity that has not been demonstrated at this scale. And it risks repeating the same pattern we have already seen -- procurement followed by gradual decay.

Let’s fix it!

Omar Shehab is a theoretical quantum computer scientist at the IBM T. J. Watson Research Center, New York.