5-Year Presidential Term: More Accountability, Not Comfort

Ghana’s constitutional reform conversation has a habit of getting reduced to slogans: “give the president more time,” “reduce elections stress,” “stabilise government.” Prof. Henry Kwasi Prempeh, Chair of the Constitution Review Committee, is pushing back on that framing. His point is simple and sharp: a proposed 5-year presidential term isn’t designed to make life easier for presidents—it can actually be harsher.

That claim matters beyond politics. In our “Sɛnea AI Reboa Adwumadie ne Dwumadie Wɔ Ghana” series, we keep coming back to one hard truth: AI in Ghanaian workplaces and schools won’t scale on vibes. It scales on systems. And systems get better when leadership is held to clear standards, with measurable results, predictable timelines, and real consequences.

So if you care about AI adoption in Ghana—whether you’re running a business, managing a public institution, or working in education—this term-length debate is not “just politics.” It’s a debate about accountability design, and accountability design is the foundation of effective governance.

Why a 5-year term can be tougher on presidents

A longer term increases the distance between elections, but it also increases the weight of what happens between them. That’s the “harsher, not kinder” logic.

With a 4-year term, a president has a tight political cycle: form a government, stabilise, show early wins, and then shift quickly into re-election mode. Whether we like it or not, that rhythm shapes decision-making. Short cycles reward visibility over durability.

A 5-year term changes the incentives:

• More time to be judged on outcomes, not announcements. Big projects can’t survive on press conferences for five years.
• Longer exposure to policy consequences. If a programme is poorly designed, citizens and markets live with it longer before the next election reset.
• Harder excuses. “We just came in” stops working earlier, and “we need more time” sounds weaker when you’ve already had five years.

A longer term doesn’t automatically create better performance. It raises the standard of proof.

That’s why Prempeh’s statement is worth taking seriously. It’s less about sympathy for any president, and more about building a structure where performance is harder to fake.

The real issue: election pressure vs policy discipline

Ghana’s governance challenge isn’t only the length of the term; it’s how much of the term gets swallowed by election survival.

Here’s the pattern many institutions recognise:

• Year 1: settling in, appointments, new priorities, early direction
• Year 2: implementation begins, but procurement and coordination delays hit
• Year 3: politics starts heating up, “quick wins” dominate
• Year 4: campaign season, spending pressure, policy becomes messaging

A 5-year term can give room for policy discipline, but only if it’s paired with mechanisms that prevent complacency.

What actually makes a longer term “harsher”

A longer term becomes harsher when you combine it with stronger mid-term accountability. That’s the part many people skip.

If Ghana extends to five years but keeps weak performance reporting, vague targets, and low transparency, then yes—leaders might feel more comfortable.

But if reforms come as a package—term length plus tougher reporting and institutional checks—then the president has more rope and fewer hiding places.

Practical examples of “harsher” accountability design:

1. Mandatory annual performance scorecards for ministries (published, comparable year-on-year)
2. Legally enforced timelines for procurement, project delivery, and audit responses
3. Stronger parliamentary committee powers to demand data and sanction non-compliance
4. Clear public service delivery benchmarks citizens can track (health, education, roads, digitisation)

This is where the AI conversation naturally fits.

Governance is the operating system for AI adoption in Ghana

AI tools don’t fail first because of algorithms. They fail because of weak workflows, unclear ownership, and poor data discipline. That’s governance, just in a different outfit.

If Ghana wants AI to improve public services and workplace productivity, we need leadership that can do three unglamorous things consistently:

• Set measurable targets (not just intentions)
• Track performance with real data (not speeches)
• Correct course fast (not defend mistakes for political pride)

That’s why term-length reform connects to AI in Ghanaian workplaces and schools. When leaders are accountable over a longer horizon, they’re pushed toward:

• building durable digital infrastructure
• investing in civil service capability
• standardising data systems across agencies
• enforcing procurement and cybersecurity discipline

The uncomfortable truth about “digital transformation”

Most institutions say they want digital transformation, but they still:

• store records across disconnected spreadsheets
• collect the same data multiple times from citizens
• lack a single source of truth for decision-making
• treat ICT as “support,” not core strategy

AI amplifies whatever environment you put it in. If the system is chaotic, AI scales chaos faster. If the system is disciplined, AI scales productivity.

So when Prempeh says a 5-year term can be harsher, I hear this: leaders will have fewer excuses for not fixing the boring system problems that block progress.

What Ghana should demand if term length changes

If we’re changing the calendar, we should also change the scoreboard. A five-year term without accountability upgrades is just a longer wait.

Here are reforms that would make a 5-year term genuinely pro-performance—especially for national competitiveness in the digital age.

1) Publish “delivery metrics” that ordinary people can understand

Don’t bury performance in technical reports. Use clear metrics like:

• average time to register a business
• passport processing turnaround time
• percentage of schools with functional ICT labs and internet
• number of public services accessible end-to-end online
• audit issues resolved within set deadlines

If a ministry can’t measure it, it usually can’t manage it.

2) Treat data as public infrastructure

AI in Ghana depends on data that’s clean, consistent, and legally usable.

That requires:

• common data standards across MDAs
• data-sharing agreements with privacy safeguards
• routine data quality audits
• training public workers to collect and use data correctly

A longer term should come with zero tolerance for “we don’t have the data.”

3) Build mid-term accountability, not just end-of-term judgment

If elections are every five years, then mid-term checks become more important.

Options include:

• mid-term national performance review debates with published datasets
• stronger citizens’ feedback mechanisms tied to service delivery
• mandatory parliamentary reporting on flagship programmes twice a year

The goal is simple: reduce the distance between citizen pain and government response.

4) Protect long-term projects from political resets

AI readiness—especially in education—needs continuity.

When curricula, procurement, and teacher training are reset every election cycle, nothing scales. A five-year term can help continuity, but only if Ghana strengthens:

• bipartisan commitments around core digital infrastructure
• multi-year budgeting discipline
• procurement transparency (to reduce project abandonment)

How this connects to AI in workplaces and schools (practical angle)

Better accountability produces better service delivery—and that creates room for AI to actually help.

Here’s the chain reaction I’ve seen in organisations that do AI well:

1. Leadership sets clear goals (customer service time, cost reduction, quality targets)
2. Teams fix basic workflows (standard forms, clear approvals, clean data)
3. Then AI is introduced (document automation, chat support, analytics)
4. Performance is tracked and improved continuously

Ghana as a country needs the same sequence.

What a “5-year harsher term” could push ministries to do

If the political system expects results over a longer horizon, ministries are more likely to prioritise:

• AI-assisted document processing to cut backlog (permits, registrations, claims)
• predictive analytics for budgeting and resource allocation
• education planning dashboards to track teacher deployment and learning outcomes
• fraud and anomaly detection in revenue collection and procurement

These are not sci-fi ideas. They’re operational. And they depend on governance discipline.

People also ask: will a 5-year term reduce election costs in Ghana?

Yes, fewer elections over time can reduce direct election administration costs, but cost reduction isn’t the main benefit—and it’s not guaranteed.

If political competition intensifies, parties can still spend heavily. The bigger win is not “saving money,” it’s saving institutional attention—less time spent in campaign mode and more time spent delivering measurable outcomes.

People also ask: doesn’t a longer term risk weakening democracy?

It can, if checks and transparency stay weak. A longer term increases the need for stronger oversight, open data, independent institutions, and enforceable consequences for non-performance.

A longer term without accountability upgrades is risky. A longer term with stronger accountability can be healthier than what we have now.

What to watch in 2026: the reform package, not the headline

We’re at the end of 2025, and Ghana’s policy conversation is already drifting toward the next electoral cycle. That’s exactly why Prempeh’s framing is useful: don’t judge the proposal by the number of years alone. Judge it by what it forces leaders to deliver—and what citizens can verify.

For readers following this series on AI at work and in education in Ghana, here’s the stance I’ll defend: AI adoption is a governance test. Countries that can manage projects, measure outcomes, and protect public value will use AI to grow productivity. Countries that can’t will buy tools and still get stuck.

So the bigger question isn’t whether the term is four years or five. What would make leaders measurably accountable every single year of that term—and how do we build systems (including AI systems) that make performance impossible to hide?