What 64 years of Nigerian data say about the country's best-performing government, when you take out the vibes.
About 5 min · Nigeria, 1960 to 2024
It's the kind of question Nigerians argue about in traffic, in beer parlours, on Twitter. Everyone has an opinion. Not enough of them have the data.
We gathered 64 years of GDP, governance scores, regime changes and oil prices. Fed it into structural break tests, lag correlations, time-series models. Asked one thing: when did Nigerian governance actually improve, in a way that wasn't just noise?
The algorithm kept returning the same answer.
We tested every major regime change. A "structural break" lights up when the trend genuinely shifts, not just wobbles. Watch them come in.
"Significant" means: if you ran this experiment a million times with random data, you'd see a change this large less than once. The p-value for 1999 is under 0.000001.
moment of governance improvement that wasn't statistical noise.
The name attached to it is Olusegun Obasanjo.
Real GDP per president, in constant 2015 dollars. The bars below tell you how much each one added to the economy on their watch. Per-year, Obasanjo ran at a rate nobody else has matched.
Source: World Bank WDI, constant 2015 USD. Buhari's number ends 2023; effects of the 2023 reforms are still landing.
So we broke it apart sector by sector. If Obasanjo's growth was just an oil story, agriculture, services and telecom wouldn't have moved. They moved.
Log scale on the y-axis. Telecom's number breaks every linear chart. From 0.02 to 26 lines per 100 people in eight years.
The government held a competitive GSM spectrum auction. Multiple operators won. The licence terms were enforced.
The fastest sectoral growth in Nigerian history followed. Hundreds of thousands of jobs, the rails for mobile banking, the floor under fintech and e-commerce. Your phone is a product of a governance decision made in 2001.
It could easily have gone the other way. The government could have kept NITEL as the only operator. It could have handed licences to cronies without auction. It chose the harder, more transparent option. The data has been thanking it ever since.
When we first wrote this, Obasanjo was the only Fourth Republic president whose tenure ended with a positive governance score. The numbers have since been re-baselined and the picture has shifted. Across the full Fourth Republic, two presidents presided over real improvements in the governance composite.
Obasanjo still wins on the bigger picture. He presides over the statistically significant 1999 break, the highest GDP growth rate of any Fourth Republic president, the sector explosions, and the policy decisions still paying off today. But the simple "only one" framing is no longer accurate. Jonathan's term, on the institutional data, was the second improvement.
One caveat we owe you. The World Bank's governance scores carry a standard error of roughly 0.2 for Nigeria. Obasanjo's +0.20 and Jonathan's +0.21 are the only two term changes large enough to clear that noise. The other three are, statistically, flat. Measured governance has barely moved in twenty-eight years. That is its own kind of verdict.
If Nigeria had kept Obasanjo's growth rate from 2007 onwards, the economy in 2023 would have been around twice the size it actually was.
Rough back-of-the-envelope, compounding the Obasanjo CAGR of 7.7% for sixteen years. Not a prediction. A reference point. The gap is the cost of choosing differently.
Treat that as an order of magnitude, not a figure. Priced properly against a peer benchmark, the gap centres near $600B, but the honest confidence band is wide enough to include almost no effect at all. What survives the uncertainty is the direction, not the decimal. Choosing differently had a cost. How big, we cannot pin down.
Correlation isn't causation.
Oil prices rose. China was buying everything. Paris Club debt relief in 2005 was a one-off windfall. Some of Nigeria's growth would have happened anyway. The data captures institutional outcomes; it doesn't capture Obasanjo's military networks, his political instincts, or whether anyone else could have made the same calls in the same room.
And there is plenty to hold against him. Odi. Zaki-Biam. The third-term ambition. Selective prosecutions. None of that disappears because the numbers are good.
But when you strip out the bias and let the data argue with itself, one period of Nigerian post-independence history stands above the others. The algorithm keeps choosing it.
The data exists. Governance can be measured. Outcomes can be tracked. The next presidential decision should be judged the same way.
What 64 years of data say about Nigeria's best-performing government.
We didn't set out to write about Obasanjo.
We set out to answer a simpler question: Does governance actually matter for Nigeria's economy? It's the kind of question Nigerians argue about in traffic, in beer parlours, on Twitter. Everyone has an opinion. Not enough of them have the data.
So we gathered years of Nigerian economic history. GDP figures, governance indicators, regime changes, all of it, from 1960 to 2023. We fed it into statistical models: Granger causality tests, structural break analysis, lag correlations, and time-series regressions.
We asked the algorithm a simple question: When did Nigeria's governance actually improve, and does it matter for the economy?
The algorithm kept returning the same answer.
1999.
And the name attached to 1999 is Olusegun Obasanjo.
We ran structural break tests on Nigeria's governance data, a statistical technique that identifies moments when a trend fundamentally shifts. [Aso Rock Index]
We tested every major political transition:
The results:
| Year | Event | Governance change | Significant? |
|---|---|---|---|
| 1966 | First Coup | −0.19 | No |
| 1979 | Second Republic | +0.04 | No |
| 1984 | Buhari Regime | +0.10 | No |
| 1999 | Fourth Republic | +0.48 | YES (p < 0.000001) |
| 2015 | APC Takes Power | +0.28 | No |
In 64 years of Nigerian history, coups, civil war, oil booms, oil busts, military dictators, civilian presidents, there is exactly one moment when governance improved in a way that was not just noise.
1999.
The probability that this is random chance is less than one in a million.
Your uncle will tell you Buhari's first government was disciplined. Your auntie might argue Shagari was underrated. Everyone has memories.
But memories aren't data.
When we say the 1999 improvement is "statistically significant," we mean: if you ran this experiment a million times with random data, you would see a change this large less than once. It is what the numbers say when you remove human bias from the equation.
Obasanjo inherited an economy worth $171 billion (constant 2015 US$). When he left in 2007, it was worth around $310 billion. [World Bank WDI]
That is about $139 billion added in eight years, at an annual growth rate of 7.7%.
In context:
| President | Tenure | GDP added | Annual growth |
|---|---|---|---|
| Obasanjo | 1999 to 2007 | +$138.6B | 7.70% |
| Yar'Adua | 2007 to 2010 | +$76.2B | 7.61% |
| Jonathan | 2010 to 2015 | +$107.1B | 5.02% |
| Buhari | 2015 to 2023 | +$26.8B | 0.66% |
Obasanjo added more to Nigeria's economy than Jonathan and Buhari combined.
But this was not just oil.
When oil prices rise, any government looks good. That's the lazy critique of the Obasanjo years.
So we disaggregated the data. We looked at what happened sector by sector. [World Bank WDI]
| Metric | 1999 | 2007 | Change |
|---|---|---|---|
| Agriculture GDP | $30B | $70B | +133% |
| Annual growth | n/a | n/a | 11.2% CAGR |
Nigeria's farms produced more than twice as much value in 2007 as in 1999. This isn't oil. This is groundnuts and cassava and palm oil. Real things grown by real farmers.
| Metric | 1999 | 2007 | Change |
|---|---|---|---|
| Services GDP | $71B | $151B | +114% |
| Annual growth | n/a | n/a | 10.0% CAGR |
Banks, insurance, transportation, hospitality. The entire services sector more than doubled.
| Metric | 1999 | 2007 | Change |
|---|---|---|---|
| Manufacturing GDP | $21.0B | $23.1B | +10% |
| Annual growth | n/a | n/a | 1.2% CAGR |
Manufacturing grew modestly. Not the explosive growth of other sectors, but it grew. This matters because of what came later.
| Metric | 1999 | 2007 | Change |
|---|---|---|---|
| Mobile lines per 100 people | 0.02 | 26.4 | +131,700% |
In 1999, for every 5,000 Nigerians, one person had a mobile phone. By 2007, roughly one in four had one. [World Bank WDI / ITU]
In August 2001, the Obasanjo administration auctioned GSM licences. MTN, Econet (now Airtel), and NITEL's mobile arm won. [NCC / Pulse Nigeria]
This was a governance decision. The government could have kept telecom as a state monopoly like NITEL, or handed licences to cronies without auction, or created regulations that strangled competition. Instead, they held competitive auctions, licensed multiple operators, created a regulatory framework, and enforced the rules.
The fastest sectoral growth in Nigerian history followed. An industry that now employs hundreds of thousands and has built the infrastructure for mobile banking, fintech, and e-commerce.
Your phone, the one in your pocket right now, is a direct product of a governance decision made in 2001.
We ran lag correlation analysis to ask: how long does it take for governance changes to affect the economy?
The shorthand answer everyone reaches for is "five years." We did too, in the first version of this piece. We've stopped headlining the precise number. Picking the lag that maximises a correlation is a trap; you can produce neat results by accident if you keep trying delays. The honest read is that governance effects take years to show up in economic data, and the relationship is real on the long horizon. The exact timing is messier than a single number suggests.
The point still stands. When you vote for a president, you are not just voting for the economy you will experience under them. You are voting for the economy that arrives years later.
The economic growth of 2005 to 2011, the years Nigeria was "booming," was largely the delayed effect of governance improvements made in 1999 to 2003. The stagnation of 2016 to 2022 is the delayed effect of governance declines that began earlier.
The lag does not absolve sitting presidents. It gives them two responsibilities: full ownership of the future they are building, and partial ownership of how they manage the present they inherited.
Buhari inherited the oil price crash of 2014 to 2015. That was not his fault. But the response, delayed devaluation, import bans, forex rationing, those were his choices. The 2016 recession was not inevitable. Policy made it worse.
The lag means presidents inherit consequences, but it does not mean they are passengers. When we evaluate a president, we ask: given what they inherited, did they make it better or worse? And what did they build for the next generation?
We ran another analysis: which component of governance matters most for economic growth?
The World Bank measures six governance dimensions:
On the current data, Rule of Law correlates with Nigerian GDP at r = 0.72, the strongest of any single component. [World Bank WGI]
Not "fighting corruption," the thing every Nigerian president campaigns on. Not "effective government," the technocrat's dream. Rule of Law. Can contracts be enforced? Can property be protected? Can disputes be resolved fairly?
The unglamorous work of building institutions. It does not make headlines or win elections, but the data says it matters more than anything else.
Obasanjo is not universally loved. He is accused of:
These critiques are valid. This analysis does not erase them.
On the original World Bank vintage we used when this piece was first drafted, Obasanjo was the only Fourth Republic president whose tenure ended with a positive governance change. The data has since been re-baselined and the table now looks different:
| President | Governance change |
|---|---|
| Obasanjo | +0.20 |
| Yar'Adua | −0.13 |
| Jonathan | +0.21 |
| Buhari | −0.08 |
| Tinubu* | −0.01 |
* Tinubu's term is still running. Source: World Bank Worldwide Governance Indicators (current vintage) and the Aso Rock Index canonical record.
The honest update is that two presidents, not one, presided over real institutional improvement. Jonathan's term, on the current data, registered a slightly larger composite improvement than Obasanjo's, partly because his window includes the 2015 peak of the WGI series. Two presidents after Obasanjo left governance worse than they found it.
One more caveat, and it cuts against easy conclusions. The World Bank's governance scores carry a standard error of roughly 0.2 for Nigeria. Obasanjo's +0.20 and Jonathan's +0.21 are the only two term changes large enough to clear that noise. The other three sit inside a single year's error, which means, statistically, they are flat. The blunt reading of the whole table is that measured governance has barely moved in twenty-eight years.
What the algorithm still chooses Obasanjo for is everything around the governance line: the statistically significant 1999 break that kicks off the whole Fourth Republic, the fastest GDP growth rate of any president, the sector explosions, and the policy decisions, GSM chief among them, that are still paying dividends two decades on.
A flawed man presiding over the period of measurable institutional change with the largest visible payoff. That is the paradox.
We built a rough counterfactual. What if Nigeria had maintained Obasanjo-era growth from 2007 onwards?
The arithmetic:
If Nigeria had grown at 7.7% from 2007 to 2023 (sixteen years):
| Scenario | 2023 GDP |
|---|---|
| Actual | $520B |
| Obasanjo-rate growth | ~$1,040B |
| Difference | ~$520B |
Half a trillion US dollars.
It is not a prediction. It is a reference point, and a rough one. Treat the figure as an order of magnitude, not a number: priced properly against a peer benchmark, the gap centres in the hundreds of billions, but the honest confidence band is wide enough to include almost no effect at all. What survives the uncertainty is the direction, not the decimal. Choosing differently had a cost. How large, we cannot pin down.
Correlation is not causation. The algorithm found that governance and growth moved together under Obasanjo. It cannot prove that governance caused growth. Oil prices helped. Global conditions helped. Debt relief helped. The early 2000s were good years globally. China was booming, commodity prices were rising, and some of Nigeria's growth would have happened anyway.
Pre-1996 governance data is estimated, not measured. The WGI itself is perception-based, which means it can lag reality. And the data captures institutional outcomes, not the personal qualities of leaders. Obasanjo's military background, his networks, his political instincts. None of that shows up in a regression.
Stop voting on vibes. The data exists. Governance can be measured. Outcomes can be tracked. Demand that candidates be judged on institutional improvement, not slogans.
Rule of Law is the lever. Not anti-corruption campaigns. The boring, institutional work of making courts faster, contracts enforceable, and property rights secure. The telecom story proves it is possible. When governance decisions are made well, explosive growth follows. Nigeria has done it before.
Sitting presidents will be judged twice: once for how they handled what they inherited, and once for what they left behind. The lag is not an excuse. The Nigeria of 2030 is being built right now. The unsigned reforms, the weakened courts, the compromised institutions, the algorithm will trace it all back.
We ran 64 years of Nigerian data through statistical models.
The algorithm kept returning one finding: 1999 was the only statistically significant governance improvement in Nigeria's post-independence history.
The president who presided over that period, Obasanjo, also delivered the highest economic growth rate of the Fourth Republic, sector explosions across agriculture, services and telecom, and policy decisions like GSM liberalisation that still shape Nigeria today.
This does not make him a saint. It does not erase Odi, or the third-term ambition, or the selective prosecution of enemies. But when we remove human bias and let the data speak, one period stands out.
The algorithm chose Obasanjo.
Data sources: GDP from the World Bank World Development Indicators (constant 2015 US$). Governance from the World Bank Worldwide Governance Indicators (1996 to 2024 vintage). Extended governance (1960 to 1995) is a researcher-constructed proxy based on regime type, events, and academic literature.
Statistical methods: Structural break analysis (Chow test variants); Granger causality tests (VAR models, lags 1 to 5); lag correlation analysis; regime-based correlation analysis; component correlation analysis.
Limitations: Pre-1996 governance is proxy data. WGI is perception-based. Confounding variables (oil prices, global conditions) are not fully controlled. Sample sizes for individual administrations are small. Per-president governance and growth figures use the current Aso Rock Index canonical record and may differ from earlier versions of this article.