Kweku Opoku-AgyemangWe predict everything - weather patterns, disease risks, grades in school and more. AI has accelerated this trend, finding patterns in enormous amounts of data and using them to make increasingly sophisticated predictions.
But there is a question that prediction alone cannot answer:
Knowing what will happen is one thing. Knowing what happens when we intervene is another.
That distinction matters enormously for Africa, where development is not simply about understanding what is likely to happen. It is about changing what happens for the better.
Imagine a city where umbrella sales and traffic accidents both increase on rainy days. Just because both happen at the same time doesn’t mean one is causing the other. Obviously umbrellas are not causing most of the accidents. Rain is affecting both.
The lesson seems obvious. But much of the challenge in development is usually less clear: two things can move together for perfectly real reasons without one causing the other. The harder question is what happens when we actually change one of them.
In development, we are constantly surrounded by correlations. Countries that invest more in education may or may not grow faster. Regions with better roads may or may not have higher incomes. Experts have to decide where their resources would have the best result.
But observing that two things occur together does not tell us what would happen if we changed one of them.
That is where causal reasoning becomes important. A useful starting point is to ask two increasingly difficult questions.
First: What happens if we intervene?
If a government builds a road, changes a tax, or expands a school programme, what effect does that intervention actually have?
And second: What would have happened otherwise?
This last question is difficult. We can observe the world we live in, but we cannot directly observe the alternative world in which the intervention never happened. Yet that alternative is often exactly what African policymakers need to know.
Observation alone is not enough
Suppose an African expert introduces a programme designed to help farmers produce more food and harvests rise afterward. That is encouraging, but it does not automatically mean the programme actually caused the increase. What if it was a coincidence? Perhaps rainfall was unusually favorable. Perhaps crop prices rose. Perhaps farmers were already becoming more productive.
To understand the effect of the programme, we need to ask a counterfactual question: How much better — or worse — would these farmers have done without it?
We can see this logic in research well beyond agriculture. Consider a study of low-income adults in Ghana during the COVID-19 challenge. Researchers wanted to know whether being able to communicate more easily could improve people's well-being in that complex time.
Rather than simply observing a relationship between communication and mental health, they provided some random people with mobile calling credits and compared their mental health outcomes with those of a control group that didn’t have these extra credits.
The only difference between the two groups was the fact that one had credits and one didn’t. The intervention was followed by a significant reduction in mental distress.
The lesson is broader than the specific intervention: when we want to know whether changing something will change an outcome, observation alone is not enough.
We need to ask what happens when we actually intervene. It changes how we think about development.
For years, development debates have sometimes been framed around predicting what Africa will look like decades from now: how fast populations will grow, how many jobs will be needed, how large cities will become, or how climate change may affect economies.
Those forecasts can be valuable. But a forecast is not a strategy.
If someone predicts that millions of young Africans will enter the labor force, the important question is not simply whether that prediction comes true. It is what we can do today to change the opportunities those young people encounter tomorrow.
That requires asking a different kind of question:
What can we change to produce the outcome we want?
What can we change?
This is where Africa's development challenge becomes an opportunity. The continent is not merely a place where predictions about the future are tested. It is a place where the future is still being built.
New transport networks can change where businesses locate. Digital payments can change how people transact. Energy infrastructure can change which industries become viable. Better agricultural technologies can change what farmers can produce. Education can change the skills available to an economy.
Each intervention creates possibilities and sometimes consequences that are difficult to see in advance.
The goal, therefore, should not be to abandon prediction. Prediction remains enormously useful. But we should stop treating prediction as the final destination.
The most important question is often not “What will happen next?”
It is: “what can we do to make something different happen?”
That shift from forecasting the future to understanding how we actually change it is one of the most important ideas in modern development.
Africa does not need only better predictions about the future.
It needs better ways of building it.
Disclaimer: The views expressed by the author do not necessarily reflect the opinions, viewpoints and editorial policy of TRT Afrika.
