The National Institute of Statistics of Rwanda (NISR) has long been known for its censuses, household surveys, and economic measurements. But it is evolving, with a shift from traditional statistical production to increased reliance on big data or alternative sources of data.

To understand this shift, we sat down with Ivan Murenzi, Director General of the NISR. Our conversation made it clear that the real story is about people, their skills, their adaptability, and their willingness to rethink how official statistics are produced.

How has the national demand for data evolved over time, and what role does the NISR play in meeting these demands?

Initially, we focused on basic statistics across sectors. We were mostly guided by national priorities, medium and long-term development plans that define aspirations across sectors. These plans, along with sector-specific strategies, helped us identify data gaps and prioritise accordingly. For example, in health, we did demographic and health surveys; in poverty and social well-being, other specialised surveys. These are largely standard globally, with local adaptations for context.

Around 2015, the shift was to increase the frequency of data collection and analysis. Some indicators like employment or inflation cannot wait three years. They must be measured quarterly, or even monthly. We started investing in more regular data collection to keep up with decision-making needs.

But producing data more frequently also revealed new challenges: the need for faster methods, richer sources, and above all, more flexible people. That pressure pushed us to explore new methods. At the same time, Rwanda’s investment in digitisation made using administrative data a viable alternative.

The NISR’s outputs have been increasingly used in the last seven years. We are also seeing growing engagement from people wanting to understand the statistics. For example, after the last national census in 2022, we received many requests from various ministries but also from different districts. People wanted us to explain the data findings and expand on the implications for their sector or district. The demand was bigger than we could meet.

Can you give an example of where alternative data has shown real promise?

Satellite images are particularly exciting. They allow us to see how weather patterns affect agriculture and could eventually reduce the need for extensive [data collection] fieldwork. We work with the Rwanda Space Agency and others. It’s still exploratory, but promising.

We take time to confirm consistency because what we release must be reliable. By the time we say, “this is the national picture”, we must be certain.

All our national plans, including Vision 2050 and NST2, are now quantified with clear baselines and targets. Sector plans – like in agriculture – are much more detailed, tracking inputs, productivity, and practices.

How are staff adapting to these shifts?

Technology has helped; moving from recording data on paper to tablets has reduced human error. Today, field data is transmitted live, with GPS verifying the location of enumerators in the field. Automated quality checks prompt real-time corrections to information being captured. On the analysis side, we’re adopting better programming tools, enabling efficient draft report generation and creating more time for review.

We plan to set up an engagement and communications department. Until now, a few staff have handled that, but we know the need for a fully-fledged department. Like Padraig Dalton (the former head of the Central Statistical Office in Ireland and a consultant to the NISR) emphasised, having someone translate data into accessible narratives makes a big difference and we see the same here.

Around the world, statistical offices are balancing established expertise with the demands of data science. NISR’s response has been to invest heavily in people: creating a Big Data and Data Science Department, training analysts, and hiring top graduates for intensive mentoring.

What are some of the other institutional changes required?

We’ve proposed a full institutional restructuring. We realise that the way we are configured now does not respond to where we are going; we will have two new departments dedicated to administrative data.

And that restructuring does not stop with the statistics office. We’ve also proposed creating data analysis units across ministries – staffed by people accountable to the NISR. These units would ensure alignment and quality. The shift to administrative data means working with other ministries from the start, helping design their systems so the right data is captured.

The innovation, big data, and national impact aspects are drawing in young people who are highly motivated. That’s a real advantage. We recruited 37 top graduates last year, trained them intensively, and paid them well. They learn our methods and values before being seconded to ministries.

Staff retention is still a challenge however; many staff get picked up by agencies, international organisations, or local private firms due to their skills and bilingualism. So, we’re building a pipeline, creating a clearer career path within NISR and continuously training new cohorts. We’ve asked the relevant authorities to allow other ministries/agencies to request secondments from us directly. That way, we can quickly place trained people who understand our approach.

This will help us retain talent and allows us to fill vacancies in the pool we’ve trained in without lengthy external recruitment processes. It’s efficient and ensures cultural fit.

What are your personal priorities for the next five years?

Engagement is a major one. We want senior staff to spend more time helping ministries, academia, and journalists understand and use our data. At the same time, we must complete our restructuring and make major headway in shifting from primary to administrative data. That transformation from reliance on surveys to reliance on integrated systems is what I want to achieve.