Speech
Opening Remarks – 13th Irving Fisher Committee Conference
Michele Bullock
Governor
13th Biennial Irving Fisher Committee on Central Bank Statistics (IFC) Conference
20 August 2026
–
Online
First up I would like to extend a warm welcome to all participants joining us for the 13th Biennial
Conference of the Irving Fisher Committee on Central Bank Statistics (IFC). I’m sorry I can’t
be there in person.
It’s great to see such a diverse group of statisticians, economists, data scientists, policymakers
and researchers from central banks, international organisations and academia gathered here in Basel.
I’d also like to thank our hosts at the BIS and all those who have contributed to organising this
conference.
As many of you’ll know, Irving Fisher was an American economist and statistician who was very
prominent in the late 1800s and into the 1900s. He understood that good economics begins with good
statistics. Long before the era of big data, he recognised that progress in public policy depends on our
ability to measure economic phenomena accurately, consistently and objectively.
But Fisher’s work also reminds us that statistics are not an end in themselves – indeed, he
cautioned that ‘mere empiricism is seldom very fruitful’. 1
Rather, statistics are a tool and
need to be combined with economic frameworks and judgement. To implement good policy, we need both facts
and interpretation.
The work of the IFC reflects many of the principles that motivated Fisher’s own research: a
commitment to careful measurement, a willingness to embrace new methods, and a belief that better
information leads to better decisions.
The tools that are available now for data construction and analysis are dramatically different than they
were 100 years ago. But beneath all the new technologies and new data sources lies the same
fundamental challenge that economists and statisticians always face: how do we construct, transform and
utilise data to help us answer policy questions.
That challenge sits at the heart of both the IFC and the discussions we will have over the next two days.
This conference is being held in a challenging time for central banks, not only in terms of technology but
also the global geopolitical and economic environment. Because policy is being set in a more uncertain
environment, policymakers require faster insights, more granular data and a deeper understanding of
increasingly complex economic and financial systems. At the same time, technological change, particularly
in artificial intelligence (AI), is opening up new possibilities while also creating new challenges.
Against that backdrop, I would like to focus my remarks on three topics today:
the role of the IFC and some of its recent work
the IFC’s new strategic direction
the key themes that will shape our discussions over the next two days.
The role of the IFC
For nearly 30 years, the IFC has been a forum for collaboration on statistical issues across the
central banking community.
Its objective is straightforward: to promote the exchange of views on statistical and data issues of
interest to central banks and to foster a global community of practice among economists, statisticians
and data professionals.
Historically, this work focused primarily on official statistics. But the environment in which central
banks operate has evolved considerably over recent years.
Reflecting this, the IFC’s Terms of Reference were updated in 2025. The revised mandate explicitly
recognises that the interests of central banks extend beyond traditional statistics to encompass broader
issues related to data, data governance, technology and the use of AI in central banks.
Central banks today are both major producers of official statistics and intensive users of data. We face
common questions about data quality, governance, interoperability, privacy, security and increasingly the
responsible use of AI.
The IFC provides a unique forum where members can learn from one another’s experiences, share
practical solutions and help develop common approaches to emerging challenges.
Over the past year, the IFC’s work has focused on four major areas: microdata, data science and AI,
data governance, and sustainable finance.
On microdata, the IFC has contributed to international efforts to improve access to, and use of, granular
information. This includes work on corporate financial statement data and broader initiatives to improve
standards, interoperability and data sharing.
On AI and data science, the IFC has continued to provide a forum for central banks to exchange practical
experiences. Recent work has examined the use of generative AI in areas such as information retrieval,
text analysis, supervision, statistical production and policy analysis. The IFC also published a report
on a membership survey of AI adoption across central banks.
On data governance, the IFC has explored how institutions can govern data more effectively, including how
governance frameworks support trustworthy AI, data stewardship and the integration of multiple data
sources.
And on sustainable finance, the IFC has worked with international partners to address climate-related data
gaps and improve the measurement of risks and exposures relevant to central bank mandates.
Taken together, this work reflects an important reality: good policy increasingly depends on good data,
and good data increasingly depends on effective governance, modern technology and strong international
collaboration.
The IFC’s new strategy
As the IFC’s mandate has broadened, it has also become clear that we need a sharper focus on what we
are trying to achieve. And so the Executive Committee has been workshopping a strategy to guide the work
of the IFC for the next three years. One important piece of input to this has been the survey of IFC
members that was conducted at the end of last year.
The result is a new strategy centred on a simple idea: data for the future. It is about strengthening
statistical capabilities, advancing data innovation and supporting robust data governance.
The strategy identifies three linked priorities.
The first is
modernising central bank statistics
.
Many of our members are involved in producing statistics. And the survey told us that members continue to
place high value on improving statistical production, measurement and dissemination.
A key challenge, however, is how to combine traditional statistical approaches with new information
sources and analytical techniques. The objective is not to replace official statistics but to strengthen
them.
The second priority is
AI and data innovation for central banking
.
The membership survey was clear on one point: AI is now the highest priority innovation topic across
central banks. Governors and senior policymakers want to understand not only the opportunities, but also
the practical realities of adoption and the risks.
While most central banks are already experimenting with AI, some are more advanced than others in using it
for policy purposes. But all are increasingly moving towards implementation, and they are looking for
learning from peers.
The IFC can play a valuable role by helping members learn from one another’s experiences and by
identifying practical, replicable use cases.
The third priority is
data governance and the future data ecosystem
.
The quality of our analytical work and our policy decisions ultimately depends on the quality of the
underlying data. We therefore need to ensure that we have strong governance around our data. We need to
be sure that the data are correct and can be trusted. In practice, this means we need policies and
frameworks that provide assurance that the data can be trusted, that we are applying consistent standards
and that there appropriate guardrails and transparency around the use of AI.
Importantly, the IFC has recently brought the Central Bank Data Collaboration Group (CBDCG) under its
umbrella. The CBDCG is a central bank network of data leaders that focuses on data governance, analytics,
AI and data capability. The IFC and the CBDCG have overlapping interests. Hosting it through the IFC will
strengthen collaboration, avoid duplication and, at the same time, preserve operational autonomy.
The strategy will be implemented through a small number of focused priorities where members want
collaboration. And we are also looking to gauge whether there is interest from member central banks in
participating in limited scope working groups on these topics.
Conference themes
The priorities set out in the new strategy are reflected strongly in the program before us over the next
two days.
The conference theme is ‘Filling information gaps in central banking: effectively leveraging data and
innovative tools’.
Information gaps have always existed. But as technology has improved, we have acquired new ways to address
them.
Three broad themes stand out across the papers.
The first is the growing role of AI and advanced analytics.
Our opening session examines how central banks are using AI to support policy analysis, improve data
discovery, monitor risks and enhance operational effectiveness. This theme is evident in many other
sessions too where we’ll hear further examples of how AI is being applied to central banking
problems.
The second theme is the use of novel and more granular data.
Many presentations explore how non-traditional sources can fill data gaps and help us understand economic
developments more quickly. For example, payments data, firm-level information, market data and even
satellite imagery. These approaches have the potential to fill longstanding information gaps and improve
the timeliness of policy insights.
The third theme is trust.
Whether we’re discussing inflation, financial stability, sustainability, globalisation or AI, the
underlying challenge remains the same: ensuring that data are reliable, understandable and fit for
purpose.
Innovation matters. But confidence in the integrity and quality of our information matters just as much.
As central banks, our credibility depends on it.
Conclusion
In the preface to
The Purchasing Power of Money
from 1920, Fisher said: ‘Only by knowledge,
both of the principles and of the facts involved, can such fluctuations in future be prevented or
mitigated’. 2
He was writing about price-level instability, crises and depressions. But I think it makes a broader
point – policy without adequate facts, runs the risk of making preventable mistakes.
While today’s central banks have access to vastly more information than Fisher could ever have
imagined, the challenge is much the same. Central banks need data that are accurate, well governed and
useful for policy decisions.
The IFC exists to support that objective.
I hope this conference provides an opportunity to share ideas, challenge assumptions, learn from each
other’s experiences and identify new ways of addressing the information needs of central banks.
Thank you for your participation.
I wish you a productive and engaging conference.
Endnotes 1
Fisher I (1925), ‘Our Unstable Dollar and the
So-Called Business Cycle’,
Journal of the American Statistical Association
,
20(150), pp
179–202
. 2
Fisher I (1920),
The Purchasing Power of
Money
, Macmillan Co, New York.