As Prepared for Delivery
Good morning. Welcome to the 6th annual IMF Statistical Forum,
where we give new meaning to the term “data mining.”
Thanks to all of you for joining what promises to be very interesting
discussions. And my appreciation to Louis Marc and his team for their
efforts to bring this conference to fruition. In the halls of the IMF, the
statisticians are often the unsung heroes. Our work is all the better for
the efforts of the Statistics Department, and this event gives them the
opportunity to take center stage.
A year ago, this conference delved into the measurement of GDP in
the digital age. We talked about addressing the statistical challenges we
face in an economy where e-commerce, mobile finance, and impossibly fast
data crunching are achieving dominance. This year we take our discussion to
the next level with the topic “Measuring Economic Welfare in the
Digital Age: What and How?”
Measuring welfare is not a new challenge. Almost 60 years ago, Paul
Douglas—a chairman of the Joint Economic Committee, and who you may
know as the Douglas in the Cobb-Douglas production function —was quoted
as saying that “a large part of satisfaction cannot be plumbed by
figures.”
Simply put, traditional indicators don't show the true size of the economy,
and they don’t do a good job of gauging how well we are doing. In other
words, they fall short in measuring welfare.
This shortcoming has economic and political ramifications in an era in
which policymakers are facing the challenges of inequality and inclusive
growth. And the growth of digitalization only raises the stakes. An IMF
Policy Paper published earlier this year on Measuring the Digital Economy describes the challenges
digitalization poses for measuring macroeconomic statistics, and points to
the need to have enhanced welfare indicators.
So the need to measure economic welfare in the digital age is clear. The
question then becomes, what to measure, and how. These
are more difficult to answer, particularly in light of the limited
statistical capacity of many countries and institutions.
There are calls for new measures such as “GDP-plus” that would capture the
positive welfare effects from the digital economy. Such measures can help
guide policies that can both foster innovation and manage how technology
affects daily life. For example, knowing the welfare effects might help
clarify the regulatory tradeoffs between data privacy and incentives for
innovation-driven growth. Teleworking in traditional employment and in “gig
economy” jobs through online platforms is changing the ways in which many
of us work, especially young people and women with children. In African
countries, mobile money has brought millions into the market economy,
helping women to start and grow small businesses.
We need to take into account the implications for household nonmarket
production and gender participation in the work force. This means both the
size of the economy and how it affects daily life.
In addition, free and nearly-free services from digital platforms are not
measured in household consumption even though they have reduced the cost of
living and raised the standard of living. We can debate how much the iPhone
has improved the quality of life. I’m not sure how beneficial it is to have
Solitaire at our fingertips 24 hours a day.
But I like having free video calls to my one-year-old grandson in
Texas. I would add that the savings that consumers reap from these
services suggest an overestimation of inflation in consumer price
indices and underestimation of productivity growth.
This takes me back to the point on what to measure and how. Technological
improvements have clearly improved welfare and our quality of life, but it
is not clear how much they do so, and how to measure the change. After all,
as Robert Gordon points out, “some inventions are more important than
others” and the digital revolution may not give as much of a boost to our
quality of life or productivity from as earlier innovations.
That is good motivation for our conference this year, where we will aim to
advance our understanding of the issues surrounding measurement of economic
welfare. Among the questions to be considered this year are:
why does welfare from consumption grow more than consumption itself in
the digital age?,
what is the value of free software, free online services, and digital
data?, what are the highest priorities?, and what is feasible to
measure and realistic to ask national compilers?
In 2008, Joe Stiglitz, Amartya Sen, and Jean-Paul Fitoussi coordinated a
study commissioned by the French government that focused on inclusive
growth and sustainability. The relevance of these measures to the work of
the IMF is shown by findings in IMF research over the past decade that
higher levels of inequality are associated with less sustainable economic
growth and less financial stability.
We already can glean considerable information about welfare by paying more
attention to measures of real income, consumption and wealth included in
such international standards as the System of National Accounts. And big
data is opening doors to granular levels of information.
So, we are seeing progress—some of which we will hear about in
presentations at this conference in the next two days. Today we will
address conceptual frameworks for measuring welfare beyond GDP; the current
state of play and the way forward for official statistics; the value of
unpriced software and data; and measuring welfare growth. Tomorrow we will
take a more detailed look at e-commerce and fintech. The last session will
ask whether all is to the good in the digital age, with presentations on
how technology could raise inequality and challenge development
opportunities.
Today’s session on official statistics is especially useful because we need
to keep in mind what is practical for statistical agencies around the world
to pursue, given limited resources and other constraints. It is our
responsibility to work with these agencies to operationalize some of this
research.
This is a milestone event on a cutting-edge topic. How we come to
understand the interaction of technology and welfare will have implications
for the very accuracy of data on growth, productivity and inflation. It may
help to improve the quality of life over time.
In sum, this is highly relevant to the work of the IMF and likely will grow
in importance across the field of economics in the coming years. We look
forward to working with you. Thank you.