IL40: Why the Economy Feels Broken... Even When It’s Growing ft. Diane Coyle

What if our most trusted economic statistic is pointing us in the wrong direction? Diane Coyle joins Kevin Coldiron to explore why GDP - long treated as a proxy for progress - now obscures more than it reveals. As economies shift toward services, intangibles, and unpaid digital labor, much of today’s value creation falls outside the frame. Drawing on her new book, The Measure of Progress, Coyle makes the case for a new way of seeing - one that captures time, trust, and the real foundations of growth. The question isn’t how fast we’re moving. It’s whether we’re measuring the right road.

-----

50 YEARS OF TREND FOLLOWING BOOK AND BEHIND-THE-SCENES VIDEO FOR ACCREDITED INVESTORS - CLICK HERE

-----

Follow Niels on Twitter, LinkedIn, YouTube or via the TTU website.

IT’s TRUE ? – most CIO’s read 50+ books each year – get your FREE copy of the Ultimate Guide to the Best Investment Books ever written here.

And you can get a free copy of my latest book “Ten Reasons to Add Trend Following to Your Portfolio” here.

Learn more about the Trend Barometer here.

Send your questions to [e-mail]

And please share this episode with a like-minded friend and leave an honest Rating & Review on iTunes or Spotify so more people can discover the podcast.

Follow Kevin on SubStack & read his Book.

Follow Diane on X and read her book.

Episode TimeStamps:

02:19 - Introduction to Diane Coyle

03:55 - How did the current systems of national accounts came to be?

06:55 - Why the national statistics doesn't add up

11:22 - The underlying problems of GDP

18:42 - How software pose a problem for measuring GDP and economic activity

23:25 - The challenges of cloud computing

26:39 - What is disintermediation and why is it a problem for economic...

Episode Transcript

(00:07):
We're in an interesting timebecause the system of national accounts
gets revised every 10 years orso. We've just had a revision earlier
this year, but almostimmediately the Secretary General
of the UN announced a newcommission to look at how to do better

(00:27):
economic statistics becausethere is, I think, a deep sense of
among many populations thatthe GDP growth figures don't reflect
their experience. GDP growthhas been going up. A lot of people
feel their lives have not beengetting better. So,the popular demand,
if you like, for a differentapproach to measurement is definitely
there.
Imagine spending an hour withthe world's greatest traders. Imagine

(01:09):
learning from theirexperiences, their successes and
their failures. Imagine nomore. Welcometo Top Traders Unplugged,
the place where you can learnfrom the best hedge fund managers
in the world so you can takeyour manager due diligence or investment

(01:29):
career to the next level.Beforewe begin today's conversation,
remember to keep two things inmind, all the discussion we'll have
about investment performanceis about the past, and past performance
does not guarantee or eveninfer anything about future performance.
Also, understand that there'sa significant risk of financial loss
with all investment strategiesand you need to request and understand
the specific risks from theinvestment manager about their products
before you make investmentsdecisions. Here's your host, veteran
hedge fund manager Niels Kaastrup-Larsen.
For me, the best part of mypodcasting journey has been the opportunity

(02:09):
to speak to a huge range ofextraordinary people from all around

(02:35):
the world. In this series, Ihave invited one of them, namely

(03:00):
Kevin Coldiron, to host aseries of in-depth conversations
to help uncover and explainnew ideas to make you a better investor.
Inthe series, Kevin will bespeaking to authors of new books
and research papers to betterunderstand the global economy and
the dynamics that shape it sothat we can all successfully navigate
the challenges within it. Andwith that, please welcome KeviCnoldiron.

(03:44):
Okay, thanks Niels. Andwelcome everyone to the July 2025
episode of the Ideas Labpodcast. I thought I'd start today's
show with two quotes from theauthor of the book we're going to
be discussing. Quote numberone, “The more research I've done
on economic statistics, theless certain I am that we know anything
solid about today's economy.”And quote number two, “The invisibility
of the economy, as it is now,and the statistics available, is
extraordinary. Well,todaywe're going to delve into why this
is the case. Why someone whohas spent her career studying economic
statistics, and has evenwritten a book called GDP, A Brief
but Affectionate History, isnow warning us that the system of
accounts we use to produce GDPand to track economic progress is
no longer working for us.Ourguest is Diane Coyle. She is
the co-director of the BennettInstitute for Public Policy at the
University of Cambridge. She'sa former Advisor to the UK Treasury.
She was awarded a CBE in 2018for her contribution to the public
understanding of economics.Andshe's here today to talk about
her latest book, one of manythat she's written, and it's called

The Measure of Progress: Counting what Really Matters. So, (03:45):
undefined
Diane, thanks for joining ustoday and welcome to the show.
Thank you for the invitation.I'm always really excited when somebody
gets interested in economicmeasurement questions because I think
it's so important.
Well, it is important and it'sfundamental to so much of our lives.

(04:17):
And I think it's even moreimportant right now with everything
that's going on. So, I'mreally excited to dig into this.
Now,your book is split intothree parts. There's the first part
that kind of documents all theways in which the modern economy
is kind of invisible orpartially invisible in economic statistics.
The second part of the booktalks about why that matters to us,
and then the third suggests anew framework that's designed with
a modern economy in mind. AndI’d kind of like to have the conversation

(04:46):
mirror that, but I thought wecould start by setting the stage.
Canyou explain to us how thecurrent system of national accounts,
which is what we use tocalculate GDP and all its components,
how it came into being, andthe nature of the economy that it
was originally designed to measure?
Of course. So, about a decadeago I wrote a book called GDP: A

(05:25):
Brief but AffectionateHistory, as you said. And one of

(05:46):
the insights I got from thatwas that we've thought about the

(06:12):
economy in very different waysover time. So, GDP is an idea. It

(06:39):
isn't a natural object of anykind. Andit's an idea that emerged
starting in the late 1930s andcrystallized during the Second World
War and immediatelyafterwards. The ambition was to have
a sense of the aggregatecapacity of the economy to produce,
and what that meant wasavailable for consumption, but during
the wartime years, what wastherefore available for wartime production
as well. And so, it's verymuch a creature of that environment,
of the importance ofmanufacturing production at the time.
Andit was cemented intoplace, in the immediate post war
years, through a processinvolving the United Nations built
for economies that were, to amuch greater extent than now, manufacturing
economies or still, in lowerincome cases, agricultural economies.
So, very material sets ofideas. So,one of the points I make
in the new book is that'sbecome less and less suitable, over
time, as the economic valuethat's being created is more and
more intangible rather thanmaterial. We still use materials,
of course, more than ever. Butthe value that's captured by the
GDP metrics is intangiblevalue. So,the three word statistics
is derived from the same rootas state. So, the framework that
we used is how statesunderstand the societies that they
are governing. And that's whatmakes them important, that's what
they're used for policy. Andthey’re used to understand how we
think things are going in oursociety in an aggregate sense.
You make a very importantpoint that the current system “adds

(07:03):
up” and that users of economicstatistics, like myself, can rely
on that. But you also say thatit doesn't add up when we take inflation
into account. And that's veryimportant. Canyou just explain the
kind of adding up concept ofnational statistics? And then why,
when we sort of adjust forchanges in prices, that doesn't hold?

(07:29):
I'm really glad you raisedthat. I think it's a very important
point that not even manyeconomists think about enough. If

(08:04):
you are trying to measure theeconomy in terms of dollars, you
can do that by adding up allof the money people spend (the expenditure),

(08:46):
adding up all the money theyearn (the incomes), or adding up
everything that gets produced.And there's a lot of kind of detailed

(09:25):
adjustment needed, but inprinciple those are the same. So,
we've got this, in effect,quadruple entry bookkeeping that

(09:56):
goes on. Butwhat economistsdo is then divide price indices or
deflators into each of thosecomponents to put them into, what
we would call, real terms,which is a misnomer because these
are not real things at all.They're even more conceptual than
the dollar figures that wemeasure. Andonce you apply the right
price index to each component,then the things that you get out
are not going to add up atall. In fact, very often it's not
even obvious what kind ofnumber you've created by using a
price index to turn into thisconcept of real value. Thereason
for doing it is that we wantto understand, in some quite fundamental
sense, what economic value isbeing created and who's getting it.
But the idea of the priceindex comes from trying to understand
what level of prices will keepa consumer's utility constant. Soit
goes back to the utilitarianfoundations, economics, and this
idea of constant cost ofliving. And the price indices have
always been incrediblypolitical. We've Seen that recently
from 2022 on, with theinflation surge because it has distributional
consequences in all kinds ofways. It undermines people's standards
of living. Andyet in thinkingabout economic growth and productivity
or market structures, we'reusing these slightly odd constructs
that are often just indexnumbers rather than units of cars
or washing machines oranything. And not enough thought
is given to what on earth itis we're trying to do with these
price indices. Andif youreally want to understand, in some
normative sense, how people'slives are going, is this the right
way to go about it with thisobscure technical set of index numbers
that nobody really understands?
Yeah, well, thanks for that.And I do want to talk about inflation
a little bit later, but I haveto say, when I read your section

(10:26):
on that, the more I read, themore I thought, do we even really…
Is there even really a conceptof inflation? I mean, there are so

(10:52):
many different ways that youcan measure it. Andyou make the
case that different people aregoing to have different inflation
rates depending on theirincome or their age because they
buy different stuff and theprices of that stuff changes. And
even they buy… You and I mighteven have the same “consumption basket”.
We might, at a high level, buythe same things, but one of us might
buy a lower quality productand one of us might buy a higher
quality product. And so, asyou said, it is almost inherently
a political decision, when wechoose a particular inflation index
to measure across the entire economy.
There are really deepproblems. So, there's an assumption,
in constructing the priceindex, that all of the goods that
you put in the price indexsubstitute for each other. So, if
the price of your food goesup, but you're getting TikTok for

(11:14):
free, and they might cancelout in some way, but of course, you
can't eat your TikTok videos.
Although I’m sure some peoplewill have tried. So, let's, okay,
going back to the issue of GDPand some of the problems with it.

(11:46):
I mean, you said that it wasdeveloped in a world where physical
capital dominated a materialworld of industrial production and

(12:09):
agriculture. And one of thesections of the book is literally
called dematerialization. Iknow this is an area that you've

(12:32):
written a lot about. So,Ithought maybe we could start with
that in terms of some of theproblems with GDP in the kind of
economy that we're operatingin now. And I thought we could use
one of the examples in yourbook of Rolls-Royce, since I think
that's a salient example and Iknow you've spent some time with
them. So,you know,Rolls-Royce is a company that makes
sophisticated turbines, butyou say that actually most of their
profit, two thirds of theirprofit, comes after the sale of the
turbine has been made. And Iwas wondering if you could just explain
to the listeners how thatworks and then why that's an issue
for the statistics and for GDP.
I've been to visit theirfactory in Derby in the Midlands,

(13:07):
here in England, and it's afantastic site. It's huge. They have

(13:39):
all these impressive windtunnels and the things that you would

(14:08):
expect you need to constructturbine engines. And in one corner

(14:38):
of the site, there's a hutwith rows of people inside sitting

(15:10):
at their screens, just likeany other office in the world. And

(15:42):
that is the driver ofRolls-Royce's revenues and profits,

(16:06):
because they sell servicesaround the engines. Ifyou operate
an aircraft, you might buy aplane with Rolls-Royce engines. And
you'll buy the services tomonitor them in real time so you
can be tracked, and problemscan be spotted immediately and sorted
out really quickly. Andso,there's a service relationship whereby
the airlines can rely on theengine maker, who knows, in much
greater detail, how tomaintain and operate the engine safely,
and it takes away the hassleof having to have your own maintenance
for engines if you run theairline. So, it's a positive sum
deal. It'scalledservitization, which is a rather
unlovely word for thisprocess. But there are quite a number
of what you would think of asmanufacturing companies, now, that
are making a lot of theirrevenues from services of different
kinds. And you see it in theeveryday economy as well. Often,you'll
see people advertisingsolutions. Office furniture solutions
means they'll sell you a chairand desk, but they'll also maintain
it, and replace it when itgets broken. And so, these sorts
of relationships between themanufacturer and the customer are
occurring more and more.So,it matters in a number of ways.
One is that we think of thedeclined manufacturing as being a
significant problem ineconomies like the US and the UK,
and in some ways it certainlyis because of a loss of engineering
expertise that goes withhaving outsourced all of your manufacturing
overseas. Butthe figuresprobably overstate the extent to
which manufacturing hasdeclined because of things like these
service activities taking up alarger share of revenues and profits.
And, you know, there's been alot of outsourcing in manufacturing.
Thereare other forms ofregeoning of production patterns
away from the conventional.The manufacturer designs and makes
things, and then thewholesaler and retailer sells them,
and then somebody elsemaintains them. So, contracting out
has become a really familiarpart now of the economic landscape.
And that means that companieslike Apple or Nike don't manufacture.
So, we might think of them asmanufacturers, but they're not. Differentcountries
would classify those indifferent ways in their statistical
system. So, if you want tocompare manufacturing sectors across
countries, you would reallyneed to go into the detail to understand,
well, where did they put thiscompany? Which sector did they put
it in? Andthen, of course, wemight get on to the question of what
does this imply about supplychains and important export figures?
Because depending on whetheror not the ownership of the components
that go into a finishedproduct changes or not, they get
counted as imports andexports, or they just get counted
as intra company transfers.So,interpreting the production structures,
because they shifted in waysthat don't map onto the old categories
that we have, it's becomereally difficult to get a really
good handle on what actuallyis the production capacity of a national
economy? Where are the profitsand wages being generated from, and
what actually is happening tothe trade balance?
So, is this an issue just withRolls-Royce? How are they then classified?

(16:31):
Are they classified as amanufacturer? Are they classified
as a service company? Aretheir activities kind of somehow
split between those sectors?
For that particular company, Idon't know, because obviously company
data is confidential. But theprinciple is usually that you ask
a company to identify itself,to identify its major sector of activity,

(16:55):
and so they might choosedifferent categorizations for themselves.
I see, yeah. So, is this aquestion, you know, of classification
rather than scale? In otherwords, is this an issue, this particular
issue (we'll talk about otherones), is this like, hey, we're still
measuring the aggregate sizeof the economy correctly, but we're

(17:19):
not measuring kind of whereit's coming from? Or are there both
issues happening here?
So, just as any accountantwill tell you to look at a company's
cash flow, I think if you lookat the dollar amounts, we've got

(17:48):
a pretty good handle on whatthat is. It's when you want to understand
what's happening at the levelof particular industries or particular
companies and types ofactivity. There, I think, we really

(18:15):
don't have a good handle onwhat's been going on. And part of
it is about digitalization andthe way that's changing where value
is added and how activitiesget categorized. Oneobvious example
is all the free services thatwe have on our smartphones now, all
those free search and emailand so on. They’re incredibly valuable
to people. There's clearlyeconomic value being created by them.
But because they have a zeroprice, it's not obvious how to put
those into the quadruple entrybookkeeping system. Youcan count
the wages of people who workfor those companies. Maybe you can
trace the profits, althoughthat's a bit complicated because
of the way companies movetheir profits around. But how do
you count that in the headlineGDP figures when there's no price

(18:36):
to add in?
Yeah, now since you've broughtit up, that's another section of
your book. Kind of all thefree services we get. And maybe we

(19:00):
could, since you brought itup, we could talk about that a little
bit more right now.Thingslike Gmail that everyone uses
or open-source software. Howdo those things pose a problem for
our measurement of GDP andeconomic activity?
The problem at heart is thatthey're valuable things and we don't

(19:39):
know how to count them becausethey've got a zero attached to them.

(20:08):
And they're also substitutingfor things that people use to pay

(20:33):
for. Open-source software,which many economists would now use

(20:53):
to do our work. RorPythonreplaces the paid for proprietary
software that everybody usedbefore. People will spend their dollars
on something else. So,thedollar figure for GDP might not be
much affected, but there areobviously things that make it complicated
to understand what's theproductivity of the software producing
sector? Is it going up ordown? Do we want to count those free
activities in it or not?Andit would be interesting just
to know what is the value ofall of the unpaid open-source effort
that people are putting intothe economy. And it's everything
from software to both consumersoftware and the kind of software
that underpins a lot of theInternet. Butalso, all of the free
entertainment we get nowbecause people are creating fantastic
content that they're puttingonline. A lot of what's called user
innovation or things likecreating new medical devices or sports
devices, which might gain acommercial market eventually, but
when they start, they're aniche thing and enthusiasts or people
who need medical attentionshare them online freely. This is
valuable stuff. We might wantto know what's going on. Wemight
even want to shape policies toencourage some of these activities,
but we've got no way ofknowing how much or what exactly
is being done.
Like this podcast, right. Itincreases everyone's wellbeing. But
it's free.
Absolutely.
One thing I thought about inthe “free section” was that we are

(21:14):
of course paying for thingslike Gmail because we're handing
over our data. And as you say,you know, the most valuable companies
in the world are “giving away”products for free. So obviously,
there's value and we'reexchanging something. How does that
exchange of data… Presumablythat also doesn't get reflected in

(21:36):
GDP.
No, that's right. I don'tthink there are any good figures

(21:58):
on how much data is being usedor what's going in and out of data

(22:26):
centers over thetelecommunications networks. It's
obviously increasing a lot.So, it is creating value. What isn't

(22:57):
clear to me is the extent towhich that is the result of market

(23:23):
power. Andso, if you've gotan advertising funded service and
there are two companies,Google and Meta in UK markets, they
dominate online advertisingand capture most of that revenue.
That's what our data isenabling. They're selling Google
app services around insightsinto what people are doing online.
Butdata is what an economistwould naturally see as a public good.
It's intangible, it can befreely shared without being depleted.
And if there were not thekinds of legal and technical protections
around data use, then it mightnot have a price at all. Once it's
out there, it's out there. Itmight be free, so we'd still have
privacy concerns and so on.Butthinking about it just in pure
economic terms, it's not clearwhere that value comes from. To what
extent is that market power? Ithink it's not wholly that because
the services that they provideare useful. Ifyou think advertising
products to people who areinterested in them is a useful thing,
then that's also got economicvalue. We consumers get our share
of the value through theservices that we really do rate highly.
So,there's a kind ofcomplicated assessment there about
if your aim is to understandprogress and what's going to be good
for people, over time, in theeconomy, that's just quite complicated
to understand.
Let's go back to when we weretalking about the dematerialization.
One thing I wanted to ask youabout there was, we talked about
Rolls-Royce, and there kind ofmaking money from, you know, not

(23:47):
just from selling engines butfrom servicing them over time. There's
another issue that you broughtup which seems to me going that's
going to be extraordinarilyimportant, which is cloud computing,
which is literally acompletely or almost completely dematerialized

(24:11):
product. Imean,what are themeasurement issues that cloud computing
presents in terms of, youknow, not just the aggregate GDP,
but also kind of the trade balance.
So, the services being soldare software and things like storage.

(25:05):
So, a whole range of productsoffered by the cloud providers. There

(26:01):
are not official statistics onthe cloud sector. Thedata centers
are privately owned, some ofthem by the big tech companies, some
of them by others. So, I wouldguess that national security agencies
know where all those datacenters are and where the cables
are going into and out ofthem. Butthey're not economic statistics.
So, we don't know that. Wedon't know about volumes. A lot of
countries have datalocalization laws that require all
data generated in theirboundaries to be stored within those
boundaries. And I think thetech companies largely abide by that.
But it's not possible tomonitor, really, what's going on

(26:37):
there. It'sreally complicatedto understand what prices are being
charged, partly because theproduct range changes all the time
and it's very complex, andpartly because that data is not given
to statistical agenciesconstructing price indices. So, on
the one hand, we know that alot of governments are switching
to using cloud services.Theydon't buy their own servers
and hire their own peopleanymore. That in itself, at least
for a while, is going toreduce GDP growth, because buying
your servers gets counted asinvestment, which adds to GDP. Paying
a cloud service gets countedas an intermediate expenditure, which
deducts from GDP. So, there'sa kind of funny statistical artifact
going on there. Butalso, Ithink, particularly at times like
these, where there's growingconcern about questions of geopolitics
and market power, just a reallack of regular statistics on what's
going on with the sector.There are some competition inquiries,
there's one going on in the UKat the moment into the cloud sector.
And that's a means for theauthorities to get some insight into
the kinds of statistics I'vejust been talking about. But that
won't deliver us the kind ofregular price index for cloud services.
So,I've constructed that witha colleague. We did it by web scraping
Amazon Web Services, which isthe market leader. And that's being
used by central banks in theUK and European Central Bank. But
that's not the same as thekind of experts at the official statistics
agencies being able to do that.
Let's talk a little bit aboutanother area which you call disintermediation.
And this seems, again, thisreally resonated with me. So, I guess

(26:58):
what you're saying is thateconomic statistics, they kind of
envisioned an economy wherethere's this boundary between what
you do for work and theactivities you do for yourself. Andyour
point has been that there'sthis kind of transformation in these
activities, so that thisboundary is pretty blurred now or
maybe even nonexistent in somecases. And that means that the measurement
of the economy isn't accurate.So, could you kind of explain that
idea to people in some detail?

(27:22):
Sure. So, we refer to this asthe production boundary, and things

(27:52):
that are transacted for moneygo on the GDP side of the boundary,

(28:24):
and things that don't involvemonetary transactions go on the household

(28:51):
side of that boundary. Andthere has always been contention

(29:15):
about not measuring some ofthe things on the household side

(29:42):
of the boundary.Clearly,there are a lot of care
and other things done in thehome that are economically valuable
and societally valuable. Andsince the 1930s, feminist scholars
have been pointing out thatthese are invisible to policymakers
because they're not counted inany way. So,there's that background
level of concern about whatgoes on what side of the boundary.
Even on the monetary side ofthe boundary, we put some things
that don't really have marketprices. Andthe big example is we
put what's called animputation for the rents that you
would pay yourself if youdidn't own your own home. So, if
you're an owner occupier,you're not paying rent. The statisticians
estimate how much rent youwould pay, and they put that on GDP.
Theidea being that shiftsbetween owner occupation and rental
shouldn't change GDP. But thatkind of artifact of shifts across
the boundary applies to otherthings as well. Thefact that a lot
of women stopped doing most oftheir work in the home and started
doing much more of their workin the paid economy through the 1970s
and ‘80s will have boostedmeasured GDP growth. But some of
that was just a transfer ofactivities, because if you're going
out to work rather thancooking at home, you'll be paying
for a ready meal or a takeoutmeal instead, for example. Nowthe
digital technologies arecausing a lot of shifts across these
boundaries. And you can thinkof examples like doing more of your
banking online or your travelagency online. There are still obviously
activities in the market, butsome of that you're now doing for
yourself. Wetalked earlierabout the example of rather than
buying newspapers, you'll belooking at some of the free services
online. So, a lot of theseshifts have started to happen and
it's probably, to some extent,reducing the productivity as measured
in the monetary economy - inthe monetary side of the economy.
How does that work?
Because the activities arestill taking place, but they are

(30:19):
occurring in the part of theeconomy that doesn't have prices

(30:49):
attached. Back in 1965, GaryBecker, Nobel prize winning economist,

(31:16):
wrote a paper about this andhe pointed out that his example is

(31:50):
a safety razor. Andso, peoplestopped going to the barber to get
their shave, they bought asafety razor and shaved at home.
Eventually barbers figured outhow to offer other services instead.
Butat least for thattransition period, they would have
had less revenues, lessgrowth, less productivity, less measured
productivity. And hehypothesized this was quite a big
part of the productivitypatterns. Anotherexample that I've
been thinking about is the waythat we shop for our groceries has
changed. And back in the ‘50sor’ 60s you'd go to the greengrocer
who would have all the goodsbehind the counter, and you'd say,
could I have half a pound ofbutter please? And it'd be measured
out for you and handed over.So, not much capital, a lot of labor
involved in that. Thenwe'vegot supermarkets, they used to have
people who would bag yourgroceries for you, so you picked
things off the shelvesyourself, but all the rest of the
work was done by paid labor.Now we have these automatic teller
machines which have a lot ofcapital and software, but rather
than a paid employee scanningand bagging the groceries, you're
doing all that yourself. So,this is a shift of labor involved
in shopping for groceries.Goingsteadily from paid labor and
no capital, to some paid laborand more capital, to a lot of capital
and no paid labor, which youare doing for yourself. What's this
going to do to the measuredproductivity of the retail sector
over time? Ithinkat themoment it'll be increasing the productivity
of the supermarkets becausethey've got roughly the same kind
of sales, but they're payingfewer wages, but the same work is
being done. So how should webe thinking about that?
Yeah, no, I'm glad you broughtthat up because I think about that
a lot and there's kind of abroader issue which is, I don't know,

(32:14):
I always have to check myself,like how much of what I'm saying
is insightful and how much isjust being a grumpy old man. But
there's this, you know, I feellike so much of the customer service

(32:36):
activities of companies areessentially now outsourced to the
customer. Andas you say,that's it's sort of like that job
is still being done, but nowit's being done by you and I trying
to figure out how to fixsomething if it goes broke, as opposed
to being able to call up thecustomer service and say, hey, you
know, can you help me with this?
Well, it's not just you beinggrumpy. It's been labeled a time
tax and engaging with a lot ofpublic services and private services.

(33:03):
You can spend a lot of timetalking to a bot or trying to get
through to a human being onthe telephone to sort out your problem.

(33:27):
And there are automatic menus.So,you end up having to spend a
lot of your time on thisactivity and not even necessarily
to get a good outcome, but asyou say, to fix a problem that's
gone wrong because of all theautomation that goes on. But this
will get worse, I think,because companies are using bots
more and more for customerservice because of the advances in
AI. Andso how are we going tothink about it or measure it when
my agent, my agentic bot istalking to the bank's agentic bot?
Is that real economic activityat all? What value is that creating?
Yeah, no, that's right. And Imean, I know this is quite an important
concept for you, right? Andwhat you would like to see happen
going forward, I mean, you'dlike to see time (I believe, if I'm

(33:53):
right) as a kind of a baseunit of measurement. Because that's
sort of a resource that isfinite and we all only have 24 hours
in the day. Howwould you liketo see time incorporated in kind
of a reformed set or anexpanded set of economic statistics?
So, I've started to thinkabout this both from the perspective

(34:48):
of us as consumers and fromthe perspective of production and

(35:26):
productivity. So, as we werejust saying, we spend our 24 hours,

(36:07):
you can't carry them over,you've got to allocate them to different

(36:49):
activities. Youneed to spendtime to work, but you also need to
spend time to consume, whichisn't usually thought about. And
you need to spend time as wellas money to engage in leisure activities.
So maybe we should think abouthow much well being do people get
out of their activities overthat 24 hours. Alotof countries
have time use statistics forconsumer and household activities,
mainly used to get an insightinto things like what's the value
of caring activities in thehome. But you could think about doing
it more regularly and in amore detailed way to understand these
allocations and to think abouthow to make, you know, each of them
more enjoyable or at leastless unpleasant for people. And then
that would bring the time taxright into the frame. And we might
start thinking about, do wereally want companies to be setting
up their business models inthis way? Fromthe production perspective,
you can look at this in twoways. A lot of productivity improvement
over time has been about doingthings faster. So, we focus on the
shiny gadgets, but actuallyit's the fact that the steamship
can cross the ocean muchfaster than the sailing ship, or
that manufacturing a car sincethe introduction of Just-In-Time
systems, has become muchfaster than it used to be in the
‘50s and ‘60s assembly line.So, those process innovations are
how technology drivesproductivity forward. And they're
often about time saving.Butthere are also, in a service
economy, lots of activitieswhere you want a lot of time spent,
but it's got to be highquality time. If you're thinking
about, say, a medical settinghaving a routine test, you want it
to be done quickly, you don'twant to have to wait around for it,
and you want your results asfast as possible. So, productivity
gains there are speeding upall of those procedures. Butif you're
really ill and you're in anintensive care unit, you want to
have a lot of attention fromhighly skilled medical staff. And
so, you can think about twokinds of productivity. There's a
sort of time and speedproductivity, and there's a quality
productivity. Andif wethought about how to divide up the
economy on that basis, we geta different kind of perspective from
conventional ways ofmanufacturing versus services or
industrial sectors.
What would it require toproduce those types of statistics?

(37:10):
Are we talking about justexpanded types of surveys, or are
there other ways that you cancapture? Andthe reason I ask that
is that, you know, we did haveJulia Lane on the show in, in January,
and I know that you quotedsome of her research in your book…
She does a, you know, verygood job of explaining the challenges

(37:33):
with the survey method ofcollecting data. Andthen she was
talking more about collectingdata on, you know, other types of
economic activities. But itdoes raise a sort of degree of skepticism
in terms of, you know, theaccuracy of surveys, at least as
they're done now.
Her research is fantastic, andshe's absolutely right about the
way that surveys arecollapsing at the moment. So, we

(37:58):
have to think about newmethods to do this. I think the technologies
may offer some ways to thinkabout it. You know, they might be

(38:20):
considered to be surveillancetechnologies, but you can often map
what people are doing throughtheir outlook diaries online, for
example. Wedo have a pilottime use survey for public sector
workers where they actuallyrecorded how much time they thought
they spent in useless meetingsand how much of their activities
were productive. So, it's anopen question. I don't know the answer
and I'm thinking at the momentabout how to monitor that in the
emerging AI economy. Ifit'sthe case that generative AI is going
to transform the economy andproductivity, what should we be looking
for? How should we lookingabout those kinds of processes at
work to change?
You talked about, I think inthe introduction of the book, just

(38:50):
how kind of new technologiesand new innovations which clearly
would improve people'sstandard of living but would have
kind of a counterintuitiveimpact on GDP. The example you gave

(39:15):
was new treatments for maculardegeneration. Switching from a drug
that costs $2,000 a month toone that costs $55 a month. Thatreduces
spending, so potentiallyreduces GDP, but at the same time
clearly improves wellbeing.Can you just explain that idea in
detail how these kind oftechnological innovations can, I
guess, mess up or interferewith our ability to kind of measure
GDP accurately?

(39:35):
It's about ideas, really. So,there are other medical examples.
You might think about therealization that aspirin can be used

(40:04):
to avert cardiovascularproblems rather than just being a
painkiller. So, there areplenty of medical examples of new

(40:30):
ideas. Here are reallyeveryday examples as well. Idon'tknow
if you ever get sucked intothe Instagram vortex of videos, but
for some reason I get servedquite a lot of videos about home
hacks. And it's things likeyou can turn your milk carton into
a trowel to use in the garden.Those are great ideas, but they're
not adding anything to theeconomy. If anything, it's reducing
a measure of GDP because I'mnot buying the trowel to use in the
garden. So,it's really aboutwhat has driven human progress for
at least a quarter of amillennium. And it's using ideas.
So, productivity is all aboutthat you've got a certain amount
of resources. How effectivelycan you use them to deliver things
that people value and make itfeel like life is getting better.
That's what it's fundamentally about.
That's a good, I think, segueto what your new framework that you'd

(40:54):
like to see. Because it reallyis, I guess, the base of it is getting
a much broader understandingof what that resource base of the
economy is. And so, you havethis concept that you call ‘six capitals’,

(41:18):
which I know you've done some…that's kind of your research framework.
Canyou explain that to us,the six capitals? How you've chosen
the categories and why they'reimportant? And, I guess, maybe which
of those capitals we'recurrently measuring and which ones
were not.
So, a simple way to thinkabout, well, at least the way I think

(42:09):
about what we might measure.We've got the system of national

(42:42):
accounts, GDP. You can thinkof that as being like a company profit

(43:20):
and loss account, and thatinvolves various manipulations. We

(43:47):
talked about time. You canthink of that as a kind of cash flow

(44:28):
that monitors things day today. And then the missing bit is

(45:04):
a balance sheet. Andso, theidea of the six capitals, which in
literature is calledcomprehensive wealth, sometimes inclusive
wealth, is that you're lookingat all of the assets that an economy
needs to continue producingGDP growth in future. So, in that
very broad sense, it's aboutcan you continue to grow? Andso,
think about what we need. Andpeople would think about financial
capital, I guess, butcertainly productive capital - the
buildings and machinery thatare used to deliver goods and services.
Also, the infrastructure ofthe economy, which isn't particularly
well measured anywhere, butwithout it, there's no economy. Naturalcapital,
we get a lot of services, frompollination on farms, to cooling
by trees in cities, to cleanair that stops people getting asthma
because the pollution is sobad. Human capital, which is the
term for people. And whatskills do they have, but also what
health do they have thataffects their ability to work productively
and be satisfied with theirlives? Andthen sort of more the
fuzzier bits. It's the ideas,it's the knowledge capital. What's
that knowledge, that base thatwe have that allows us to carry on
doing things? That does seemto be able to erode. And I think
we're seeing some examples ofthat at the moment. And then how
do we organize that allcollectively – so, social or institutional
capital? DaronAcemoglu andhis colleagues won the economics
Nobel Prize for focusing onthe importance of institutions for
economic growth. So, we knowthat they matter. Can you think about
how to measure whatconstitutes a good set of institutions
for transactions to take placeand contracts to be observed and
people to trust each other? Sothat's the sort of balance sheet.
Waterfront. What do wemeasure? Well,the bit we measured
best, I suppose, is physicalcapital, but even then it's pretty
complicated because you'rethinking about combining the equipment
a hairdresser has in theirsalon with the machine tools in a
standard manufacturing plant,with the really advanced stuff needed
to make silicon chips. Butthat's probably the one that we measure
best. Alotof progress hasbeen made in measuring the natural
resources that we use. And theUS started collecting those statistics
a couple of years ago. We'vedone it in the UK for 10 years or
so, and there's a set ofstandards for doing so across countries.
Humancapital people typicallymeasure as either what skills have
people gained or what wagesthey earn that reflect the assumptions
about their skills. It doesn'tusually measure health. I think,
because of aging societies,it's becoming more important to understand
the health component of humancapital. And then the others, we
don't measure all that well atall. Andinfrastructure, as I said,
seems pretty basic, but wedon't measure it in terms that understand
how well maintained it's been.And of course, bridges can collapse
overnight, as they sometimes do.
I want to ask a question aboutnatural capital. As I was thinking

(45:35):
about your expanded framework,that one, I guess, seemed the easiest
for me to think about becauseI think it's fairly straightforward

(46:11):
to think about - a naturalasset producing services that's input
to economic growth. So,iswhat you're imagining a system whereby
you would have a kind of abalance sheet of natural capital,
I don't know, today, and thenyou would kind of adjust that as
the natural capital gets usedor depleted? I mean, I guess I was
thinking about something likea mineral deposit, copper or lithium,
something like that, that wewould need for renewable energy.
Areyou imagining that ifthat's extracted, that depletes natural
capital? And then also there’sgot to be an interaction with technology,
right? Because maybe the valueof that deposit is a function of
our ability to extract it, butthe extracting, you know, technique
might then have environmentalconsequences, presumably, which also
affects the balance. So, am Iconceptualizing that correctly?

(46:35):
Yes, well, you're pointing tosome of the challenges and trying

(47:33):
to think about how would youmeasure the whole of the economy's

(48:37):
natural capital? Andif youcome at this from a national accounting
background, you think, well,I'd need to know the quantity and
I need to know what price toapply to that. And that gives me
my balance sheet figure. So, Iwill estimate the amount of reserves
of whatever, lithium, orwhatever it is. And what the national
accountants do is they thenapply a market price or something
close to a market price tothat. Ifyou're thinking about it
from the point of view ofbroader societal needs, you might
want to think about the kindsof technological substitutions that
can occur. You're going tothink about it more as an asset in
a portfolio of thingsavailable to the economy. So, there
are two different ways ofthinking about it. The literature
is on the simple, we'll justlook at the quantities and find some
kind of price. Tounderstandwhat does this imply for future capabilities
to produce and grow? You needto think about it in a different
way. And that's not happenedyet. That's work in progress, I would
say. Idothink it's becomingimportant because in the geopolitical
environment now, or given theshocks of the past 10, 20 years in
the economy, and the impact ofbottlenecks in supply chains (thinking
more about as a nation), whatare your capabilities? What are the
assets that you have tocontinue to produce? This becomes
much more important than ithas been. So, I think that's a really
interesting agenda. There'snot a lot of research around thinking
about it like that, butfundamentally important. Thefarming
sector in the Midwest dependscritically on the ecosystems that
deliver the bugs that keep thesoil healthy. And nobody's looking,
well, environmentalists arelooking at it. Nobody in economics
or national statistics islooking at that. But if that reached
a tipping point, it woulddestroy the productivity of the sector
really quickly.

(49:21):
With human capital, you quote,I guess the World Bank has said that
something like 60% of totalwealth is human capital, I think
if I got that right. And thenyou say, well, that's not in the
national statistics at all.Andyou also talk about, you know,
how spending on education isconsidered consumption as opposed
to investment. So, it's notbuilding capital. Can you talk a
little bit about, you know, ifwe wanted to value human capital,
how we would go about doing that?
The typical way of doing it isto look at the different levels of

(49:45):
educational skills andattainment in groups of the workforce
and to (using a somewhatcomplicated formula) basically apply

(50:06):
the wage that people get forthose different types of skilled
activities and add it up thatway. That's obviously quite a narrow

(50:32):
way of thinking about humancapital. As I mentioned, it doesn't
include health, it doesn'tinclude particularly people's well

(50:54):
being and how motivated theyfeel. So,you could think about it
in a much more expansive way.So, what's there is quite a narrow
measure, but that 60% is also60% of quite an incomplete measure
of this broad sense ofnational wealth. So, I think that
we're slightly at sea in thefog with those kinds of figures.
Butif ideas (and this is oneof the questions about AI), if you
think ideas drive productivityand progress over the long term,
which has been the case sincethe Industrial revolution. Ideas
are attached to people, andwell educated people tend to have
the kind of ideas that producenew technologies that continue driving
economic growth. Nowmaybe AIfor science will change all that
and the machines will startdoing all that innovation and so
the people can stop thinking,and the human capital will get substituted
by agentic AI capital. I'm notsure I believe that.
So, if you do say, and I thinkwe get the sense (we're just listening
to it) that it's a challengingagenda that you're putting forward.
I mean, what is the currentstate of play on some of these things?

(51:19):
And if people listening areinterested in kind of following up,
what are some of the placeswhere they can kind of take a look
at, I don't know, say, wherewe are with natural resource accounting
or human capital accounting? Imean, is there public data that you
think at least useful in some sense?

(51:40):
Yes. So, all theindustrialized economies have their
own national statisticsagencies that put out lots of data

(52:02):
and interpretation. Theinternational agencies like the World
Bank or the OECD are goodplaces to go for information. We're

(52:27):
in an interesting time becausethe system of national accounts gets
revised every 10 years or so.And we've just had a revision earlier

(52:56):
this year approved by theUnited Nations. So, it's become the
official set of standards.Butalmost immediately, the Secretary
General of the UN announced anew commission to look at how to
do better economic statistics.Because there is, I think, a deep
sense among many populationsthat the GDP growth figures don't
reflect their experience. GDPgrowth has been going up. A lot of
people feel their lives havenot been getting better. So, the
popular demand, if you like,for a different approach to measurement
is definitely there. Butit'sa bit like a technical standard.
You know, if you've got twopin plugs and you want to switch
people away from two pin tosomething else, you've got to know
what the something else is.And it's got to happen at the same
time. So, you need aconsensus. AndI don't think we're
anywhere near having aconsensus among economists and statisticians
on what is the right way to dothis? The things I talk about in
my book have pretty solideconomic theory underpinning them.
So, my hope is that theeconomists will go, that might be
a good way to go. Let'sexplore that and let's think about
developing standards aroundthat. ButI might be wrong and somebody
else might come up withsomething better.

(53:16):
When you personally try togauge where the economy is strong,
weak, growing or not, whatstatistics and data sources do you
still value given the factthat as you said that so much of
the economy is invisible?
Well, you have to use a mix.The official statistics are suffering,

(53:50):
as you mentioned, becausesurvey response rates are going down.
There's a lot of turmoil inthe US statistical system at the

(54:17):
moment because of cuts andproposals to merge the agencies.
But there have been budgetcuts everywhere and we're going to

(54:44):
have to think about new waysof doing it. So,my team used a lot
of techniques like webscraping, or trying to think about
accessing mobile phone data,or online trends data, and so on.
So, a lot of experimentationis going on with different ways of
doing things. And I thinkwe're just in quite a confused phase.
Ihavea facsimile edition ofthe 1885 Handbook of Statistics for
the United Kingdom. With 1885think dark satanic mills, and Dickens
novels, and all of thatindustrial misery. And out of 120
pages there are maybe 10 pageson all of the aspects of the Industrial
revolution, the railways, andthe cotton mills, and the coal mines,
and all of the rest isagriculture. It took a long time,
it took about 50 years to getfrom that to thinking about what
became modern GDP. Ithinkwe're in one of those periods
driven by big technologicaltransformations in our economy and
really big social andpolitical changes as well. So, it'll
keep me busy for the rest ofmy career.
Well, that's good. You'll beproductive. Butit does mean that,
you know, for a lot of people,especially those in the markets,

(55:07):
but just in general that we'rekind of operating in a fog and the
fog isn't going to clearanytime soon. But it also means that
we should take a lot ofinterest in this sort of research

(55:33):
because it will hopefully,eventually, give us a much better
system for understanding wherewe're at. So,I think that's a good
place to leave it today.Diane, I just wanted to say thanks
for writing the book and inparticular thanks for taking your
time. I hope it wasn't tootaxing. I hope it wasn't a big time
tax to share your ideas withus. We really appreciate it.
On the contrary, it was a realpleasure. Thanks so much.
Okay. Well, the book is calledThe Measure of Progress. Pleasemake
sure to go and get a copy andto follow Diane's work, because I

(55:53):
think you can tell not onlyare these very important ideas, but
they're also not beingdiscussed enough on mainstream media.
So, for all of us here at TopTraders Unplugged, thanks for listening
and we'll see you next time.
Thanks for listening to TopTraders Unplugged. If you feel you
learned something of valuefrom today's episode, the best way

(56:15):
to stay updated is to go onover to iTunes and subscribe to the
show so that you'll be sure toget all the new episodes as they're
released. We have some amazingguests lined up for you, and to ensure
our show continues to grow,please leave us an honest rating
and review in iTunes. It onlytakes a minute, and it's the best
way to show us you love thepodcast. We'll see you next time

(56:36):
on Top Traders Unplugged.