it's great to see so many people to see just some of the people who have who have come because I understand there are some others in other room so hello to everybody I'm franisco gazelli I'm a professor here at the LC I have the pleasure of introducing tonight's speaker Diane Co she has she's a graduate from Oxford University she's a PhD from Harvard she she has had a long and distinguished career in public service she has she's a founder of enlightment economics she's now a professor at the University of Manchester and she's here to tell us about GDP um she says she feels affection for um GDP that's very mild for my standards I Feel Love for GDP um so thank you well good even everybody who would have thought that this number of people would turn out for a talk about GDP on a dismal Monday evening in November so thank you very much um but there is a surprising amount of interest in GDP now I've given talks to uh audiences of 300 old age pensioners who are interested in GDP and what's a little strange about it is that it's part of the furniture of the conversation we have about what's going on in the news um people always talking about GDP up 0.2% this quarter and so on and economists of course use it a lot they feel great affection for this data series that they can now download from the internet for lots of countries over long periods of time and run regressions but uh when I was young which was about a million years ago economists were taught national income accounting in their undergraduate courses and I think that's largely Fallen by the wayside now and people are using GDP fig and talking about GDP figures without anybody having paid very much attention to them until just recently with all this interest so why now and I think one reason is the environmental concern environmentalists who feel that there's just a bit too much GDP growth because they're worried about the impact on environment and climate change in particular but another reason is that because of the financial crisis there are other people who think actually there's not been enough GDP growth the recovery from the crisis has been slower and more feeble than recoveries from past um uh downturns it's been a longer more drawn out more serious recession there are also people who think there's not enough GDP growth because we're entering an era of secular stagnation and it's tied up with the debate about uh what's happened to the trend in productivity growth which is typic measured using the GDP figures and then as well as that there are all the digital people Silicon Valley people who say there is a lot of GDP growth we can tell you because we're delivering it and it's just not being recorded in the figures the figures are wrong so I think all these things are converging and it's one of the reasons uh Charles Bean has Sir Charles Bean has been asked to do his review of the economic statistics and it's important because it's how we assess progress It's the headline measure we use to know as voters is the economy getting better or worse are are politicians delivering for us and the debate about the statistics is a proxy for a debate about the character of the economy and how well it's doing but the statistics are not objective measures of reality it's not like trying to measure the speed of light it's not like trying to assess the velocity of a moving object in any way they are definitions and interpretations of something that's happening in the world but also that in turn shaped reality because we respond to them it changes our behavior when we hear what's going on so today I want to focus on the digital changes and how that's affecting um the structure of the economy and how that might be playing out in GDP figures or in something that would be better than the GDP figures that we have and I've got three points I want to make really one is that the statistics are significantly lagging structural change in the economy the second is that there's a growing gap between GDP and um a better concept of economic welfare and the third is that the character of the digital Goods in particular means that that wedge between GDP and economic welfare is growing and possibly growing quite quickly now statistics are for States they're State istics um and they reflect what governments want to know and think they need to know to run the country and the structure of them always lags the reality and it can be quite a big gap this is a page that you won't be able to read from um an 1886 publication I've got a faim of it that the office of national statistics published in 1986 and it's got 200 Pages most of which are like this one acreage of crops and number of livestock it's headed so this book tells you about imports and exports of Commodities from the colonies it tells you about the sales of different kinds of grain in Market towns around England and Wales it gives you all these figures about uh grains and carrots and potatoes and the number of livestock in the country out of the 200 Pages there are 12 on mines factories canals Railways and then a couple of pages on the public finances too this is at the height of the Industrial Revolution it's a couple of generations since William Blake was writing about the dark satanic Mills and yet the statistics the gap between what they were recording and what was happening in the econom had become enormous because it was a period of very rapid change and everybody knew that and at the time uh Parliament commissioned a lot of what we call Blue Book studies looking at um conditions in the factories number of children working and so on and I think we have a similarly big gap now the current statistic the current framework that we use for our statistics dates from the depression and the second world war and these two people were particularly significant in that debate the guy on the right I'm sure you know is KES the guy on the left is Simon kusut and in the 1930s he and somebody in the UK called Colin Clark were the first people to put together an aggregate measure of the economy they needed to do it because it was the Great Depression and governments on both sides of the Atlantic and indeed in other countries wanted to know what was the extent of the economic management problem they faced in responding to the depression they had lots of figures they had the number of bank failures they had the number of Steel bars that were being produced by the factories uh They Knew by that time how long the railway system was the railway network was but they didn't have an aggregate measure they didn't know uh what extent of action they might have to take to bring down unemployment or um tackle the the bank failures and um there was quite a debate about um measuring that aggregate it was the first time anybody had thought they needed it because of the extension of the democratic franchise there had been vicious trade Cycles in the 19th century and nobody thought the government should do anything about it so it was the first time that people had a concept of the economy that the government ought to ought to try and um do something about and KET in particular had a very strong view about what this aggregate ought to be measuring and he wanted it to measure economic wealth so he said for example you would exclude advertising because people didn't like it you'd excl you'd exclude commuting it was a regrettable necessity that people had to do to get to their jobs you'd exclude other regrettes like policing or paying for defense they had to be done but nobody got any economic welfare out of it his view though was overtaken by the demands of the second world war when the imperative very much became having an aggregate measure of the economy the that could be used and this is where kanes came in he wrote a pamphlet called how to pay for the war he wanted a measure that could be used to give the government on both sides of the Atlantic an indication of what was available to put into the war effort and what consumption sacrifice domestic consumers were going to have to make to sustain the war effort he wanted an activity measure and in particular the administrators in the United States wanted a measure that did not exclude arms spending because they didn't want anything that made it look like the war effort was harming the economy so this debate was quite vigorous and actually it lasted into the 1950s but because of the wartime imperatives kan's won that debate and the aggregate that we got um that was the Forerunner of today's GDP was much more an activity aggregate than an economic welfare Aggregate and some other key judgments were made around that time about what's called the production boundary what do you put in and what do you keep out of GDP um one thing that was uh going that went in was government the aggregate activity measure measured activity in the Market at market prices and of course government doesn't have market prices there was also quite a debate about which parts of government spending you could validly include because they were services to Consumers they were final final goods and services and which parts you want to try to net out because that was just spending that had to be done these regrettes to sustain activity um it proved too hard to do that I mean if you have a road that you spent that the government spends money on how can you divide that into uh what part of it is uh people moving along the road for enjoyment and what part of it is people moving along the road for commuting is Impractical so all of government went in but all of housework home production was left out and if you look back at the um old literature it's quite interesting to see that there's quite a consensus that that housework home production ought to have been included but they decided it was just too difficult to do that you couldn't Sur survey all the housewives it was always Housewives of course and there weren't enough market prices for those services to be able to Value them even if you knew how much time people were spending on them so housework got left out and over time of course we just got used to these definitions and people um put behind them the debate that there had been in the 1930s and 40s about the kind of construct needed initially all of the interest was in the level of GDP and the level of its components because that was What mattered what was left over um after the war effort after wartime spending when the war was over the interest was about the level of employment and the stability of employment GDP grew up alongside Keynesian demand management and that well-known equation that I'm sure most people know here C plus I plus G Plus xus m and it was the level that mattered but before too long attention switched from the level of GDP to its growth rate does anybody have happen to know the level of GDP in the UK anybody want to call out well done full marks um but actually it's not a very interesting number is it it's not one that we pay attention to it's always the change that we pay attention to but that switch from levels to rates of change actually had a very political cause it was was the Cold War and uh chrush made a famous speech the Russian leader in which he said growth is the battering ram with which we will smash the capitalist system this was the time of Sputnik the um miss the Cuban Missile Crisis and JFK and his uh chair of his Council of economic advisors Walter heler were very clear then that they had to focus on American economic growth how fast it was increasing and that growth Target got written in at their behest into the founding Charter of the organization for economic cooperation and development and they had a target for um 50% growth in GDP over the first decade of its existence so there was a growth turn at that time but it wasn't very long before people then um started to have some concerns about focusing too much on economic growth and it was the environmental critique this very famous report the limits to growth was published in 1972 so only a decade after after the switch to worrying about growth and not levels of GDP and this critique was in the specific context of the environmental externalities of economic activity and um the fact that those externalities are omitted from measures of GDP but are an important part of economic welfare now very often economists will say um externalities are important the environment is obviously important but GDP growth isn't meant to capture economic welfare it's just a measure of economic activity at market prices but that's not true apart from anything else we use it as a welfare measure all the time constantly and anyway why else would you ever bother to look at real GDP over periods of time or compared AC cross countries because that can only be about economic welfare or progress in some sense and I think very often people are quite modeled about the distinction and you'll get statements like the ones from um havarian who is Chief Economist at Google who says uh GDP is growing faster than the statistics show when what perhaps he does mean is that GDP excludes some things in his case all the digital activity um in the environmentalist case it would be that it includes some things so it's uh growing at a different meas GDP is growing at a different rate from economic welfare and we need to start to get that distinction more clearly into our minds and the environmental externalities are certainly an important part of the gap between what we measure in GDP and a wider conception of economic welfare and of course they're negative and there are lots of indices that try to deduct environmental externalities from GDP and they all show um that the adjusted index whatever it is they have different names stops growing in in the mid 1970s uh largely because of the impact of the oil price rise then and they used that to Value the Gap but it's not the only one there are other externalities other uh parts of the GDP and economic welfare wedge that might go the other way and that's what I want to talk about now I want to talk about the digital some of the problems about digital are pretty straightforward in theory even if they're hard in practice GDP put together from hundreds literally hundreds of different surveys collected from different kinds of people and um over different time periods with different seasonality patterns and they all have to be combined together if you get a change in the population of people filling out the surveys or who ought to fill out the surveys for a more accurate measure of GDP then you have a bit of a problem through the 50s and 60s into the 70s uh even into the 80s actually there's a relatively small number of large firms filling out surveys about things like investment levels or the prices being charged for products and of course that's supplemented by other indicators actually those surveys might not have been as reliable as as expected I was doing a talk in Manchester for some uh ordinary citizens who'd come along to hear um how the national account statistics work what it all means and an elderly gentleman put up his hand and said I used to work for ICI and it was my job to fill out the survey that came from the office of national statistics but I was much too busy so I looked at the year before and I just added a bit to the numbers when I put them in so there's always measurement error but if the population of um the firms in the economy is changing and in particular if you've got more small firms more self-employed people then it's harder to capture what they're doing uh through all the surveys so you might miss them and there would be an increasing level of uncertainty about the figures there's also the really interesting issue of classification now there's something called the standard occupational classification that lists to a great great level of detail all the jobs that you can do and when they uh collect the statistics the labor force survey people filling out the surveys assign themselves to a category this table this uh list of classifications uh has been updated but was initially drawn up in an economy that was uh very much focused on manufacturing and much less on Services than it is today so if you're a painter there are literally 50 categories that you can assign yourself to so let's have a look you can be a painter of cars you can be a painter of boats you can be a painter and decorator you can be a painter of watercolors or Miniatures or artificial flowers you know literally here's the list so there are all uh all sorts of categories of painters there's a category called maker up about 20 uh slots for maker makers up I tried searching online and this job doesn't seem to exist anymore if anybody knows what it is I'd be interested to find out afterwards so presumably nobody now ticks those boxes but I asked myself what if you're a social media consultant of some kind where would you put yourself so I had to look through the whole list and I found maybe what is that a dozen categories that you could choose and there are four of the high level occupational categories here um but for all kinds of jobs now being a Web Master uh being a programmer being a video games designer or or coder there just aren't obvious categories to put yourself in and it's not clear at all where those people if they're uh found by the surveys are assigning themselves to we just don't know and it's not that easy to fix partly because actually a lot of those job titles aren't standardized yet if you're doing big data you can be called a big data scientist or you can be called a chief data evangelist and we don't want a category for every title until until the job category settle down so there's bound to be a bit of a delay in getting that fixed the other problem is that um these surveys these are lists are done by International agreement through the United Nations and it's a very slow process it's been done quite recently there's not much appetite in that Community to go through at all again and not much appetite among a lot of countries to start worrying about where do you put the chief data evangelist in these categories so that's not going to get fixed quickly internationally I think it's something that we need to start looking at nationally very quickly because there are people doing all kinds of jobs and we've got no idea how many or where they're putting themselves at the moment and that seems quite important a bit more definitionally challenging is the move away from bricks and mortar and there's a lot of retailing now substituting for um High Street retail sales poor Blockbuster one of the many to go under Woolworth uh and others that you'll know about now there's no clear effect of this on consumer spending it's a change in the channel through which people do their shopping and we don't really know whether there's a general price fall as a result of moving to online retailing and if there is are the volumes of things that people buy staying the same so nominal GDP is coming down or are the volumes going up because the prices have decreased and are the prices actually being measured properly anyway there has been a lot of work in the ons to look at online pricing um but a lot of their work on the price indices still relies on um going to uh supermarkets to check prices and getting in um till data from High Street stores so there's a lot of uncertainty about that but also um this is going to contribute to a decline in business investment in commercial property because people aren't building High Street stores and in fact there are lots of other businesses that are moving away from physical retail channels so the banks have closing branches turning them into wine bars um books and CDs are increasingly uh dematerialized you don't get travel agencies anymore people do that online so the these these this demetalization of uh retailing and similar activities is probably contributing to a slower measured GDP growth but consumers are no worse off maybe better off and there's improved productivity of bricks and M because uh uh digital retailing is served by just a few Warehouse buildings rather than all of these uh High Street stores so that's another one in fact the dematerialization of the economy is quite significant and it's something I first noticed in around mid 1990s this is a chart that shows real GDP and what's called the total material requirement of the economy which is the um amount of physical material it's the weight of the UK economy the weight of production in the UK economy it doesn't take account of material and resources embedded in imports and exports and we have a trade deficit so that if you took that into account it would close the gap this is about the structure of domestic production in the economy but that's quite a significant move away from the material this is telling us that all of the real GDP growth domestically generated domestically over that period 25 years now has been intangible in some way and if you think about it that's not surprising it's the design the research and development the software the consultancy the lawyers all the services service sector now 80% of the economy now the statisticians know that measuring intangibles is something they have a problem with and they've been taking steps to address that so last year with revisions to the methodology they introduced um investment in software and investment in R&D and patents and copyright into the GDP measure but if you think about it what is a unit of output in a service what is a unit of output of a lawyer um or a a nurse is a nurse more productive if she gets through twice as many patients a day or if she gets through half as many patients but gives them better care and they get better faster and the answer is it depends if she's doing blood tests or sight tests the faster the more she gets through the better and if she's caring for very sick people then the fewer she has to deal with the better as long as that care delivers better outcomes and it's really hard to separate conceptually never mind measuring quantity price and quality in lots of services so what do businesses have in mind when they are uh responding to the on surveys about the price for their services a lawyer will give a billable hour measure but what does that actually mean what is the output of a billable hour now the effect of um innovation in particular is a long understood problem with the GDP figures so here I've got a picture of living rooms in 1895 and 2015 and they look pretty similar don't they you've got chairs to sit on on uh you've got a flat screen TV in the Ikea catalog and a piano to stand around and sing in the 1895 version but they look quite similar but in many other ways our lives have been absolutely transformed over that period of course we've got electric lighting indoor plumbing we've got uh all the modern medicines mobile phones the internet and even very trivial Innovations never mind things like uh tamoxifen or mobile phones are really valued by consumers to give you a very trivial example nylon stockings were introduced uh by Dupont on the 15th of May 1940 in America and they sold 800,000 pairs on the first day and 64 million pairs by the end of the first year now those numbers speak to um some very happy female consumers that is uh a new product an Innovative product that was serving a very deeply felt need replacing the old the old kind of stockings when people talk about real GDP measures over time they sometimes don't quite understand what it is it's measuring so if I say to you in 1895 real GDP per capita um was $115,000 in today's money that does not mean having $15,000 today and being able to spend it it means having $15,000 today and being able to spend it only on products ail ailable in 1895 so you couldn't have decided that your mobile phone was your top priority and spent some of the money on that and these um these changes in um the quality of life the character of life through Innovations aren't captured in Real GDP measures statisticians change the base Year from time to time to capture new products and put them in there and that can lead to some quite dramatic effects Nigeria hadn't updated its Bas year figures since 1960s when it did so last year and it added I think it was 89% to its level of GDP overnight now of course nobody was any better off nobody there were there weren't fewer poor people um but they had um weights that assigned a weight of zero to Nollywood movies and the mobile phone sector and hadn't been counting all of that activity most um developed oecd statisticians update much more frequently they were doing every five years now they're doing it um annually with chain waiting they update that mix every year and this came about in 1995 when the Bureau of Labor Statistics in the US decided that because it was using 1987 prices at that time to Value the computer industry it was overvaluing it because prices of computers had come down so much and um therefore it would switch to updating the weights every year for everything to get away from these odd basee effects they then introduced what's called honic pricing as well which takes account of the quality characteristics of consumer goods um like computers if you buy the one for the same amount of money today that you did three years ago it'll have many better characteristics more speed better me more memory and better camera and so on um and that makes the measured prices for that sector even lower thing is that when you do that and this is a phenomenon that Simon kar's first pointed out in his work on GDP figures it looks like you've got very slow growth as a share of the economy in the most dynamic sectors where the prices are falling fastest and there's nothing um wrong about that statistically but it really offends people's intuition so all the guys in Silicon Valley look at the American figures and see that the share of the computer industry in GDP doesn't appear to have been going up and they know in their hearts that's wrong um it's that intuition that's offended so there are lots of measurement issues but there there's also the issue of consumer surplus that that that that joyful women able to wear proper nylons instead of rollup stockings um and that can be large consumer surplus is the gap between uh what how how much consumers value a new product or service what they would be willing to pay for it and what they actually have to pay in the market and if you've done the economics you know it's that chunk um above the market price and below the demand curve that can be large one of my favorite examples is Nathan Mar roosild richest man in the world when he died in 1836 of an AB infected abscess on his tooth and you ask yourself if antibiotics had been introduced in 1836 what would he have paid for them that would have been large consumer surplus um this trap is on the International Space Station now I don't know if you know this story but his spanner broke and it stopped his experiment and in the old days they would have had to launch a new rocket to send him a new spanner which costs tens of millions of dollars and uh they would always wait a few weeks to fill up the rocket with other things that they wanted to send up there as well so lots of time lots of lost experimentation and tens of millions of dollars this time they emailed him some software code and he printed it out on the 3D printer on board almost instantaneously so that too um is a qualitative change speaks to a very large increase in consumer surplus that isn't captured in GDP figures it never has been but there's a lot of innovation happening now and my indicator of the speed of innovation is the price declines and I've got two examples for you here the first one is uh by uh bill nordhouse is the price of computer processing power Mo law is still continuing this says it's a logarithmic scale so it's an so um this isn't a constant rate but an accelerating rate um very one of the most dramatic declines in the price of a new good that we have seen But beaten if you look at the right hand chart by the cost of sequencing a genome this is from the National Institutes of Health and the white line shows you the pace of decline implied by Moors law and this again is a logarithmic scale and you can see that there has been a much bigger decrease in prices for genome sequencing the MS law would imply and I could have and I should have because I from Manchester have put the graphine price um on here as well very similar p so across a range of Technologies absolutely astonishing price declines the GDP by definition doesn't include the consumer surplus that that's generating and I think this also casts a light on the secular stagnation debate which uses GDP figures to assess uh productivity and isn't paying attention to any potential changes any potential increase in consumer surplus and if you look at past new techn Technologies and look at how much they've contributed to GDP growth it looks really small people have looked at electrification they have looked at the introduction of the railways and found it very hard to identify big effects on GDP growth but electrification brought us mass production it brought the possibility of the assembly line the railways brought urbanization because food could be brought into populations and urban centers it didn't all have to be grown in in the city and the GDP me just doesn't try to and doesn't capture that so at a time when there's a lot of innovation that particular part of the wedge may be opening up a special form of innovation is variety and here's the yogurt section of a supermarket in New York uh variet is a kind of innovation um it's a a better match of people's specific preferences to the supply available in the market and nobody CS it so I've put on the left here a list of things that the Dallas Federal Reserve board counted in 1998 and I haven't found any subsequent attempt to count it and between the 70s and the late 90s they found uh a lot of U proliferation of variety in the particular products that they looked at there has certainly um been a lot more since then and we've got these kind of yogurt displays now there is an argument about the Paradox of choice that says that actually all this Choice makes us worse off and you should deduct from economic welfare and I think that's absolute nonsense it's based on experiments where you put um three kinds of jam in one bit of the shopping mall and 12 kinds of jam in the other bit of the shopping mall and people will be more likely to make a purchase if they got less to choose few fewer items to choose from they are befuddled by an excess of choice and that might well be true at an individual level it obviously there's a fallacy of composition it obviously doesn't aggregate because there's no reason my three favorite flavors would be the same as your three favorite flavors and you only have to see them to think about how would it be if we said um there's too much choice in book titles let's only sell the bestseller list you certainly wouldn't be able to buy my book in that case I'm afraid um or uh let's not bother introducing personalized medicines because that's too much choice for people there's you know clearly welfare from from this variety and that's not measured in GDP either now this isn't to say there isn't digital hype because there is there is lots and lots and lots of free stuff including the mar Manchester policy blogs which I hope you'll read as well as the lsse blogs um but if there's a zero price obviously the consumer surplus is large but a lot of these are the result of a business model choice and they're funded by adverts which consumers don't like and will even pay for ad blockers to stop and if you try to impute a cost of the adverts you find that actually there's a very little increase in economic welfare or they are cross subsidized by uh freemium and buy subscriptions from other people so there's a cross subsidy between different kinds of consumers going on there but some of it's true some of it's genuinely voluntary activity that's making people better off economically um so take for example this beauty vlogger many of you will know better than I do what a beauty vlogger does um but there are people who sit at home make videos post them to YouTube for free and a lot of people enjoy watching them but they will hope that they will sell adverts around those or they'll be asked to endorse products and they'll make some money out of it or there are people who will create creative common software but hope to sell consultancy services around it or they'll do a TED Talk and hope that will help them get onto the lecture circuit or sell more copies of their books so the boundary between these kinds of activities that are voluntary and paid work is starting to get a bit fuzzy and I would argue that as's a good case to include some of the free particularly things like free economic soft Ware or Wikipedia into an economic welfare measure and indeed we include a lot of government spending that um is is paid for um by taxes um which one might argue people object to as well um but if you're including writing software for free why not housework and as I said obviously this is how I dress when I do the housework um as I said that's always been excluded from GDP because it was thought to be too difficult um but now there are lots of Market Services competing with home production you can easily have somebody comeing to clean your house and pay them or um you can pay for child care if you want if both Partners want to go out to work and people are choosing a lot on that margin between home production and Market production and actually some of it if you make a product it's counted in GDP at least in theory because in many developing countries growing food or making clothes at home is an important part of economic activity so the products are included but the services are not so if you volunteer in the charity shop or the school or if you do services like cleaning and cooking that're they're not included but this margin is important and actually I think another margin is going to become more important as the population Ages which is substitution between home production of services um by a human being uh Market production of services by a human being whose cost is going to go up um and um production of services by a domestic robot domestic capital and these margins are very important policy Choice margins as well but we have a data blank on them uh the O owners has done one survey of how much time people spend on housew work they are due to publish another one next year but having them done every 10 years doesn't really help you get to these important choices and it will all become much more important as we move more into what's been called the gig economy this uh Beauty vlogging part-time or you have a job but you drive an Uber as your second job or or you rent out your spare room for Airbnb the variety um of uh availability of products in the digital economy I think leads to another important bit of economic welfare that we haven't counted before and that's matching and that's I'm thinking about um markets like exactly Airbnb and Uber they called various things sharing economy collaborative consumption but I think of it as being about matching this is an example some of you might know about it's Robert Jensen's well-known study of the introduction of mobile phone masks around in different ports around the coast of carola and the right hand chart shows the point at which the must was introduced in each of the areas and you can see that afterwards there's much less dispostion in prices and his study found that there was also an increase in Fisherman's incomes and a decrease in prices paid by consumers because there was better matching of demand and supply for particular fish the fishermen could phone ah head to each port and they wasted did less of the fish that they had caught because they would be more sure of selling them at the ports that they went to a pure efficiency gain and I think there's a lot we need to understand here in the um Airbnb type sharing economy markets um are people are they directly substituting for existing activities or are they extending the range of choices is Airbnb only a substitute for hotel rooms or is it actually adding kinds of accommodation and experience that weren't available otherwise so that people for example who who've got small children who don't want to stay in a hotel and don't have that kind of money will now take a trip that they wouldn't have done otherwise um so we have we have very little data on this so I've gone through a number of reasons why the wedge between GDP and economic welfare might be um increasing and there's one last one I want to talk about I'm probably out of time and that's Network effects um it kind of Echoes the environmental critique in reverse it's about an externality the environmentalist said you're excluding the important um environmental externalities which affect economic welfare but we're also excluding actually all other externalities too and these Network externalities are very important in the digital economy um you know GDP um rests on assumptions about markets that we know are not true in real life and particularly excludes all the traditional market failures it doesn't include any externalities and in the knowledge economy there are multiple externalities there's not just Network effects there is the fact that the social value of knowledge greatly exceeds the private value of knowledge that uh ideas are non-rival in use they are public good that they're often unexperienced good you don't know whether you like them until you've tried them and then the network externalities too so they quite a lot of reasons why you might think that in uh these kinds of markets there are multiple market failures and the gap between GDP and Welfare is getting bigger now there are lots and lots of other issues that I could have talked about I could have talked more about the importance of measuring assets another part of what environmentalists complain about uh famously a natural disaster increases GDP growth because we count the Reconstruction cost but we don't count the um the value of the of the Lost assets um I haven't talked about um globalization and whether or not having a national framework for these statistics is as useful as it as it used to be I think it's time for actually quite a big rethink about how we understand the economy and therefore how we count the economy and I want to end with this quote from John Hicks who took part in all the debates of the 1940s and 50s about what kind of measure we should be using um as economists this is the main product that we're selling to people it absolutely dominates political debate uh before the election earlier this year um all of the newspapers depending on their point of view we interpreting the GDP growth figures in their particular way we have a responsibility for the quality of the product and now I think it's time for a rethink now I've probably gone a little bit over but I hope we've got some time left for questions and comments thank you okay I was delighted to hear that um GDP misses out a lot of negative stuff from environmental problem and then it misses out a lot of positive stuff from uh variety Innovation Network effects so my conclusion that on average you get it right uh now before we go to uh to questions I I do want to say that Dian will be signing her book outside this room at the end if you are interested and you can contribute to the expenditure side measure of GDP by buying it um okay any questions there the lady with a black shirt hello um I just wanted to um um question you saying that GDP is anybody thought about by governments because they're looking at um welfare and I feel that GDP be measured to look at how you can um expect growth in the economy so you can spend now so there'll be further tax receipts in the future do you think that's relevant so GDP still matters obviously um the government cares a lot about the tax base and tax revenues and um that's one of the reasons actually the globalization and looking at an international framework for statistics is important because all the activities of of multinationals so I think that's certainly true and the origins of the national accounts way back in the 17th century were all about measuring how much tax the Monarch could raise to fight Wars successfully so I think that's correct um but I think I think the the growth of the GD the growth of GDP is a central part of the debate that we've been having and actually this statistic that matters has changed over time in the um 1980s it was the unemployment rate for a while it was baned payments for a while but there becomes some to some a single figure that becomes totemic for the public for the political debate and also for financial markets at the moment I think that is the quarterly change in GDP which of course is nonsense because the margin of errors are so large that they they swamp the quarterly figure and the bank of England chart projecting GDP actually backcast the uncertainty so at the moment they're 90% sure that growth is between one and a half and three and a half% which is quite a wide range there the back yeah uh I'm from India and uh from the skill development Ministry uh we use uh National statistics that is nsso which is quite related to the isso which is the international statistic uh for job classification so um so I was a little surprised that you know that we don't have standardization in uh job classifications because um what are we using if if there is no standardization and why can't the same isso be used for uh GDP calculation as well and why can't we keep that updated the the classifications do get updated but they get updated at um long intervals and there are just so few service sector categories compared to manufacturing categories in any of them and as I was saying a lot of these service sector jobs in the digital sector are are new and not standardized and actually you can't ask stat statisticians to develop a classification until you get to that point of being able to standardize them so I think that lag is is inevitable um but it means that I think we need to supplement the classifications with additional work on exactly what kinds of jobs people are doing conf front here how can we standardize the digital IND the kind of jobs digital industry for example considering the fact it moves so quickly um so for example the um it world of 20 years ago is completely different to as it is now it's virtually unrecognizable uh the rates of growth are astronomical in the size of the internet and the number of websites how can we measure it accurately well I don't think we can at the moment it's it's the same kind of replies to the one about standardizing the job classifications that that we just can't in fact when I um started doing econometrics seriously I had to program in Fortran to run regressions there was only one statistics package available and it didn't suit what I needed to do and I discovered last week that actually um NASA needs forra programmers because they can't find anybody to hire to update certain old bits of code they have a lot of the banks have a lot of code code written in fortron and cobal and they too can't find people to update that so um perhaps we oughtn't be too fuss in updating these things but I think it's just going it's just going to take time they're not standardized yet and it will take a while um hey there uh if we want GDP to be a more accurate measure of um economic welfare um is there an argument that there are some uh products or that we should discount from GDP measures based on judgments about the welfare they bring so I know you mentioned advertising was meant to be discounted originally yeah that's a very interesting question and I don't quite know what I think about that the um assumption or judgment that advertising is negative to Consumer welfare is um really well supported in the early literature uh Nikki kalor was another Economist who who was very clear that it harmed consumers to seek advertising what makes that particularly interesting is that governments are now Keen to use advertising techniques in implementing policies nudge policies and that's popular and it looks like it can have some good results makes me quite uneasy and um I think speaks also to this question that you raised about about the judgment on that but it's not just advertising there might be all kinds of things where people come to different judgments and and that's one of the arguments for not having a single measure but breaking down the statistics into a range of measures a dashboard of measures and I have some sympathy with that too trouble is that there are lots of dashboards and they've got lots and lots and lots of indicators in them and you could come to any conclusion you like from looking at them so I think we just have a lot of work to do to think about what do we actually mean by economic welfare and how do you make the value judgments more transparent than they are at the moment thank you if if GDP is underestimated then I would say that the problems of uh jobless growth or weight stagnation and Rising inequalities are also underestimated and they are much worse than we thought would you agree with that or not I would agree with it I don't know that I think GDP is is underestimated I just don't know what happens when you net out all of those different externalities and whether whether it's right to say it gets it about right um but I think looking at income distribution is pretty key to economic welfare there's always been a presumption in economics that you can separate the two questions you can separate the question of the level of activity and distribution and I would agree with you that you you that you can't uh you certainly shouldn't in practice and actually I don't think you can in he um if you um see GDP as problematical do you see another measure that could replace it or that would be um more suitable for instance the kind of work that um measures social inequality um as Thomas picky does and get an aggregate of that and use that um to capture um what welfare consists of rather than trying to to mechanically play around with um the problems that GDP causes currently causes I'm not ready yet to say there's one single measure that's better um if I would introduce just one change it would be to look at net national product rather than growth and to include in the depreciation that you do the depreciation of um natural capital and public infrastructure capital so if I had to do one thing that would be the one and I haven't talked about that this evening actually um but apart from that I I just don't think we're ready to go to an alternative isn't there a problem in that when we when different organizations use a measure of GDP it's kind of subjective on what they need it to do so a couple years back when the EU was calculating uh income for its constituent memb numbers they included a criminal activity and prostitution among them and lo and behold the UK's contribution to the EU went up after that um isn't there a problem that we should we haven't got like a harmonized index of GDP I think the problem arises because there's increasing what's called administrative use of GDP so things like the mustri rules say budget deficit as a share of GDP or IMF adjustment looks at issues as a share of GD uh measures as a share of GDP and that should make us really uneasy because there are so many judgments and so much uncertainty around the figures I mean the margin of error on the level of GDP is enormous so to pin um austerity measures for the Greek people to the budget deficit as a share of their GDP doesn't really concerns me a lot and actually the use of GDP relative to the output Gap in monetary policy uh when the output uh potential output is even more judgmental than than GDP itself so I think that's I think that's the problem really it's the way that they are used and of course you're right introducing the illegal activities gives us the Absurd position that we we try to estimate prostitution and I don't think the staticians are going around the red light districts with clipboards they get data from police forces so who knows what the quality of the data is and how much do you charge for that service Madam um but we're not we're not counting home production which is a really important part of of of economic activity and economic welfare so it is absurd but the most worrying thing is the way that the formal GDP figure is being used and you're quite right about the incent incentives to make it bigger to make your budget deficit ratio look smaller um I was very happy to hear you mention briefly natural Capital I'm just wondering I do quite a bit of work with on and treasury at the moment trying to look at how we might start incorporating natural Capital um and two big debates come from forth one is do we really want to include a capital measure in a flow account uh and when we value natural Capital we use non-market valuation techniques which capture economic welfare consumer surplus how well does that uh mesh with the sna processes um it seems to me it's not my area of expertise which is why I didn't talk about it I I it seems to me problematic to try to value a Capital stock so looking at changes would be the way to go um and well as you will know better than I those valuation techniques have their own problems and the question is really the the classic question is some number better than no number and I come down on yes some number is better than no number because it's clear what the effects of not paying any attention to Natural capital or natural Capital changes has been so far so we need to do something about it um what is the ideal GDP growth rate uh I don't have a number for you um people who say it's zero like um Tim Jackson caricaturing him a little and sorry um seem to me to forget that the really important thing about growth is innovation it's not having more handbags or more cars or even bigger houses it's actually what Innovations change the quality of life and I don't know that I actually want to put an upper bound on how much Innovation there is um and that makes it very hard to say and given all the measurement issues and the consumer surplus issues anyway it's very hard to translate that into a GDP growth figure so not zero but as we're struggling to get as high as two I think the answer is somewhere between zero and two can I ask a question about that because I always wondered I don't think you allow to ask questions are you sorry just one um so you you mention Innovation and Innovation means productivity improvements among other things and uh there is something like OK law used to be at least that links um uh unemployment changes to growth changes and what one I I always wonder whether one of the problems with the zero growth argument is that it implies unbounded increases in unemployment yeah and what do we do with all those people yeah um I think it I think that's right um but obviously the character of what we think of as work has changed over time so Innovation seem to bring uh new kinds of work in them and that's the experience every gen generation has that their m doesn't understand what job they do and that's because of Innovations so if you've got anything to do with computers your grandma has no idea and it looks like quite fun to her compared to working in the factory that she worked in so so yes I agree with that but I think work work itself continually gets redefined with Innovation um what's your view on positional Goods which in my opinion add little to social welfare but only really serve to exacerbate social hierarchies yeah you know um there's no whiteboard the the traditional quadrant of of goods um where there you've got rival Rous and non-rivalrous and excludable and non-excludable I kind of want to extend that to um the super non-rivalrous Goods with network externalities and the super rivalrous Goods that that are positional goods and so the question really is can you um frame that as an usefully as an ex ity that you could incorporate in you're thinking about economic welfare um there are people of course he would tax positional Goods but the thing is we never know what they are in advance and you are too young to remember that Norman Lamont as chancelor taxed mobile phones as a nasty positional Goods that yuppies were spending far too much money on in the 1983 budget and that looks pretty silly in retrospect but things that are obviously always going to be positional like large Country Estates I think taxing those makes a lot of sense I wanted to I wanted to uh defend the zero growth environmental position for for a moment um you say that the most important thing about that that that argument misses is that what's important about growth is innovation and I think the the argument looks at the way GDP is measured and all the all the things that you said in your talk about how GDP misses how Innovation makes life better and then um says okay well we can still have Innovation with the zero rate of growth and we could maybe even still have relatively good employment or as good as we have now and and work less and I was wondering if um you think that we need a number to to grow we need a a a a number a way of taking into account these innovations that make life better and and including them in our GDP and and growing that or if we could say that these things aren't aren't measurable um I think the argument makes a lot more sense if it's reformulated in some way and doing it doing it in terms of GDP gives you the kind of absurdity that if somebody has a fabulous new innovation you've got to take something else away to preserve the zero growth but if you can do it in a way um that looks at not the total material requirement but the resource footprint and if you formulate it as the resource footprint of economic activity needs to not grow then that strikes me as a much more persuasive kind of argument but that's not typically how it's presented um hi um I was just thinking that GDP is only the numerator in the measuring of living standards and what real really ought to be concerned about as GDP per capita and isn't it just as difficult especially now um to measure the denominator and I just wonder you know whe whether we ought to be thinking just as hard about what we count as because after all we could in we could lose GDP if Scotland went independent or we could lose um English G you know UK GDP could shrink quite easily or we if we annexed um the Republic of Ireland uh we could grow it immediately but there is a whole return to the Fine Days of British Empire indeed yes which I just about remember some of them um but I I just thinking that with all the movement of people that's going on and um it's becoming increasingly hard just to Simply count the denominator as well you're absolutely right that we should be thinking about it in per capita terms to think about economic welfare and I should have said so um is an obvious Omission um on the migration advisory committee when we were looking at economic impacts of the migration policy we asked the Home Secretary specifically which people ought to be included to calculate GDP per capita and she was very clear that it was people already in the UK but you're absolutely right that there's a a moral judgment and an economic judgment to be made about which people you counting well there are errors there are errors in the census but we do have a census and it's thought to be the best way counting people maybe a couple more and I think we're losing our audience bit by bit I shouldn't have said that anyway I should say I'm going to um stick around for a bit outside if people want to talk afterwards and if anybody is watching in the overspill room and wants to send questions by the Twitter hashtag I will I will respond to them in due course remind me yeah so um you talked about uh sustainability and environmental concerns earlier and I was wondering um do you think that governments are too concerned with their own political objectives than to affect a structural change in the economy in order to achieve sustainability because it's a more longterm change than we think and therefore we won't be achieving it in the short term so for example the conservatives said that when we come into power we're going to make Britain Britain less green we're going to not allow a new runway in Heath for example but um now they're getting rid of subsidies on wind farms and uh yeah there's obviously a huge Collective action problem about tackling environmental policies because um The Horizon over which things happen is much longer than any individual government it's not just our government it's all all democratically elected governments have shorter terms than the time Horizon of the phenomena that that we're dealing with um and that's why it's so hard to get these International agreements I did listen to Thomas shelling talk about this one time as a game theory problem and he said he thought a big part of the problem was that the focal point of the game was something that was really hard to measure and could only be measured long in retrospect so the two degree warming um there's some disagreement about how to measure it at all and you only get the figures long after the actions um that affect them so his argument was that you can help that Collective action problem that you're talking about by switching to a different different kind of of um mechanism for all these International agreements okay I think we can call it the night for now we please join me in thanking Diana again for a